setop
/
dataviz
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
dataviz/docs/data-visualization-guide.pd...

2642 lines
196 KiB
HTML
Raw Normal View History

pdf
2021-04-11 12:56:38 +00:00
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<title>Data Visualization Guide for Presentations, Reports, and Dashboards</title>
<link rel="stylesheet" type="text/css" href="pdf.css">
<script src="https://unpkg.com/pagedjs/dist/paged.polyfill.js"></script>
</head>
<body>
<div class="cover-page">
<h1>Data Visualization Guide</h1>
<h3 class="cover-header">
IBCS Association
<br>Anton Zhiyanov
</h3>
<h3 class="cover-header">CC BY-SA 4.0</h3>
</div>
<div class="separate-page">
<h1 id="data-visualization-guide-for-presentations-reports-and-dashboards">Data Visualization Guide for
Presentations, Reports, and Dashboards</h1>
<p><em>Based on <a href="https://www.ibcs.com/standards/">IBCS Standards</a> 1.1 by
<a href="https://www.ibcs.com/">IBCS Association</a>, licensed under <a
href="https://creativecommons.org/licenses/by-sa/4.0/">CC BY-SA 4.0</a>. Adapted for the web and
other
formats by <a href="https://antonz.org/">Anton Zhiyanov</a>.</em>
</p>
<p>This is a highly practical and example-based guide on visually representing data in reports and dashboards.
It is
based on the work of authors such as Barbara Minto, Edward Tufte, and Stephen Few.</p>
<p>The guide consists of seven chapters:</p>
<ol>
<li>Convey a message.</li>
<li>Organize content.</li>
<li>Choose proper visualization.</li>
<li>Avoid clutter.</li>
<li>Increase information density.</li>
<li>Ensure visual integrity.</li>
<li>Apply semantic notation.</li>
</ol>
<p>Applied together, they will help you to design concise, clear, and actionable reports.</p>
</div>
<div class="page-break"></div>
<h2 id="say--convey-a-message">SAY – Convey a message</h2>
<p>SAY covers all aspects of conveying messages to the recipients of reports and presentations.</p>
<p><em>Conveying messages</em> means that reports and presentations, both as a whole as well as within their
individual components, intend to say something to the recipients. Messages in this sense can be determinations,
explanations, clarifications, recommendations, and other forms of statements.</p>
<p>This chapter covers introducing, delivering, supporting, and summarizing messages with respect to the objectives
of senders and receivers.</p>
<ol>
<li><a href="#sa-1-know-objectives">Know objectives</a></li>
<li><a href="#sa-2-introduce-message">Introduce message</a></li>
<li><a href="#sa-3-deliver-message">Deliver message</a></li>
<li><a href="#sa-4-support-message">Support message</a></li>
<li><a href="#sa-5-summarize-message">Summarize message</a></li>
</ol>
<h3 id="sa-1-know-objectives">SA 1 Know objectives</h3>
<p>Good reports (presentations) successfully achieve both the goals of the writer (speaker) and of the readers
(audience).</p>
<h3 id="sa-11-know-own-goals">SA 1.1 Know own goals</h3>
<figure><img src="img/sa-1.1.png" alt="Figure SA 1.1: Know own goals">
<figcaption>Figure SA 1.1: Know own goals</figcaption>
</figure>
<p>Do not start creating a report or presentation without a clear vision of what to achieve with it. The least goal
is to inform about an interesting detection. A higher goal is to make the reader (audience) understand a problem
by explaining it. The ultimate goal is to get a decision on a suggestion provided and to cause corresponding
actions.</p>
<h3 id="sa-12-know-target-audience">SA 1.2 Know target audience</h3>
<figure><img src="img/sa-1.2.png" alt="Figure SA 1.2: Know target audience">
<figcaption>Figure SA 1.2: Know target audience</figcaption>
</figure>
<p>A good report (presentation) will try to answer the questions of the readers (audience). So it is important to
know the target audience (e.g. their function, position, network, knowledge, experience, attitude, behavior,
worries, cultural background) and their goals, preferences, and expectations.  Do they only want to get informed
about interesting detections, or are they looking for an explanation to a problem? Are they willing to make
decisions and to act accordingly? Who might object to the message and why?</p>
<h3 id="sa-2-introduce-message">SA 2 Introduce message</h3>
<p>The addressees appreciate an introduction mapping the actual situation followed by an explanation of the given
problem. Raising a question will focus on the given message.</p>
<h3 id="sa-21-map-situation">SA 2.1 Map situation</h3>
<figure><img src="img/sa-2.1.png" alt="Figure SA 2.1: Map situation">
<figcaption>Figure SA 2.1: Map situation</figcaption>
</figure>
<p>Mapping the situation means compiling and presenting the related facts. Be sure to cover all relevant aspects and
obtain a general consensus concerning the facts. In general, this means not yet describing the given problem but
presenting facts and goals already known to the reader or audience. It is advisable to begin with a positive and
generally accepted description of the situation in order to prevent early contradictions.</p>
<h3 id="sa-22-explain-problem">SA 2.2 Explain problem</h3>
<figure><img src="img/sa-2.2.png" alt="Figure SA 2.2: Explain problem">
<figcaption>Figure SA 2.2: Explain problem</figcaption>
</figure>
<p>After mapping the situation, introduce the challenge or complication, affecting the reader or the audience. It
should make everyone aware of an interesting, critical, or even dangerous problem.</p>
<h3 id="sa-23-raise-question">SA 2.3 Raise question</h3>
<figure><img src="img/sa-2.3.png" alt="Figure SA 2.3: Raise question">
<figcaption>Figure SA 2.3: Raise question</figcaption>
</figure>
<p>A good introduction raises the relevant question from the perspective of the recipient of how to solve the
complication in the described situation. The question at the beginning of each report or presentation then leads
to the message, i.e. the answer to the question.</p>
<h3 id="sa-3-deliver-message">SA 3 Deliver message</h3>
<p>Delivering the message means answering the question asked at the end of the introduction. Messages detect,
explain, or suggest something the report or presentation later explains in detail.</p>
<h3 id="sa-31-detect-explain-or-suggest">SA 3.1 Detect, explain, or suggest</h3>
<figure><img src="img/sa-3.1.png" alt="Figure SA 3.1: Detect, explain, or suggest">
<figcaption>Figure SA 3.1: Detect, explain, or suggest</figcaption>
</figure>
<p>Messages in reports and presentations can detect, evaluate, explain, warn, complain, threaten, excuse, suggest,
or recommend something interesting. Make sure to deliver these messages in a complete sentence in order to be
understood.</p>
<p>Today, many messages in business reporting are pure <em>detections</em>. Since detections are statements that can
be checked whether they are true or false, they should be formulated as precisely as possible.</p>
<p>Explaining the reasons for a detection (<em>explanation</em>) or even deriving a <em>suggestion</em> on how to
solve the problem or at least on how to further proceed can add value.</p>
<figure><img src="img/sa-3.1-1.png" alt="Figure SA 3.1.1: Classification of messages">
<figcaption>Figure SA 3.1.1: Classification of messages</figcaption>
</figure>
<p>This figure shows a classification of messages with examples from the business environment (Source: Hichert, R.
and Kornwachs, K.)</p>
<h3 id="sa-32-say-message-first">SA 3.2 Say message first</h3>
<figure><img src="img/sa-3.2.png" alt="Figure SA 3.2: Say message first">
<figcaption>Figure SA 3.2: Say message first</figcaption>
</figure>
<p>Every report, every presentation, and every single page or exhibit can be summed up with a clear overall message.
This message usually comes first and is proven afterwards. For the readers or the audience it is more difficult
to follow the storyline if the message comes at the end.</p>
<p>Be cautious applying this rule in presentations (not in reports) with bad, unexpected, or unpleasant messages
(e.g. layoffs) or in a cultural environment, where directness is considered impolite.</p>
<h3 id="sa-4-support-message">SA 4 Support message</h3>
<p><em>Supporting the message</em> covers some technical and practical aspects of message conveyance.</p>
<h3 id="sa-41-provide-evidence">SA 4.1 Provide evidence</h3>
<figure><img src="img/sa-4.1.png" alt="Figure SA 4.1: Provide evidence">
<figcaption>Figure SA 4.1: Provide evidence</figcaption>
</figure>
<p>Substantiate the message in order to prove the message by facts and figures. If possible, a presentation slide
should itself explain or prove the speaker’s message and not – as very often seen in practice – be explained by
the speaker. This can be done by spoken sentences possibly supported by charts, tables, and pictures.</p>
<h3 id="sa-42-use-precise-words">SA 4.2 Use precise words</h3>
<figure><img src="img/sa-4.2.png" alt="Figure SA 4.2: Use precise words">
<figcaption>Figure SA 4.2: Use precise words</figcaption>
</figure>
<p>The more unambiguous the language, the clearer the message. Only precise words will be understood. Speaking about
“relevant” or “significant” (in common speech, not as a statistical term) content leads to misinterpretations
and misunderstandings. Speaking about facts and figures will prevent them.</p>
<h3 id="sa-43-highlight-message">SA 4.3 Highlight message</h3>
<figure><img src="img/sa-4.3.png" alt="Figure SA 4.3: Highlight message">
<figcaption>Figure SA 4.3: Highlight message</figcaption>
</figure>
<p>Visually highlight messages in the communication objects presented – namely in charts, tables, graphs, and
pictures. This facilitates comprehension and reduces the time needed to understand complex situations. In most
cases, it should be possible to highlight the important parts of the content by underlining the most important
facts or emphasizing interesting details. Objects and pages without highlighting indicators tend to be a
statistic rather than a report.</p>
<h3 id="sa-44-name-sources">SA 4.4 Name sources</h3>
<figure><img src="img/sa-4.4.png" alt="Figure SA 4.4: Name sources">
<figcaption>Figure SA 4.4: Name sources</figcaption>
</figure>
<p>Naming sources for the material presented increases the credibility. Projected slides can omit them but written
reports and handouts must include them.</p>
<h3 id="sa-45-link-comments">SA 4.5 Link comments</h3>
<figure><img src="img/sa-4.5.png" alt="Figure SA 4.5: Link comments">
<figcaption>Figure SA 4.5: Link comments</figcaption>
</figure>
<p>Use comments in written reports and handouts to add explanations, conclusions, and similar statements. Projected
slides in presentations rarely need any comments because the comments are given by the speaker.</p>
<p>Number comments related to specific parts of a page (e.g. words, numbers, or visualization elements) and link
them to the respective parts. Post numbered comments in text boxes on free areas of a page. General comments
concerning the whole page are not numbered. Post them as a footnote at the bottom of a page.</p>
<h3 id="sa-5-summarize-message">SA 5 Summarize message</h3>
<p>Conclude a presentation with the overall message, including the next steps and an explanation of the
consequences.</p>
<h3 id="sa-51-repeat-message">SA 5.1 Repeat message</h3>
<figure><img src="img/sa-5.1.png" alt="Figure SA 5.1: Repeat message">
<figcaption>Figure SA 5.1: Repeat message</figcaption>
</figure>
<p>Avoid the phrase “Thank you for your attention” at the end of a presentation. Instead, presenters should briefly
sum up their message one last time – in one sentence, if possible. At the conclusion of a successful
presentation, the audience will be thanking the presenters for the information. Repeating the message from the
beginning of a presentation at the end helps the audience check the quality of the storyline and brings the
presentation full circle. In reports, on the other hand, such repetition is not necessary as the reader can
quickly browse back to the respective summary at the beginning.</p>
<h3 id="sa-52-explain-consequences">SA 5.2 Explain consequences</h3>
<figure><img src="img/sa-5.2.png" alt="Figure SA 5.2: Explain consequences">
<figcaption>Figure SA 5.2: Explain consequences</figcaption>
</figure>
<p>Conclude reports and presentations with proposals for decisions to be taken and an explanation of their
consequences. This is the real objective of a presentation: Convince the audience of both the message and the
suggested steps to be taken next.</p>
<h2 id="structure--organize-content">STRUCTURE – Organize content</h2>
<p>STRUCTURE covers all aspects of organizing the content of reports and presentations.</p>
<p><em>Organizing the content</em> means that reports and presentations follow a logical structure forming a
convincing storyline.</p>
<p>This chapter covers using consistent elements, building non-overlapping elements, building collectively
exhaustive elements, building hierarchical structures, and visualizing their structure properly.</p>
<ol>
<li><a href="#st-1-use-consistent-elements">Use consistent elements</a></li>
<li><a href="#st-2-build-non-overlapping-elements">Build non-overlapping elements</a></li>
<li><a href="#st-3-build-collectively-exhaustive-elements">Build collectively exhaustive elements</a></li>
<li><a href="#st-4-build-hierarchical-structures">Build hierarchical structures</a></li>
<li><a href="#st-5-visualize-structure">Visualize structure</a></li>
</ol>
<h3 id="st-1-use-consistent-elements">ST 1 Use consistent elements</h3>
<p>Listings and groupings of any kind of elements (items, terms, pictures, symbols, etc.) used to organize content
in charts, tables, and texts should contain consistent elements only. This pertains for example to items,
statements, wordings, and the appearance of symbols and pictures.</p>
<h3 id="st-11-use-consistent-items">ST 1.1 Use consistent items</h3>
<figure><img src="img/st-1.1.png" alt="Figure ST 1.1: Use consistent items">
<figcaption>Figure ST 1.1: Use consistent items</figcaption>
</figure>
<p>Items in a group should be of the same type, i.e. consistent. Consistent items can be different types of cars,
houses, traffic signs, or – as shown in Figure ST 1.1, on the right hand side – different national flags
representing the corresponding nations. The left hand side of this figure includes other types of items besides
national flags, destroying the consistency.</p>
<h3 id="st-12-use-consistent-types-of-statements">ST 1.2 Use consistent types of statements</h3>
<figure><img src="img/st-1.2.png" alt="Figure ST 1.2: Use consistent types of statements">
<figcaption>Figure ST 1.2: Use consistent types of statements</figcaption>
</figure>
<p>A list of statements will be easier to understand if all statements are of the same type. The right hand side of
Figure ST 1.2 shows four suggestions. By contrast, on the left-hand side of this figure the third statement is a
detection, not a suggestion.</p>
<h3 id="st-13-use-consistent-wording">ST 1.3 Use consistent wording</h3>
<figure><img src="img/st-1.3.png" alt="Figure ST 1.3: Use consistent wording">
<figcaption>Figure ST 1.3: Use consistent wording</figcaption>
</figure>
<p>Structure all phrases – especially in listed arrangements – in a grammatically consistent manner to facilitate
quicker understanding. The right hand side of Figure ST 1.3 shows a group of four consistent suggestions, an
imperative verb paired with a noun. By contrast, on the left hand side of this figure the second suggestion uses
verbal substantive instead of an imperative.</p>
<h3 id="st-14-use-consistent-visualizations">ST 1.4 Use consistent visualizations</h3>
<figure><img src="img/st-1.4.png" alt="Figure ST 1.4: Use consistent visualizations">
<figcaption>Figure ST 1.4: Use consistent visualizations</figcaption>
</figure>
<p>Visualizations such as symbols and pictures that are uniform in respect to their layouts, colors, forms, fonts,
etc. – especially in listed arrangements – facilitate faster and easier comprehension.</p>
<h3 id="st-2-build-non-overlapping-elements">ST 2 Build non-overlapping elements</h3>
<p>Elements belonging to a group should not overlap, i.e. they should be disjoint or mutually exclusive. This
concerns practical applications such as report structures, business measures, or structure dimensions.</p>
<h3 id="st-21-build-non-overlapping-report-structures">ST 2.1 Build non-overlapping report structures</h3>
<figure><img src="img/st-2.1.png" alt="Figure ST 2.1: Build non-overlapping report structures">
<figcaption>Figure ST 2.1: Build non-overlapping report structures</figcaption>
</figure>
<p>Structure reports and presentations in such a way that the parts, chapters, sections, and paragraphs do not
overlap. They should not cover the same aspects.</p>
<p>In Figure ST 2.1, on the left hand side, the following chapters of a project description overlap:</p>
<ul>
<li>expenses and costs</li>
<li>schedule, steps, milestones, and calendar</li>
<li>objective, results, and achievements</li>
</ul>
<p>At first glance, the six terms on the right hand side of this figure have no overlap in their logical structure.
Of course, a relationship exists between the <em>cost</em>, the <em>results</em>, and the <em>schedule</em> of a
project, but in regards to the content of the chapters this is not an overlap.</p>
<h3 id="st-22-build-non-overlapping-business-measures">ST 2.2 Build non-overlapping business measures</h3>
<figure><img src="img/st-2.2.png" alt="Figure ST 2.2: Build non-overlapping business measures">
<figcaption>Figure ST 2.2: Build non-overlapping business measures</figcaption>
</figure>
<p>Structure a group of business measures in lists or calculations in a way they do not overlap, i.e. business
measures on one hierarchical level should be disjoint or mutually exclusive.</p>
<p>Looking at Figure ST 2.2, on the left hand side, the following business measures overlap</p>
<ul>
<li><em>material costs</em> and <em>costs of goods sold</em></li>
<li><em>depreciation</em> and <em>fixed costs</em></li>
</ul>
<p>The calculation scheme on the right hand side has been cleaned up.</p>
<h3 id="st-23-build-non-overlapping-structure-dimensions">ST 2.3 Build non-overlapping structure dimensions</h3>
<figure><img src="img/st-2.3.png" alt="Figure ST 2.3: Build non-overlapping structure dimensions">
<figcaption>Figure ST 2.3: Build non-overlapping structure dimensions</figcaption>
</figure>
<p>The elements of the <em>structure dimensions</em> used in reports and presentations should not overlap, i.e. the
elements of a structure dimension should be disjoint or mutually exclusive.</p>
<p>Looking at Figure ST 2.3 on the left hand side, the regions <em>Norway, Sweden, Denmark,</em> and
<em>Finland</em>
overlap with <em>Scandinavia</em>.
</p>
<h3 id="st-3-build-collectively-exhaustive-elements">ST 3 Build collectively exhaustive elements</h3>
<p>A list of elements is considered to be exhaustive when they cover all aspects of a superordinate topic. For
example, dividing <em>Europe</em> into <em>Germany</em>, <em>Austria</em>, <em>Switzerland</em>, and
<em>Belgium</em>
is not exhaustive because other countries also belong to Europe.
</p>
<p>Structures with mutually exclusive (ME) and collectively exhaustive (CE) elements are known as MECE structures.
</p>
<h3 id="st-31-build-exhaustive-arguments">ST 3.1 Build exhaustive arguments</h3>
<figure><img src="img/st-3.1.png" alt="Figure ST 3.1: Build exhaustive arguments">
<figcaption>Figure ST 3.1: Build exhaustive arguments</figcaption>
</figure>
<p>If some important arguments relating to a specific question are left out, the given answer will not be
convincing.</p>
<p>Looking at Figure ST 3.1 on the left hand side the option “<em>old products, new location</em>” is missing.</p>
<h3 id="st-32-build-exhaustive-structures-in-charts-and-tables">ST 3.2 Build exhaustive structures in charts and
tables</h3>
<figure><img src="img/st-3.2.png" alt="Figure ST 3.2: Build exhaustive structures in charts andtables">
<figcaption>Figure ST 3.2: Build exhaustive structures in charts andtables</figcaption>
</figure>
<p>The elements of structures presented in charts and tables should also be exhaustive, in other words, adding up to
one hundred percent.</p>
<p>In many practical applications of this kind, adding a remainder element (“rest of...”) helps to conform to this
rule.</p>
<h3 id="st-4-build-hierarchical-structures">ST 4 Build hierarchical structures</h3>
<p>Give reports and presentations a hierarchical structure whenever possible, resulting in faster comprehension and
simplified searching. These rules help to write and present a good storyline.</p>
<h3 id="st-41-use-deductive-reasoning">ST 4.1 Use deductive reasoning</h3>
<figure><img src="img/st-4.1.png" alt="Figure ST 4.1: Use deductive reasoning">
<figcaption>Figure ST 4.1: Use deductive reasoning</figcaption>
</figure>
<p>Exhibiting deductive reasoning (<em>logical flow</em>) for a given message aids in <em>building</em> hierarchical
structures. <em>Logical flows</em> always answer the question “why” following the key message. They begin with a
statement (all men are mortal), continue with a comment (Socrates is a man), and resolve with a conclusion
(Socrates is mortal) culminating in the message (Socrates will die).</p>
<p>Deductive reasoning can be best applied in controversial discussions for arguing and demonstrating need for
action. However, it forces the readers or the audience to reproduce the deduction and the whole argumentation
can collapse if any statements are questionable.</p>
<h3 id="st-42-use-inductive-reasoning">ST 4.2 Use inductive reasoning</h3>
<figure><img src="img/st-4.2.png" alt="Figure ST 4.2: Use inductive reasoning">
<figcaption>Figure ST 4.2: Use inductive reasoning</figcaption>
</figure>
<p>Exhibiting <em>inductive</em> reasoning (<em>logical group</em>) for a given message aids in understanding
hierarchical structures. <em>Logical groups</em> are homogenous, non-overlapping, and collectively exhaustive
arguments culminating in a message. This results in a powerful argumentation that satisfies the addressees need
for an easily comprehensible logical structure.</p>
<h3 id="st-5-visualize-structure">ST 5 Visualize structure</h3>
<p>Having organized the arguments hierarchically, visualize this structure in order to make the storyline
transparent.</p>
<h3 id="st-51-visualize-structure-in-reports">ST 5.1 Visualize structure in reports</h3>
<figure><img src="img/st-5.1.png" alt="Figure ST 5.1: Visualize structure in reports">
<figcaption>Figure ST 5.1: Visualize structure in reports</figcaption>
</figure>
<p>For easier understanding, underscore the logical structure of reports and presentations with visual aids (e.g.
outlines, dashboards, summaries). Figure ST 5.1 illustrates this rule showing binder tabs on the right hand
side.</p>
<h3 id="st-52-visualize-structure-in-tables">ST 5.2 Visualize structure in tables</h3>
<figure><img src="img/st-5.2.png" alt="Figure ST 5.2: Visualize structure in tables">
<figcaption>Figure ST 5.2: Visualize structure in tables</figcaption>
</figure>
<p>Design tables in such a manner that their hierarchical structure can be recognized in both the columns as well as
the rows.</p>
<p>The right hand side of Figure ST 5.2 shows three hierarchical levels of rows in a table. The base level shows
cities, the first summary shows regions, and the second summary shows the country.</p>
<h3 id="st-53-visualize-structure-in-notes">ST 5.3 Visualize structure in notes</h3>
<figure><img src="img/st-5.3.png" alt="Figure ST 5.3: Visualize structure in notes">
<figcaption>Figure ST 5.3: Visualize structure in notes</figcaption>
</figure>
<p>Notes are also easier to understand when their structure is shown clearly (see Figure ST 5.3).</p>
<h2 id="express--choose-proper-visualization">EXPRESS – Choose proper visualization</h2>
<p>EXPRESS covers all aspects of choosing the proper visualization in reports and presentations.</p>
<p><em>Proper visualization</em> means that reports and presentations contain charts and tables, which convey the
desired message along with the underlying facts as quickly as possible.</p>
<p>This chapter covers utilizing the correct types of charts and tables, replacing inappropriate visualizations and
representations, adding comparisons, and explaining causes.</p>
<ol>
<li><a href="#ex-1-use-appropriate-object-types">Use appropriate object types</a></li>
<li><a href="#ex-2-replace-inappropriate-chart-types">Replace inappropriate chart types</a></li>
<li><a href="#ex-3-replace-inappropriate-representations">Replace inappropriate representations</a></li>
<li><a href="#ex-4-add-comparisons">Add comparisons</a></li>
<li><a href="#ex-5-explain-causes">Explain causes</a></li>
</ol>
<h3 id="ex-1-use-appropriate-object-types">EX 1 Use appropriate object types</h3>
<p>Choosing the appropriate <em>object type</em> is of prime importance for the comprehension of reports and
presentations.</p>
<p>We use tables when looking up data. Tables have a high information density. They are clear, they are honest, they
do not want to highlight, and they typically do not want to visually convey a certain message. So they do not
compete with charts.</p>
<p>Charts on the opposite are always biased. It is the selection of data, the selection of the chart type, and the
usage of highlighting which makes the difference. We evaluate charts by asking whether they transfer the
intended message effectively and in a proper way. So charts cannot be replaced by tables.</p>
<p>The following section is about choosing the appropriate types of charts and tables. It presents in detail
different types, layouts, and examples as well as their proper application.</p>
<h3 id="ex-11-use-appropriate-chart-types">EX 1.1 Use appropriate chart types</h3>
<figure><img src="img/ex-1.1.png" alt="Figure EX 1.1: Use appropriate chart types">
<figcaption>Figure EX 1.1: Use appropriate chart types</figcaption>
</figure>
<p>A <em>chart</em> is a graphical object, in which visualization elements such as columns, bars, and lines
represent
data.</p>
<p>This section discusses the types, layout, and examples of <em>single charts</em>. <em>Overlay charts</em><em>and
multiple charts</em> are discussed in the CO 4 “<a href="#co-4-add-elements">Add elements</a>” and CO 5 “<a
href="#co-5-add-objects">Add objects</a>”.</p>
<p>The most important groups of business charts are those showing development over time (charts with horizontal
category axes), those showing structural relationships (charts with vertical category axes), and those showing
x‑y charts, scatter plots, and bubble charts (charts with two value axes), see Figure EX 1.1.</p>
<p>Other chart types are of lesser interest in business communication and will be treated in a later version of the
standards.</p>
<figure><img src="img/ex-1.1-1.png" alt="Figure EX 1.1-1: Chart Types">
<figcaption>Figure EX 1.1-1: Chart Types</figcaption>
</figure>
<p>Looking at charts with horizontal and vertical category axes, the chart selection matrix displayed in the figure
aids in selecting the right chart type for time series and structure analyses.</p>
<p>In the following sections, the correct usage of <em>charts with horizontal category axes</em>, <em>charts with
vertical category axes</em>, and <em>charts with two value axes</em> is discussed in greater detail.</p>
<p><strong>Charts with horizontal category axes</strong></p>
<p>Charts with horizontal category axes (short:<em>horizontal charts</em>) typically display time series. Use the
horizontal category axis as a time axis. Vertically, the visualization elements represent the data per time
period or point of time (there is no need to show a vertical value axis as the visualization elements carry
their own values). Time category axes run from left to right and show characteristics of period types (e.g.
months or years) or points of time (dates).</p>
<p>In general, the data series of a <em>horizontal chart</em> is represented by columns (single, stacked, grouped),
vertical pins, horizontal waterfalls, or lines. <em>Vertical pins</em> can be considered very thin columns.
Because of their importance, they are dealt with in a separate section.</p>
<p>Here follows the grouping of <em>horizontal chart types</em>:</p>
<p><strong>Single column charts</strong></p>
<figure><img src="img/ex-1.1-2.png" alt="Figure EX 1.1-2: Single column charts">
<figcaption>Figure EX 1.1-2: Single column charts</figcaption>
</figure>
<p>In general, <em>single column charts</em> (short: single columns) are used to display the temporal evolvement of
one data series.</p>
<p>Single columns consist of:</p>
<ul>
<li>
<p><strong>Horizontal category axis:</strong> The <em>horizontal category axis</em> represents with its
labels the respective time periods or points of time. The part on “Semantic rules” suggests to use the
category width (see width A in the figure) for identifying the period type (see UN 3.3 “<a
href="#un-33-unify-time-periods-and-points-of-time">Unify time periods</a>”).</p>
</li>
<li>
<p><strong>Columns</strong>: One <em>column</em> per time period or point of time extends from the category
axis in accordance with the respective value. Columns are displayed in the foreground of the category
axis, so that the length of the column is not hidden. The part on “Semantic rules” suggests that the
ratio of column width to category width (see ratio B/A in the figure) represents information about the
measure type (see UN 3.1 “<a href="#un-31-unify-measures">Unify measures</a>”).</p>
</li>
<li>
<p><strong>Legends</strong>: As there is only one data series, the legend (name of the data series) is part
of the chart title.</p>
</li>
<li>
<p><strong>Data labels</strong>:<em>Data labels</em> name the values of the data series corresponding to the
length of the respective columns. Position the labels of positive values above their respective columns,
the labels of negative values below.</p>
</li>
</ul>
<p><strong>Stacked column charts</strong></p>
<figure><img src="img/ex-1.1-3.png" alt="Figure EX 1.1-3: Stacked column charts">
<figcaption>Figure EX 1.1-3: Stacked column charts</figcaption>
</figure>
<p><em>Stacked column charts</em> (short: stacked columns) represent more than one data series (e.g. multiple
products, countries, or divisions), see the figure on the left.</p>
<p>Stacked columns consist of:</p>
<ul>
<li>
<p><strong>Horizontal category axis:</strong> See single column charts.</p>
</li>
<li>
<p><strong>Columns</strong>: The columns (see single column charts) are divided into segments (Excel names
them “data points”) representing the data series (stacked columns).</p>
</li>
<li>
<p><strong>Legends</strong>: Legends (names of the data series) are positioned either on the far left side
with right-aligned text or on the far right side with left-aligned text. The column segments define
their vertical position, centered vertically with the data labels of the respective column segment. If a
segment at the far left or far right is missing or has a very small size, its legends might need
assisting lines.</p>
</li>
<li>
<p><strong>Data labels</strong>:<em>Data labels</em> name the values of the data series corresponding to the
length of the respective column segments. If the sum of the column segments of a category is positive
(column pointing upward), the label of the sum is positioned above the respective column, if negative
(column pointing downward), it is positioned below.</p>
</li>
</ul>
<p>It must be pointed out that stacked columns should only be used if all chart values are either positive or
negative.</p>
<p>This chart type might also not be a good choice if the values of each data series vary too much. The maximum
number of data series (segments of a stacked column) to be shown depends on the range of how much the values of
each data series vary: More than 5 data series will only work well in the case of little variations.</p>
<p>Position the data series of central importance or interest directly on the axis in order to best see its
development over time.</p>
<p><strong>Grouped column charts</strong></p>
<figure><img src="img/ex-1.1-4.png" alt="Figure EX 1.1-4: Grouped column charts">
<figcaption>Figure EX 1.1-4: Grouped column charts</figcaption>
</figure>
<p><em>Grouped column charts</em> (short: grouped columns) show, in general, time series for a primary scenario
(e.g. AC or FC) in comparison with a reference scenario (e.g. PY or PL). Two columns per category (<em>grouped
columns</em>) represent these two scenarios.</p>
<p>The columns of the primary scenario and the reference scenario overlap, the reference scenario placed behind the
primary scenario – either to the left or right of the primary scenario (see bottom chart of the figure as well
as the paragraph on ”Scenario comparisons” in UN 4.1 “<a href="#un-41-unify-scenario-analyses">Unify scenario
analyses</a>”). A third scenario could be displayed using a <em>reference scenario triangle</em>. All other
elements of a grouped column chart are identical to single column charts.</p>
<p>Instead of using grouped columns, the primary scenario can be represented with a single column with the reference
scenario represented by reference scenario triangles (see top chart of the figure).</p>
<p><strong>Horizontal pin charts</strong></p>
<figure><img src="img/ex-1.1-5.png" alt="Figure EX 1.1-5: Horizontal pin charts">
<figcaption>Figure EX 1.1-5: Horizontal pin charts</figcaption>
</figure>
<p><em>Horizontal pin charts</em> (short: horizontal pins) are used for the visualization of relative variances in a
time series analysis.</p>
<p>Horizontal pins consist of:</p>
<ul>
<li>
<p><strong>Horizontal category axis:</strong> see <em>single column chart</em>.</p>
</li>
<li>
<p><strong>Pins</strong>: One <em>pin</em> per time period or point of time extends from the category axis
to
the respective length. The pin has the size of a very thin column. Color the pin green or red
corresponding with positive or negative relative variances respectively. The fill of the pinhead
represents the primary scenario (see the paragraph on “Scenario comparisons” in UN 4.1 “<a
href="#un-41-unify-scenario-analyses">Unify scenario analyses</a>”). Display the pin in the
foreground, so that the length of the pin (see length X in the figure) is not hidden.</p>
</li>
<li>
<p><strong>Legends</strong>: As there is only one data series, the legend (name of the data series) is part
of the chart title.</p>
</li>
<li>
<p><strong>Data labels</strong>:<em>Data labels</em> name the values of the data series consistent with the
length of the respective pins. Position the labels of positive values above the respective pins, labels
of negative values below.</p>
</li>
</ul>
<p><strong>Horizontal waterfall charts</strong></p>
<p><em>Horizontal waterfall charts</em> (short:<em>horizontal waterfalls</em> or <em>column waterfalls</em>) analyze
root causes, over time, for the change or variance between two or more statuses. Accordingly, horizontal
waterfalls consist of two or more <em>base columns and totals columns</em>. In between a base column and a
totals
column there are <em>contribution columns</em> demonstrating what led to the difference between these two
columns. The <em>contribution columns</em> start at the end value, i.e. the height, of the preceding column, and
show the positive or negative contribution as well as the accumulated contribution of all columns up to the
respective point of time.</p>
<p>There are two types of horizontal waterfalls:</p>
<figure><img src="img/ex-1.1-6.png" alt="Figure EX 1.1-6: Growth waterfalls">
<figcaption>Figure EX 1.1-6: Growth waterfalls</figcaption>
</figure>
<p><strong>Growth waterfalls</strong>: In <em>growth waterfalls</em>, base columns and totals columns represent a
stock measure (e.g. headcount, accounts receivable) at different points in time (e.g. end of Q4 2012, 2013 and
2014). The contribution columns in between represent the changes (increases and decreases) over time of this
stock measure.</p>
<p>(There is no vertical equivalent to the horizontal <em>growth waterfall</em>.)</p>
<figure><img src="img/ex-1.1-7.png" alt="Figure EX 1.1-7: Horizontal variance waterfalls">
<figcaption>Figure EX 1.1-7: Horizontal variance waterfalls</figcaption>
</figure>
<p><strong>Horizontal variance waterfalls</strong>: In <em>horizontal variance waterfalls</em>, base columns and
totals columns represent a flow measure (e.g. sales) at different periods in time (e.g. 2015 and 2016) and/or
regarding different scenarios (e.g. PL and AC). The contribution columns in between represent the periodical
variances between the different time periods and/or scenarios.</p>
<p>The elements of a horizontal waterfall chart are the same as the elements of single column charts. In
addition, <em>assisting lines</em> connect the end of a column to the beginning of the succeeding column.</p>
<p><strong>Line charts</strong></p>
<figure><img src="img/ex-1.1-8.png" alt="Figure EX 1.1-8: Line chart">
<figcaption>Figure EX 1.1-8: Line chart</figcaption>
</figure>
<p>In general, <em>line charts</em> are used for the display of the temporal evolvement of data series with many
data
points.</p>
<figure><img src="img/ex-1.1-9.png" alt="Figure EX 1.1-9: Line chart with selective data labels">
<figcaption>Figure EX 1.1-9: Line chart with selective data labels</figcaption>
</figure>
<p>Many data points lead to small category widths. The advantage of line charts over column charts is the simplified
display of data (better <em>data-ink-ratio</em>). In addition, they can better represent positive and negative
values of more than one data series than columns. On the other hand, lines tend to imply a continuous timeline –
practically non-existent in business communication. Therefore lines should not be used for the presentation of
data series with only a few values.</p>
<p>Line charts cannot be “stacked” in order to show structure like in stacked column charts. In the place of line
charts for “stacked data”, <em>area charts</em> offer a good solution (there is no layout concept for area
charts
in this version of the guide yet).</p>
<p>Line charts with more than three intersecting lines tend to be confusing. Instead, several smaller charts with
one line each could be placed next to one another (small multiples), particularly when the general trends of the
lines are to be analyzed – not the direct comparison of two data series (e.g. in comparing seasonal developments
of several years), see also EX 2.4 “<a href="#ex-24-replace-spaghetti-charts">Replace spaghetti charts</a>”.</p>
<p>Line charts consist of:</p>
<ul>
<li>
<p><strong>Horizontal category axis:</strong> See <em>single column chart</em>. The semantic rules in part 3
suggest to use the category width (see width A in the first figure) for identifying the period type (see
UN 3.3 “<a href="#un-33-unify-time-periods-and-points-of-time">Unify time periods</a>”).</p>
</li>
<li>
<p><strong>Lines</strong>: One or more <em>lines</em> with <em>line markers</em> represent the values of the
respective data series. Use line thickness, line color, and line markers for meaning, see part “Semantic
rules”.</p>
</li>
<li>
<p><strong>Legends</strong>:<em>Legends</em> label the data series. If the line chart shows only one data
series, include the legend in the chart title. If the line chart shows two or more data series, the
legend should be positioned to the right of the far right data point (left-aligned text, see first
figure) or the left of the far left data point (right-aligned text, see second figure). Alternatively
position legends close to the lines at any other place in the chart.</p>
</li>
<li>
<p><strong>Data labels</strong>:<em>Data labels</em> name the values of the respective data points. If
possible, label maximum values (peaks) above the line markers and minimum values (valleys) below the
line markers. In many practical applications it is not necessary to clutter the line chart by labeling
every data point, see second figure on the left and SI 5.3 “<a
href="#si-53-avoid-unnecessary-labels">Avoid unnecessary labels</a>”.</p>
</li>
</ul>
<p><strong>Other horizontal charts</strong></p>
<p>Other chart types with horizontal category axes are <em>boxplot charts</em> (range charts) and <em>area
charts</em>. There is no specific notation concept for these chart types yet however it can be easily
derived from the notation concept of column and line charts.</p>
<p><strong>Charts with vertical category axes</strong></p>
<p>Charts with vertical category axes (<em>vertical charts</em>) typically show structural data. In general, present
structural data of one period or one point of time in the form of <em>bars</em>.</p>
<p>Use the vertical category as a structure axis. Horizontally, the visualization elements represent the data per
structure element (there is no need for a horizontal value axis as the visualization elements carry their own
values). Structure axes run from top to bottom and show characteristics of structures (e.g. products or
countries). The sequence of these elements depends on the intended analysis; see the UNIFY section about
“Structure analyses”.</p>
<p>In general, the data series of a vertical chart is represented by (horizontal)<em>bars</em> (single, stacked,
grouped), by <em>horizontal pins</em>, or by <em>waterfall bars</em>. Do not use lines in vertical charts as
they
could be interpreted as trends or developments, which do not exist in structure analyses.</p>
<p><em>Horizontal pins</em> can be considered very thin bars, but because of their importance are dealt with in a
separate section. A chart with horizontal pins is called a <em>vertical pin chart</em>.</p>
<p>Here follows the grouping of <em>vertical chart types</em>:</p>
<p><strong>Single bar charts</strong></p>
<figure><img src="img/ex-1.1-10.png" alt="Figure EX 1.1-10: Single bar charts">
<figcaption>Figure EX 1.1-10: Single bar charts</figcaption>
</figure>
<p>In general, <em>single bar charts</em> (short: single bars) are used for the structural analysis of one data
series (e.g., products, countries, or divisions) for one period or one point in time.</p>
<p>Single bar charts consist of:</p>
<ul>
<li>
<p><strong>Vertical category axis:</strong> The <em>vertical category axis</em> with its labels represents
the respective structure elements such as countries, products, etc. The category width (see width A in
figure) should be the same for corresponding analyses.</p>
</li>
<li>
<p><strong>Bars</strong>: One <em>bar</em> per structure element extends from the category axis to the
length
representing the respective value. Display the bars in the foreground of the category axis, so that the
length of the bar is not hidden. The part on “Semantic rules” suggests that the ratio of bar width to
category width (see ratio B/A in figure) represents information about the measure type (see UN 3.1 “<a
href="#un-31-unify-measures">Unify measures</a>”).</p>
</li>
<li>
<p><strong>Legends</strong>: As there is only one data series, the legend (name of the data series) is part
of the chart title.</p>
</li>
<li>
<p><strong>Data labels</strong>:<em>Data labels</em> name the values of the data series consistent with the
length of the respective bars. Position the labels of positive values at the right hand side of the
respective bars, the labels of negative values at the left hand side.</p>
</li>
</ul>
<p><strong>Stacked bar charts</strong></p>
<figure><img src="img/ex-1.1-11.png" alt="Figure EX 1.1-11: Stacked bar charts">
<figcaption>Figure EX 1.1-11: Stacked bar charts</figcaption>
</figure>
<p>Stacked bar charts (short: stacked bars) represent more than one data series (e.g., products, countries, or
divisions) for one period or one point in time.</p>
<p>Stacked bar charts consist of:</p>
<ul>
<li>
<p>Vertical category axis: See single bar charts.</p>
</li>
<li>
<p><strong>Bars</strong>: The bars (see single bar charts) are divided into segments (Excel names them“data
points”) representing the data series (stacked bars).</p>
</li>
<li>
<p><strong>Legends</strong>: Legends (names of the data series) are positioned either above the top stacked
bar or below the bottom stacked bar, with the bar segments defining their horizontal position: they are
horizontally centered with the data labels of the respective bar segment. If a segment at the top or
bottom is missing or has a very small size, its legend might need assisting lines.</p>
</li>
<li>
<p><strong>Data labels</strong>:<em>Data labels</em> name the values of the data series corresponding to the
length of the respective bar segment. If the sum of the bar segments of a category is positive (bar
pointing to the right), the label of the sum is positioned to the right hand side of the respective bar.
If the sum is negative (bar pointing to the left), the label of the sum is positioned to the left hand
side of the respective bar.</p>
</li>
</ul>
<p>It must be pointed out that stacked bars should only be used if all chart values are either positive or negative.
</p>
<p>This chart type might also not be a good choice if the values of each data series vary too much. The maximum
number of data series (segments of a stacked bars) to be shown depends on the range of how much the values of
each data series vary: More than 5 data series will only work well in the case of little variations.</p>
<p>Position the data series of central interest directly at the axis in order to best see its structure.</p>
<p><strong>Grouped bar charts</strong></p>
<figure><img src="img/ex-1.1-12.png" alt="Figure EX 1.1-12: Grouped bar charts">
<figcaption>Figure EX 1.1-12: Grouped bar charts</figcaption>
</figure>
<p>In general, <em>grouped bar charts</em> (short: grouped bars) show structure analyses for a primary scenario
(e.g.
AC or FC) in comparison to a reference scenario (e.g. PY or PL). Two bars per category (<em>grouped bars)</em>
represent <em>these</em> two scenarios.</p>
<p>The bars of the primary scenario and the reference scenario overlap, the reference scenario placed behind the
primary scenario – either above or below (see the bottom chart of the figure as well as the paragraph on
“Scenario comparisons” in UN 4.1 “<a href="#un-41-unify-scenario-analyses">Unify scenario analyses</a>”).</p>
<p>A third scenario could be displayed using a <em>reference scenario triangle</em>. All other elements of a grouped
bar chart are identical to a single bar chart.</p>
<p>Alternatively, instead of grouped bars, the primary scenario can be represented with a single bar and the
reference scenario by reference scenario triangles (see top chart of figure).</p>
<p><strong>Vertical pin charts</strong></p>
<figure><img src="img/ex-1.1-13.png" alt="Figure EX 1.1-13: Vertical pin charts">
<figcaption>Figure EX 1.1-13: Vertical pin charts</figcaption>
</figure>
<p><em>Vertical pin charts</em> (short: vertical pins) are used for the visualization of relative variances in a
structure analysis.</p>
<p>Vertical pins consist of:</p>
<ul>
<li>
<p>Vertical category axis: see <em>single bar chart</em>.</p>
</li>
<li>
<p><strong>Pins</strong>: One <em>pin</em> per structure element extends from the category axis to the
respective length. The pin has the size of a very thin bar. It is colored green or red when representing
positive or negative relative variances respectively. The fill of the pinhead represents the primary
scenario (see the paragraph on “Scenario comparisons” in UN 4.1 “<a
href="#un-41-unify-scenario-analyses">Unify scenario analyses</a>”). Display pins in the foreground,
so that the length of the pin (see length X in the figure) is not hidden.</p>
</li>
<li>
<p><strong>Legends</strong>: As there is only one data series, the legend (name of the data series) is part
of the chart title.</p>
</li>
</ul>
<p><strong>Data labels</strong>:<em>Data labels</em> name the values of the data series corresponding to the length
of the respective pins. Position the labels of positive values at the right hand side of the respective pins,
the labels of negative values at the left hand side.</p>
<p><strong>Vertical waterfall charts</strong></p>
<p><em>Vertical waterfalls charts</em> (in short:<em>vertical waterfalls</em> or <em>bar waterfalls</em>) analyze
structural root causes for the difference between two or more statuses. Accordingly, vertical waterfalls consist
of two or more <em>base bars</em> and <em>totals bars</em>. In between a base bar and a totals bar there
are <em>contribution bars</em> representing the contribution to the difference between these two bars. Starting
from the top base bar, <em>contribution bars</em> always start at the end of the preceding bar, showing positive
or negative individual contributions of the respective structure element as well as the accumulated contribution
resulting in the next totals bar.</p>
<p>There are two types of vertical waterfalls:</p>
<figure><img src="img/ex-1.1-14.png" alt="Figure EX 1.1-14: Calculation waterfalls">
<figcaption>Figure EX 1.1-14: Calculation waterfalls</figcaption>
</figure>
<p><strong>Calculation waterfalls</strong>: In <em>calculation waterfalls</em>, base bars and totals bars represent
base and result measures (e.g. sales and EBIT) whereas the contribution bars in between represent the additions
and subtractions of other measures (e.g. financial income and direct cost) in a calculation scheme. More complex
calculation schemes (e.g. profit and loss calculation) can have intermediate subtotals bars.</p>
<p>(There is no horizontal correspondence to the vertical <em>calculation waterfall</em>.)</p>
<figure><img src="img/ex-1.1-15.png" alt="Figure EX 1.1-15: Vertical variance waterfalls">
<figcaption>Figure EX 1.1-15: Vertical variance waterfalls</figcaption>
</figure>
<p><strong>Vertical variance waterfalls</strong>: In <em>vertical variance waterfalls</em>, base bars and totals
bars
represent values at different periods or points in time (e.g. January 1, 2013 and January 1, 2014) and/or
different scenarios (e.g. PY and AC). The contribution bars in between represent the variances in structure
between the different times and/or scenarios.</p>
<p>The elements of vertical waterfalls are the same as the elements of single bar charts. In addition, <em>assisting
lines</em> connect the end of a bar with the beginning of the succeeding bar.</p>
<p><strong>Remainder bar</strong></p>
<figure><img src="img/ex-1.1-16.png" alt="Figure EX 1.1-16: Remainder bar">
<figcaption>Figure EX 1.1-16: Remainder bar</figcaption>
</figure>
<p>If a large number of elements need to be presented, then only the most important elements can be displayed in one
chart or on one page. In order to make the analyses exhaustive, sort the elements by descending size,
accumulating the smallest elements, which cannot be depicted, in a <em>remainder bar</em> (“rest of...”).
Separate the remainder bar from the other bars by a wider gap (see gap C in the figure on the left).</p>
<p>Note: This remainder bar has to be excluded from some Structure analyses such as averaging, ranking, and
selecting.</p>
<p><strong>Other vertical charts</strong></p>
<p>Other chart types with vertical category axes are <em>vertical boxplot charts</em> (range charts). There is no
specific notation concept for this chart type yet however it can be derived from the notation of the standard
bar charts.</p>
<p>In general, do not use lines and areas in vertical charts as they might underline a continuum of data
non-existent in business communication.</p>
<p><strong>Charts with two values axes</strong></p>
<figure><img src="img/ex-1.1-17.png" alt="Figure EX 1.1-17: Charts with two values axes">
<figcaption>Figure EX 1.1-17: Charts with two values axes</figcaption>
</figure>
<p><em>Charts with two value axes</em> show two-dimensional positioning of visualization elements, which can provide
new and interesting insights. <em>Scattergrams</em> arrange points in a two-dimensional coordinate system.</p>
<figure><img src="img/ex-1.1-18.png" alt="Figure EX 1.1-18: Bubble charts">
<figcaption>Figure EX 1.1-18: Bubble charts</figcaption>
</figure>
<p><em>Bubble charts</em> (portfolio charts) have bubbles instead of points and use the bubble area to show a third
dimension. A fourth dimension could be presented via pie segments within the bubbles (bubble pie charts).</p>
<p>Besides <em>scattergrams</em> and <em>bubble charts</em> there are other chart types with two value axes, e.g.
charts with horizontal axes representing a continuous timeline (instead of fixed time categories) and charts
with columns or bars of variable width.</p>
<p>There are no specific notation rules for charts with two value axes yet. An appropriate notation concept for
these chart types can be derived from the notation of column charts, bar charts and line charts with their data
visualization elements, legends, data labels, and axes.</p>
<h3 id="ex-12-use-appropriate-table-types">EX 1.2 Use appropriate table types</h3>
<p>A <em>table</em> is a communication object in which data is arranged in two dimensions, i.e.
(vertical)<em>columns
and</em> (horizontal)<em>rows</em>. The <em>row header</em> (row name) describes the content of a row,
the <em>column header</em> (column name) the content of a column. The data are positioned at the intersections
of
rows and columns called <em>table cells</em>.</p>
<p>“One-dimensional tables” (tables with one or more columns but without row headers) are called <em>lists</em> and
are not covered here.</p>
<p><em>Table types</em> are defined by a set of <em>columns</em> and a set of <em>rows</em> in order to fulfill
specific analytic and/or reporting purposes.</p>
<p><strong>Column types</strong></p>
<p>Column types are columns with similar content falling under similar column headers. Typical column types
are <em>time columns</em> (with monthly or yearly data), <em>scenario columns</em> (with actual or plan data)
and <em>variance columns</em> (ΔPL or ΔPY).</p>
<p>The following layout principles apply to all column types:</p>
<ul>
<li>
<p><strong>Width</strong>: Columns belonging to a certain column type should have an identical width. This
column width should not depend on the text length of the respective column header.</p>
</li>
<li>
<p><strong>Orientation</strong>: Right-align columns with numerical data. Left-align columns with
non-numerical data (e.g. texts or product names). <em>Column headers</em> have the same orientation as
the rest of the column. Headers for combined columns can be centered or even left-aligned to increase
legibility.</p>
</li>
<li>
<p><strong>Vertical lines and gaps</strong>: Vertical lines separating different columns should be very
light or even omitted. Use white vertical lines or white vertical gaps to mark the columns. In the
following figures, different widths of these white lines resp. gaps are being used to separate and group
columns. Separate columns of the same type by a narrow gap (see gap B1 in the figure in section
“Scenario columns” et seq.). Use a wider gap to separate a group of similar columns from the next group
(see gap B2 in the figure in section “Row header columns” et seq.).</p>
</li>
</ul>
<p>Additional layout principals depend on the <em>column types</em> described below.</p>
<p><strong>Row header columns</strong></p>
<figure><img src="img/ex-1.2-1.png" alt="Figure EX 1.2-1: Row header columns">
<figcaption>Figure EX 1.2-1: Row header columns</figcaption>
</figure>
<p>Row header columns contain the header texts of the rows. Often, these columns are positioned at the very left of
a table. In most cases, row header columns are much wider than other column types.</p>
<p>Keep the texts of the row headers short by using abbreviations or footnotes in order to omit too wide tables.</p>
<p>Use a wider gap (see width B2 in the figure) to separate the <em>row header column</em> from columns with
numbers.
</p>
<p><strong>Scenario columns</strong></p>
<figure><img src="img/ex-1.2-2.png" alt="Figure EX 1.2-2: Scenario columns">
<figcaption>Figure EX 1.2-2: Scenario columns</figcaption>
</figure>
<p><em>Scenario columns</em> show data for scenarios (e.g. previous year, plan, actual). Use the same width for all
scenario columns (depending on the number of digits).</p>
<p>For the sequence of scenario columns see UN 4.1 “<a href="#un-41-unify-scenario-analyses">Unify scenario
ana­lyses</a>”.</p>
<p><strong>Variance columns</strong></p>
<figure><img src="img/ex-1.2-3.png" alt="Figure EX 1.2-3: Variance columns">
<figcaption>Figure EX 1.2-3: Variance columns</figcaption>
</figure>
<p><em>Variance columns</em> show data of absolute variances (e.g. ΔPL, ΔPY) or relative variances (e.g. ΔPL%,
ΔPY%).</p>
<p><strong>Time columns</strong></p>
<figure><img src="img/ex-1.2-4.png" alt="Figure EX 1.2-4: Time columns">
<figcaption>Figure EX 1.2-4: Time columns</figcaption>
</figure>
<p><em>Time columns</em> show <em>time periods</em> (for flow measures) or <em>points of time</em> (for stock
measures).</p>
<p>Use a temporal order – from left to right – for the sequence of the time columns (e.g. Jan, Feb, Mar, or 2013,
2014, 2015).</p>
<p><strong>Measure columns</strong></p>
<figure><img src="img/ex-1.2-5.png" alt="Figure EX 1.2-5: Measure columns">
<figcaption>Figure EX 1.2-5: Measure columns</figcaption>
</figure>
<p><em>Measure columns</em> show measures such as sales, headcount, or equity.</p>
<p>Displaying long measure names in column headers can be challenging. As the column width should not depend on the
length of the measure name, use the abbreviations defined in the glossary instead.</p>
<p>Use a wider gap after intermediate results to expose the calculation scheme (see width B2 in the figure on the
left).</p>
<p><strong>Structure columns</strong></p>
<figure><img src="img/ex-1.2-6.png" alt="Figure EX 1.2-6: Structure columns">
<figcaption>Figure EX 1.2-6: Structure columns</figcaption>
</figure>
<p><em>Structure columns</em> show the elements of a structure dimension (e.g. countries or products).</p>
<p><strong>“Thereof” columns</strong></p>
<figure><img src="img/ex-1.2-7.png" alt="Figure EX 1.2-7: “Thereof” columns">
<figcaption>Figure EX 1.2-7: “Thereof” columns</figcaption>
</figure>
<p>If details of an aggregated column are shown in one or more column not totaling to the aggregated column, these
columns are called “thereof” columns.</p>
<p>The design of the <em>thereof columns</em> should differ from other columns. E.g. use a smaller font (see X in
the
figure) to expose a <em>thereof column</em> and do not separate it from the mother column (see
columns <em>AL3</em> and <em>AL3.1</em> in the figure) in order to show that it is part of it. A <em>thereof
column</em> is positioned at the right hand side of the mother column.</p>
<p><strong>Remainder columns</strong></p>
<figure><img src="img/ex-1.2-8.png" alt="Figure EX 1.2-8: Remainder columns">
<figcaption>Figure EX 1.2-8: Remainder columns</figcaption>
</figure>
<p>If the set to be presented in the columns has too many elements, accumulate the less important or smaller
elements in a <em>remainder column</em> (e.g. 10 columns for the top 10 countries and a remainder column
titled“Rest of world” or “RoW”).</p>
<p>In the figure, the <em>remainder column</em> “Other cost” has the same vertical gaps B1 as the other measure
columns.</p>
<p><strong>“Percent of” columns</strong></p>
<figure><img src="img/ex-1.2-9.png" alt="Figure EX 1.2-9: “Percent of” columns">
<figcaption>Figure EX 1.2-9: “Percent of” columns</figcaption>
</figure>
<p>Use “<em>Percent of</em>” columns to present important data of another column as shares of a given total. A
typical example for a “percent of” column is data of a profit and loss statement as a percentage of sales.</p>
<p>“Percent of” columns have a smaller font size (see X) than the other columns.</p>
<p><strong>Totals columns</strong></p>
<figure><img src="img/ex-1.2-10.png" alt="Figure EX 1.2-10: Totals columns">
<figcaption>Figure EX 1.2-10: Totals columns</figcaption>
</figure>
<p>Position columns displaying <em>totals of a group of columns</em> (e.g. quarters totaling in years or products
totaling in product groups) at the right hand side of the columns belonging to this group. The design of
the <em>totals columns</em> should differ from other columns, e.g. highlighted by bold fonts or by light gray
background.</p>
<p>The column types described before refer to <em>single</em> columns. The following paragraphs
present <em>combined</em> columns i.e. <em>hierarchical</em> and <em>nested</em> columns.</p>
<p><strong>Hierarchical columns</strong></p>
<figure><img src="img/ex-1.2-11.png" alt="Figure EX 1.2-11: Hierarchical columns">
<figcaption>Figure EX 1.2-11: Hierarchical columns</figcaption>
</figure>
<p>Hierarchies in dimensions may call for columns showing multiple levels. If possible, the sibling elements
belonging to the same parent element of a dimension should be homogenous, mutually exclusive, and collectively
exhaustive.</p>
<p>Separate parents by appropriate means, e.g. wider gaps. Display the parent columns at the right hand side of
their child columns (like <em>totals columns</em>).</p>
<p>In the figure, a wider gap B2 separates the two years (with four quarters each) from each other.</p>
<p><strong>Nested columns</strong></p>
<figure><img src="img/ex-1.2-12.png" alt="Figure EX 1.2-12: Nested columns">
<figcaption>Figure EX 1.2-12: Nested columns</figcaption>
</figure>
<p>In <em>nested columns</em>, two column types are combined in such a way that the columns of one type repeat
iteratively within every column of the other type. Separate the resulting groups of columns by appropriate
means, e.g. wider gaps.</p>
<p>In the figure, wider gaps B2 separate the four years (with AC and PL data each) from each other.</p>
<p><strong>Row types</strong></p>
<p><em>Row types</em> are rows with similar content falling under similar row headers. Typical row types
are <em>measure rows</em> (e.g. sales, cost, profit) or <em>structure rows</em> (e.g. Italy, France, UK).</p>
<p>The following layout principles apply to all row types:</p>
<ul>
<li>
<p><strong>Height</strong>: Rows belonging to a row type should have an identical height (see height A in
the figure in section “measure rows” et seq.).</p>
</li>
<li>
<p><strong>Horizontal lines</strong>: Separating rows by light horizontal lines will increase the
legibility.</p>
</li>
</ul>
<p>Additional layout principals depend on the row types described below.</p>
<p><em>Time periods and points of time</em>, <em>scenarios</em>, and <em>variances</em> should be displayed in
columns
rather than in rows.</p>
<p><strong>Column header rows</strong></p>
<figure><img src="img/ex-1.2-13.png" alt="Figure EX 1.2-13: Column header rows">
<figcaption>Figure EX 1.2-13: Column header rows</figcaption>
</figure>
<p>Column header rows contain the header texts of the columns. In most cases, these rows are positioned at the very
top of a table. In order to group columns two and more column header rows might be necessary. If necessary,
abbreviate column header texts in order to fit in the preferred column width. Alternatively keep column headers
short by using footnotes.</p>
<p>In the figure the <em>column header row</em> uses two lines in order to fit the column header texts in the narrow
columns.</p>
<p><strong>Measure rows</strong></p>
<figure><img src="img/ex-1.2-14.png" alt="Figure EX 1.2-14: Measure rows">
<figcaption>Figure EX 1.2-14: Measure rows</figcaption>
</figure>
<p><em>Measure rows</em> show measures such as sales, headcount, or equity.</p>
<p>Separate rows showing final or intermediate results of a calculation scheme (<em>results rows</em><em>or totals
rows</em>) by solid lines. Display results rows in bold font or highlight them with light gray background.
An additional gap B below a results row will increase legibility.</p>
<p><strong>Structure rows</strong></p>
<figure><img src="img/ex-1.2-15.png" alt="Figure EX 1.2-15: Structure rows">
<figcaption>Figure EX 1.2-15: Structure rows</figcaption>
</figure>
<p>Structure rows show elements of a structure dimension (e.g. countries or products).</p>
<p><strong>“Thereof” rows</strong></p>
<figure><img src="img/ex-1.2-16.png" alt="Figure EX 1.2-16: “Thereof” rows">
<figcaption>Figure EX 1.2-16: “Thereof” rows</figcaption>
</figure>
<p>If details of an aggregated row are shown in one or more rows not totaling to the aggregated row, these rows are
called “thereof” rows. Place the aggregated <em>above</em> the “thereof” rows (in contrast to the <em>totals
row</em> positioned <em>below</em> the rows of its group).</p>
<p>The design of the <em>thereof rows</em> should differ from other rows. E.g. in the figure, the <em>thereof
row</em>
is of smaller height, written in a smaller font (see X), not separated by a horizontal line, and has a
right-aligned row header.</p>
<p><strong>Remainder rows</strong></p>
<figure><img src="img/ex-1.2-17.png" alt="Figure EX 1.2-17: Remainder rows">
<figcaption>Figure EX 1.2-17: Remainder rows</figcaption>
</figure>
<p>If the structure dimension to be presented in the rows outline has too many elements, accumulate the less
important or smaller elements in a <em>remainder row</em> (e.g. 7 rows for the top 7 countries and a remainder
titled “Rest of world”).</p>
<p>Exclude remainder rows from some of the Structure analyses such as averaging, ranking, and selecting.</p>
<p>In the figure, the <em>remainder row</em> has the same row height A as the other structure rows of this table
example.</p>
<p><strong>“Percent of” rows</strong></p>
<figure><img src="img/ex-1.2-18.png" alt="Figure EX 1.2-18: “Percent of” rows">
<figcaption>Figure EX 1.2-18: “Percent of” rows</figcaption>
</figure>
<p>Use “<em>Percent of</em>” rows to present important data of another row as shares of a given total. A typical
example for a “<em>percent of</em>” row is gross profit as a percentage of sales.</p>
<p>“Percent of” rows have a smaller font size (see X) than the other rows.</p>
<p><strong>Totals rows</strong></p>
<figure><img src="img/ex-1.2-19.png" alt="Figure EX 1.2-19: Totals rows">
<figcaption>Figure EX 1.2-19: Totals rows</figcaption>
</figure>
<p>Place rows displaying <em>totals of a group of rows</em> (e.g. countries totaling in regions or products totaling
in product groups) below the rows of this group and separated them by solid lines.</p>
<p>The design of the <em>totals rows</em> should differ from other rows, e.g. highlighted by bold fonts or by light
gray background.</p>
<p>The row types described before refer to <em>single</em> rows. The following paragraphs present <em>combined</em>
rows i.e. <em>hierarchical</em> and <em>nested</em> rows.</p>
<p><strong>Hierarchical rows</strong></p>
<figure><img src="img/ex-1.2-20.png" alt="Figure EX 1.2-20: Hierarchical rows">
<figcaption>Figure EX 1.2-20: Hierarchical rows</figcaption>
</figure>
<p>Hierarchies in dimensions may call for rows showing multiple levels. If possible, the sibling elements belonging
to the same parent element of a dimension should be homogenous, mutually exclusive, and collectively exhaustive.
</p>
<p>Separate parents by appropriate means, e.g. wider gaps (see additional gap B in the figure). Display the parent
rows <em>below</em> their child rows (like <em>totals rows</em>).</p>
<p><strong>Nested rows</strong></p>
<figure><img src="img/ex-1.2-21.png" alt="Figure EX 1.2-21: Nested rows">
<figcaption>Figure EX 1.2-21: Nested rows</figcaption>
</figure>
<p>In <em>nested rows</em>, two types of rows are combined in such a way that the rows of one type repeat
iteratively
within every row of the other row type.</p>
<p>Separate the resulting groups of rows by appropriate means, e.g. wider gaps (see additional gap B in the figure).
</p>
<p><strong>Table types</strong></p>
<figure><img src="img/ex-1.2.png" alt="Figure EX 1.2: Table types">
<figcaption>Figure EX 1.2: Table types</figcaption>
</figure>
<p>Table types are distinguished by their analytic purpose in time series tables, variance tables and cross tables.
Tables serving more than one analytic purpose are called combined tables.</p>
<p><strong>Time series tables</strong></p>
<figure><img src="img/ex-1.2-22.png" alt="Figure EX 1.2-22: Time series tables">
<figcaption>Figure EX 1.2-22: Time series tables</figcaption>
</figure>
<p><em>Time series tables</em> are used for time series analyses, combining time columns with measure rows or
structure rows.</p>
<p>A typical example for a <em>time series table</em> is a sales analysis by countries (rows) and years (columns).
</p>
<p><strong>Variance tables</strong></p>
<figure><img src="img/ex-1.2-23.png" alt="Figure EX 1.2-23: Variance tables">
<figcaption>Figure EX 1.2-23: Variance tables</figcaption>
</figure>
<p><em>Variance tables</em> are used for scenario analyses, combining scenario columns and variance columns with
measure rows or structure rows.</p>
<p>A typical example for a <em>variance table</em> is a sales analysis by countries (rows) showing different
scenarios and different variances (columns).</p>
<p><strong>Cross tables</strong></p>
<figure><img src="img/ex-1.2-24.png" alt="Figure EX 1.2-24: Cross tables">
<figcaption>Figure EX 1.2-24: Cross tables</figcaption>
</figure>
<p><em>Cross tables</em> are used for Structure analyses, combining structure columns with structure rows.</p>
<p>A typical example of a <em>cross table</em> is a sales table with countries in rows and products in columns.</p>
<p><strong>Combined tables</strong></p>
<figure><img src="img/ex-1.2-25.png" alt="Figure EX 1.2-25: Combined table 1">
<figcaption>Figure EX 1.2-25: Combined table 1</figcaption>
</figure>
<p><em>Combined tables</em> are used for multiple analyses. A combined table uses more than one <em>column type</em>
and/or more than one <em>row type</em> presented either side by side or nested.</p>
<p>The first figure shows a hierarchical structure of countries on three levels in the rows. The columns are nested:
scenarios and variances are the same for both time periods <em>November</em> and <em>January_November</em>.</p>
<figure><img src="img/ex-1.2-26.png" alt="Figure EX 1.2-26: Combined table 2">
<figcaption>Figure EX 1.2-26: Combined table 2</figcaption>
</figure>
<p>The second figure shows the measures of a calculation scheme in the rows. The columns are nested: The four
quarters and the annual total are the same for both years.</p>
<figure><img src="img/ex-1.2-27.png" alt="Figure EX 1.2-27: Combined table 3">
<figcaption>Figure EX 1.2-27: Combined table 3</figcaption>
</figure>
<p>The third figure shows the same rows as the second one (measures of a calculation scheme). The nested columns now
show PY and AC data as well as absolute and relative variances for two markets.</p>
<h3 id="ex-2-replace-inappropriate-chart-types">EX 2 Replace inappropriate chart types</h3>
<p>Inappropriate charts make it hard to perceive the message. Knowing the correct usage of chart types helps in
replacing inappropriate visualizations, such as pie charts, speedometer visualizations, radar charts, and
spaghetti charts, with those chart types better suited.</p>
<h3 id="ex-21-replace-pie-and-ring-charts">EX 2.1 Replace pie and ring charts</h3>
<figure><img src="img/ex-2.1.png" alt="Figure EX 2.1: Replace pie and ring charts">
<figcaption>Figure EX 2.1: Replace pie and ring charts</figcaption>
</figure>
<p><em>Pie</em> and <em>ring charts</em> are<a href="http://en.wikipedia.org/wiki/Circle">circular charts</a>
dividing some total into<a href="http://en.wikipedia.org/wiki/Circular_sector">sectors</a> of relative
proportion, but there are better ways to illustrate the numericalproportions of segments, e.g. bar charts or
charts with stacked columns, see Figure EX 2.1.</p>
<p><em>Pie charts</em> allow for one-dimensional analyses only, and therefore seldom convey revealing insights.
However, some useful applications for pie charts exist, for example when market sizes and/or market shares for
one period need to be allocated to certain regions on a map (see CH 3.3 “<a
href="#ch-33-avoid-misleading-colored-areas-in-maps">Avoid misleading colored areas in maps</a>”). As
opposed to column or bar charts, pie charts can be positioned on a specific point on a map.</p>
<h3 id="ex-22-replace-gauges-speedometers">EX 2.2 Replace gauges, speedometers</h3>
<figure><img src="img/ex-2.2.png" alt="Figure EX 2.2: Replace gauges, speedometers">
<figcaption>Figure EX 2.2: Replace gauges, speedometers</figcaption>
</figure>
<p>Often found as part of a so-called dashboard, <em>speedometers</em> are probably one of the most useless
visualizations out there. They take up way too much space and have often confusing color coding and scaling. In
general, bar charts showing the respective structures or columns charts showing the respective development over
time are better choices, see Figure EX 2.2.</p>
<h3 id="ex-23-replace-radar-and-funnel-charts">EX 2.3 Replace radar and funnel charts</h3>
<figure><img src="img/ex-2.3.png" alt="Figure EX 2.3: Replace radar and funnel charts">
<figcaption>Figure EX 2.3: Replace radar and funnel charts</figcaption>
</figure>
<p>So-called <em>radar charts</em> (also called <em>net charts</em> or <em>spider charts</em>) are frequently used
for
evaluating purposes. Having no advantage over bar charts and having, actually, many weaknesses, use them only
for two-dimensional analyses (e.g. comparing young-old with rich-poor). Willard C. Brinton wrote almost 100
years ago: “This type of chart should be banished to the scrap heap. Charts on rectangular ruling are easier to
draw and easier to understand.”</p>
<p>Of course, if the circular arrangement has meaning (such as the compass direction), this kind of chart can be
very valuable, but these types of analysis are not typical in business reporting.</p>
<p><em>Funnel charts</em> are misleading when the size of the area displayed does not correspond to the respective
numerical values – an issue applying also to other artificial chart forms (e.g. spheres) in which the length,
area, or volume do not correspond to the numerical values.</p>
<h3 id="ex-24-replace-spaghetti-charts">EX 2.4 Replace spaghetti charts</h3>
<figure><img src="img/ex-2.4.png" alt="Figure EX 2.4: Replace spaghetti charts">
<figcaption>Figure EX 2.4: Replace spaghetti charts</figcaption>
</figure>
<p>A chart with more than three or four intersecting lines (“spaghetti chart”) can be more confusing than several
smaller charts with one line each placed next to one another (small multiples), particularly when evaluating the
shape or the trend of the lines, see Figure EX 2.4.</p>
<p>However, when needing to compare exactly the height of data points of several lines, spaghetti charts cannot be
avoided.</p>
<h3 id="ex-25-replace-traffic-lights">EX 2.5 Replace traffic lights</h3>
<figure><img src="img/ex-2.5.png" alt="Figure EX 2.5: Replace traffic lights">
<figcaption>Figure EX 2.5: Replace traffic lights</figcaption>
</figure>
<p>“Traffic lights” with green, red, and yellow areas are a popular form of visualization but contain little
information per area used. However, they can be used for analyses showing “yes or no” decisions or situations
similar to real traffic lights. In all other cases replace them with more suitable means of (analog)
representation such as bar charts, see Figure EX 2.5.</p>
<h3 id="ex-3-replace-inappropriate-representations">EX 3 Replace inappropriate representations</h3>
<p>From a perceptual perspective, avoid all visual representations requiring time consuming analyses or additional
explanations, particularly the popular use of merely conceptual representations and all forms of texts,
including bullet lists.</p>
<h3 id="ex-31-prefer-quantitative-representations">EX 3.1 Prefer quantitative representations</h3>
<figure><img src="img/ex-3.1.png" alt="Figure EX 3.1: Prefer quantitative representations">
<figcaption>Figure EX 3.1: Prefer quantitative representations</figcaption>
</figure>
<p>Due to the time constraints usually involved with presentations, conceptual graphs prove less suitable than
charts, photos, maps, etc. For a one-hour presentation, do not use more than three or four conceptual
representations. Do this only if they are absolutely essential for comprehension. The audience will understand
charts and pictures (photos, drawings, etc.) better and faster, see Figure EX 3.1.</p>
<h3 id="ex-32-avoid-text-slides-in-presentations">EX 3.2 Avoid text slides in presentations</h3>
<figure><img src="img/ex-3.2.png" alt="Figure EX 3.2: Avoid text slides in presentations">
<figcaption>Figure EX 3.2: Avoid text slides in presentations</figcaption>
</figure>
<p>Avoid all forms of text slides in presentations. Texts should either be recited or written in a handout. A few
exceptions to this rule are specific texts being discussed such as definitions, quotes, etc. In general, all
forms of lists (bullet points) should appear only in the written handout, not projected on the screen. True, if
someone sees and hears something simultaneously, he remembers it better than when he just hears it, but bear in
mind texts are not considered something merely to be seen – they must be read and understood, see Figure EX 3.2.
</p>
<h3 id="ex-4-add-comparisons">EX 4 Add comparisons</h3>
<p>Visual perception is strongly based on setting one perceived object in relation to another. Adding meaningful
comparisons helps the eye evaluate faster, the main purpose of charts.</p>
<h3 id="ex-41-add-scenarios">EX 4.1 Add scenarios</h3>
<figure><img src="img/ex-4.1.png" alt="Figure EX 4.1: Add scenarios">
<figcaption>Figure EX 4.1: Add scenarios</figcaption>
</figure>
<p>Scenarios such as “plan” and “previous year” are the most common references for comparison purposes. Add them
whenever available. Use a standardized scenario notation for faster comprehension, see Figure EX 4.1.</p>
<h3 id="ex-42-add-variances">EX 4.2 Add variances</h3>
<figure><img src="img/ex-4.2.png" alt="Figure EX 4.2: Add variances">
<figcaption>Figure EX 4.2: Add variances</figcaption>
</figure>
<p>Having added scenarios for comparison purposes, the visualization of variances makes it easier to evaluate the
situation. Use a standardized notation of variances for faster comprehension, see Figure EX 4.2.</p>
<h3 id="ex-5-explain-causes">EX 5 Explain causes</h3>
<p>Present data more understandable by showing interrelations, i.e. causes and dependencies. Seeing the entire
context, especially extreme values and deviant values, helps to explain causes. Details increase not only the
level of credibility but also comprehension. Use charts to prove, explain, and render something plausible, not
to serve merely as decoration. This section focuses on the explanation of causes by using tree structures,
clusters, and correlations. A more structured approach to increasing information density is discussed in the
chapter “CONDENSE – Increase information density”.</p>
<h3 id="ex-51-show-tree-structures">EX 5.1 Show tree structures</h3>
<figure><img src="img/ex-5.1.png" alt="Figure EX 5.1: Show tree structures">
<figcaption>Figure EX 5.1: Show tree structures</figcaption>
</figure>
<p>The presentation of the assumptions or basic data upon which a business analysis is based, results not only in
better understanding, but also makes it more convincing. A good choice is the display of calculated measures in
a tree structure, see Figure EX 5.1 (see also CO 5.2 “<a href="#co-52-show-related-charts-on-one-page">Show
related charts on one page</a>”).</p>
<h3 id="ex-52-show-clusters">EX 5.2 Show clusters</h3>
<figure><img src="img/ex-5.2.png" alt="Figure EX 5.2: Show clusters">
<figcaption>Figure EX 5.2: Show clusters</figcaption>
</figure>
<p>With the help of clusters in two-dimensional and three-dimensional forms, large amounts of data very often can
provide interesting and new insights, see Figure EX 5.2.</p>
<h3 id="ex-53-show-correlations">EX 5.3 Show correlations</h3>
<figure><img src="img/ex-5.3.png" alt="Figure EX 5.3: Show correlations">
<figcaption>Figure EX 5.3: Show correlations</figcaption>
</figure>
<p>When comparing several data series, correlations are often sought in order to better understand the
interrelations. Suitable rankings and comparisons can facilitate the understanding of patterns, see Figure EX
5.3.</p>
<h2 id="simplify--avoid-clutter">SIMPLIFY – Avoid Clutter</h2>
<p>SIMPLIFY covers all aspects of avoiding clutter in reports and presentations.</p>
<p><em>Avoiding clutter</em> means that reports and presentations avoid all components and characteristics, which
are too complicated, redundant, distracting or merely decorative.</p>
<p>This chapter covers avoiding unnecessary and decorative components and replacing them with cleaner layouts,
avoiding redundancies and distracting details.</p>
<ol>
<li><a href="#si-1-avoid-unnecessary-components">Avoid unnecessary components</a></li>
<li><a href="#si-2-avoid-decorative-styles">Avoid decorative styles</a></li>
<li><a href="#si-3-replace-with-cleaner-layout">Replace with cleaner layout</a></li>
<li><a href="#si-4-avoid-redundancies">Avoid redundancies</a></li>
<li><a href="#si-5-avoid-distracting-details">Avoid distracting details</a></li>
</ol>
<h3 id="si-1-avoid-unnecessary-components">SI 1 Avoid unnecessary components</h3>
<p>Completely avoid components, such as pictures, backgrounds, and logos, not contributing to the comprehension of a
report or presentation.</p>
<h3 id="si-11-avoid-cluttered-layouts">SI 1.1 Avoid cluttered layouts</h3>
<figure><img src="img/si-1.1.png" alt="Figure SI 1.1: Avoid cluttered layouts">
<figcaption>Figure SI 1.1: Avoid cluttered layouts</figcaption>
</figure>
<p>Layout concepts often contain elements that lack meaning but merely conform to corporate design or personal
taste. Avoid all these elements, see Figure SI 1.1.</p>
<h3 id="si-12-avoid-colored-or-filled-backgrounds">SI 1.2 Avoid colored or filled backgrounds</h3>
<figure><img src="img/si-1.2.png" alt="Figure SI 1.2: Avoid colored or filled backgrounds">
<figcaption>Figure SI 1.2: Avoid colored or filled backgrounds</figcaption>
</figure>
<p>Numbers and labels are easiest to read when depicted in black on a white background. Any type of background color
or pattern makes something harder to read, see Figure SI 1.2.</p>
<h3 id="si-13-avoid-animation-and-transition-effects">SI 1.3 Avoid animation and transition effects</h3>
<figure><img src="img/si-1.3.png" alt="Figure SI 1.3: Avoid animation and transition effects">
<figcaption>Figure SI 1.3: Avoid animation and transition effects</figcaption>
</figure>
<p>Animated <em>PowerPoint</em> slides are not useful if the animation has no meaning and does not support the
message, see Figure SI 1.3. They merely distract and confuse. Only the “appear” function is recommended to be
used for the gradual development of a slide.</p>
<h3 id="si-2-avoid-decorative-styles">SI 2 Avoid decorative styles</h3>
<p>Simplify complicated visualizations in order to facilitate and accelerate their comprehension. Whereas the
section “Avoid unnecessary components” involves omitting entire layout elements, the aim here is to find the
most suitable and simplest possible style of visualization elements.</p>
<h3 id="si-21-avoid-borders-shades-and-pseudo-3d">SI 2.1 Avoid borders, shades, and pseudo-3D</h3>
<figure><img src="img/si-2.1.png" alt="Figure SI 2.1: Avoid borders, shades, and pseudo-3D">
<figcaption>Figure SI 2.1: Avoid borders, shades, and pseudo-3D</figcaption>
</figure>
<p>In general, borders, shades, and pseudo-3D convey no meaning and make comprehension more difficult. Shades and
pseudo-3D might even give a false visual impression. Avoid them because they do not add value, see Figure SI
2.1.</p>
<h3 id="si-22-avoid-decorative-colors">SI 2.2 Avoid decorative colors</h3>
<figure><img src="img/si-2.2.png" alt="Figure SI 2.2: Avoid decorative colors">
<figcaption>Figure SI 2.2: Avoid decorative colors</figcaption>
</figure>
<p>If colors serve merely decorative purpose in one instance, using them for meaning in another instance (e.g. for
highlighting) becomes difficult. Therefore use colors only if they convey meaning, see Figure SI 2.2.</p>
<h3 id="si-23-avoid-decorative-fonts">SI 2.3 Avoid decorative fonts</h3>
<figure><img src="img/si-2.3.png" alt="Figure SI 2.3: Avoid decorative fonts">
<figcaption>Figure SI 2.3: Avoid decorative fonts</figcaption>
</figure>
<p>A normal typeface and clear fonts increase legibility. Save bold and cursive fonts for making distinctions,
see Figure SI 2.3.</p>
<h3 id="si-3-replace-with-cleaner-layout">SI 3 Replace with cleaner layout</h3>
<p>Using a cleaner method of visualization decreases the amount of ink necessary to convey a message.</p>
<h3 id="si-31-replace-grid-lines-and-value-axes-with-data-labels">SI 3.1 Replace grid lines and value axes with data
labels</h3>
<figure><img src="img/si-3.1.png" alt="Figure SI 3.1: Replace grid lines and value axes with datalabels">
<figcaption>Figure SI 3.1: Replace grid lines and value axes with datalabels</figcaption>
</figure>
<p>Using integrated data labels can make value axes, tick marks, and gridlines superfluous, see Figure SI 3.1.
Gridlines, however, can be useful in charts with missing reference points as might be the case in charts with
many data series and data points, or in small charts (e.g. small multiples).</p>
<h3 id="si-32-avoid-vertical-lines-by-right-aligning-data">SI 3.2 Avoid vertical lines by right-aligning data</h3>
<figure><img src="img/si-3.2.png" alt="Figure SI 3.2: Avoid vertical lines by right-aligning data">
<figcaption>Figure SI 3.2: Avoid vertical lines by right-aligning data</figcaption>
</figure>
<p>Omit all avoidable elements to make tables more straightforward. Avoid vertical lines by right-aligning numerical
values and the corresponding column headers, see Figure SI 3.2.</p>
<h3 id="si-4-avoid-redundancies">SI 4 Avoid redundancies</h3>
<p>Avoiding redundant terms usually increases the legibility of charts and tables. In some cases, a certain amount
of redundancy can be helpful like when the time period named in the chart title also appears in said chart. But
unnecessary redundancy impedes comprehension like when naming the year twelve times in a chart with twelve
monthly category labels.</p>
<h3 id="si-41-avoid-superfluous-extra-words">SI 4.1 Avoid superfluous extra words</h3>
<figure><img src="img/si-4.1.png" alt="Figure SI 4.1: Avoid superfluous extra words">
<figcaption>Figure SI 4.1: Avoid superfluous extra words</figcaption>
</figure>
<p>Extra words such as “sum” and “total” are redundant because they add no value to the meaning of the term they
accompany. No difference exists between “Europe” and “Sum of Europe”. Extra words make it harder to read text
elements, see Figure SI 4.1.</p>
<h3 id="si-42-avoid-obvious-terms">SI 4.2 Avoid obvious terms</h3>
<figure><img src="img/si-4.2.png" alt="Figure SI 4.2: Avoid obvious terms">
<figcaption>Figure SI 4.2: Avoid obvious terms</figcaption>
</figure>
<p>Terms such as “chart analysis”, “development”, or “comment” are redundant because they name something already
shown, see Figure SI 4.2. Other obvious terms in charts and tables are “table”, “statistics”,
“report”,“visualization”, “structure”, or “trend”.</p>
<h3 id="si-43-avoid-repeated-words">SI 4.3 Avoid repeated words</h3>
<figure><img src="img/si-4.3.png" alt="Figure SI 4.3: Avoid repeated words">
<figcaption>Figure SI 4.3: Avoid repeated words</figcaption>
</figure>
<p>Repeated words in legends, axis labels, row headers, etc. such as“division” in “division A”, “division B”, etc.
or “2017” in “Q1 2017”,“Q2 2017”, etc. should be avoided, see Figure SI 4.3. Omitting repeated words usually
increases the degree of legibility.</p>
<h3 id="si-5-avoid-distracting-details">SI 5 Avoid distracting details</h3>
<p>In addition to avoiding noise and redundancy, omitting nonessential, distracting information details facilitates
comprehension. Examples include unnecessarily large numbers and disproportionately detailed information in
project or product overviews.</p>
<h3 id="si-51-avoid-labels-for-small-values">SI 5.1 Avoid labels for small values</h3>
<figure><img src="img/si-5.1.png" alt="Figure SI 5.1: Avoid labels for small values">
<figcaption>Figure SI 5.1: Avoid labels for small values</figcaption>
</figure>
<p>Labels of small values are often hard to position and rarely contribute to the comprehension of the message.
Therefore they can be avoided in most cases, see Figure SI 5.1. However, add them when special reference is made
to them. If it is necessary to label these small values or small visualization elements, <em>assisting
lines</em>
might be necessary.</p>
<h3 id="si-52-avoid-long-numbers">SI 5.2 Avoid long numbers</h3>
<figure><img src="img/si-5.2.png" alt="Figure SI 5.2: Avoid long numbers">
<figcaption>Figure SI 5.2: Avoid long numbers</figcaption>
</figure>
<p>Numbers with more than three digits in charts and four digits in tables are hard to read; moreover, such
precision is seldom necessary to understand the message, see Figure SI 5.2.</p>
<h3 id="si-53-avoid-unnecessary-labels">SI 5.3 Avoid unnecessary labels</h3>
<figure><img src="img/si-5.3.png" alt="Figure SI 5.3: Avoid unnecessary labels">
<figcaption>Figure SI 5.3: Avoid unnecessary labels</figcaption>
</figure>
<p>Omit labels for data points that do not represent extreme values or values of special importance, see Figure SI
5.3.</p>
<p>→]()</p>
<h2 id="condense--increase-information-density">CONDENSE – Increase information density</h2>
<p>CONDENSE covers all aspects of increasing information density in reports and presentations.</p>
<p><em>Increasing information density</em> means that all reports and presentations include all information that is
necessary to understand the respective message on one page.</p>
<p>This chapter covers using small components, utilizing space, as well as adding data, elements, and objects.</p>
<ol>
<li><a href="#co-1-use-small-components">Use small components</a></li>
<li><a href="#co-2-maximize-use-of-space">Maximize use of space</a></li>
<li><a href="#co-3-add-data">Add data</a></li>
<li><a href="#co-4-add-elements">Add elements</a></li>
<li><a href="#co-5-add-objects">Add objects</a></li>
</ol>
<h3 id="co-1-use-small-components">CO 1 Use small components</h3>
<p>The need for a higher level of information density requires to display all objects, elements, and signs as small
as possible, while still being legible. Different technical parameters apply to printed material, screen
displays, and projected slides.</p>
<h3 id="co-11-use-small-fonts">CO 1.1 Use small fonts</h3>
<figure><img src="img/co-1.1.png" alt="Figure CO 1.1: Use small fonts">
<figcaption>Figure CO 1.1: Use small fonts</figcaption>
</figure>
<p>In general, avoid oversize fonts. They needlessly waste space, see Figure CO 1.1.</p>
<h3 id="co-12-use-small-elements">CO 1.2 Use small elements</h3>
<figure><img src="img/co-1.2.png" alt="Figure CO 1.2: Use small elements">
<figcaption>Figure CO 1.2: Use small elements</figcaption>
</figure>
<p>Small elements increase clarity. Large-scale symbols and highlights are not more suitable than smaller symbols
and highlights, see Figure CO 1.2.</p>
<h3 id="co-13-use-small-objects">CO 1.3 Use small objects</h3>
<figure><img src="img/co-1.3.png" alt="Figure CO 1.3: Use small objects">
<figcaption>Figure CO 1.3: Use small objects</figcaption>
</figure>
<p>The size of charts and tables in reports and presentations should not be as large as possible, rather as small as
possible – yet only so small so that the objects and all its details and labels can be read easily. This
provides room for more information and therefore better understanding of the context, see Figure CO 1.3.</p>
<h3 id="co-2-maximize-use-of-space">CO 2 Maximize use of space</h3>
<p>Utilizing free space is the fastest and easiest way to increase information density. Make better use of
needlessly wide margins and frames, or blank or little used pages by filling them with helpful data pertaining
to the context.</p>
<h3 id="co-21-use-narrow-page-margins">CO 2.1 Use narrow page margins</h3>
<figure><img src="img/co-2.1.png" alt="Figure CO 2.1: Use narrow page margins">
<figcaption>Figure CO 2.1: Use narrow page margins</figcaption>
</figure>
<p>The page layout is often dominated by corporate design standards not made for high information density but for
attractive design, sacrificing valuable space to layout elements such as extra wide page margins, see Figure CO
2.1.</p>
<h3 id="co-22-reduce-empty-space">CO 2.2 Reduce empty space</h3>
<figure><img src="img/co-2.2.png" alt="Figure CO 2.2: Reduce empty space">
<figcaption>Figure CO 2.2: Reduce empty space</figcaption>
</figure>
<p>Reduce empty space to increase information density. This applies not only to the page layout (see Figure CO 2.1)
but also to the layout of report objects such as charts and tables (see Figure CO 2.2).</p>
<h3 id="co-3-add-data">CO 3 Add data</h3>
<p>Adding more data to an existing visualization increases information density and helps better understand the
context.</p>
<h3 id="co-31-add-data-points">CO 3.1 Add data points</h3>
<figure><img src="img/co-3.1.png" alt="Figure CO 3.1: Add data points">
<figcaption>Figure CO 3.1: Add data points</figcaption>
</figure>
<p>Displaying more data points does not jeopardize the comprehension of numerical data. For example, a monthly
statistic of staff numbers over twelve months in a year would be understood just as quickly as for the same data
series with twelve months for each of the last three years – in other words, a total of 36 data points instead
of twelve. Usually, interesting relationships are only detected with an increased number of elements in a data
series (see Figure CO 3.1).</p>
<h3 id="co-32-add-dimensions">CO 3.2 Add dimensions</h3>
<figure><img src="img/co-3.2.png" alt="Figure CO 3.2: Add dimensions">
<figcaption>Figure CO 3.2: Add dimensions</figcaption>
</figure>
<p>A very useful way to increase information density is to show more than two dimensions of a business situation. A
chart with only one dimension (such as in a pie chart), visualizes only mundane things easily stated in a simple
sentence. Already charts with two dimensions can yield very interesting relationships – yet those charts with
three and more dimensions yield structures leading to completely new insights (see Figure CO 3.2).</p>
<h3 id="co-4-add-elements">CO 4 Add elements</h3>
<p>It is often appropriate to use two or more basic chart types (either horizontal or vertical) to build
<em>combined
charts</em> with a higher information density. <em>Combined charts</em> are treated as one entity as opposed
to multiple charts. <em>Combined charts</em> can be built both out from horizontal or vertical charts.
</p>
<p>There are three types of combined charts depending on their type of combination:<em>Overlay
charts</em>, <em>multi-tier charts</em>, and <em>extended charts</em>. Additionally, chart elements can be
embedded in tables and explanations can be integrated.</p>
<h3 id="co-41-show-overlay-charts">CO 4.1 Show overlay charts</h3>
<figure><img src="img/co-4.1.png" alt="Figure CO 4.1: Show overlay charts">
<figcaption>Figure CO 4.1: Show overlay charts</figcaption>
</figure>
<p>In an <em>overlay chart</em>, two or more basic charts overlap. These overlapping charts always use the same
category axis.</p>
<p><em>Overlay charts</em> can facilitate comprehension such as in the combination of the development of sales (a
series of columns) and the return on sales in percent (a line). However, this approach can only be used for a
few chart combinations, see Figure CO 4.1.</p>
<figure><img src="img/co-4.1-1.png"
alt="Figure CO 4.1-1: Overlay chart with lines and columns using different valueaxes">
<figcaption>Figure CO 4.1-1: Overlay chart with lines and columns using different valueaxes</figcaption>
</figure>
<p><em>Overlay charts</em> frequently use different value axes. A <em>column chart</em> representing a measure (e.g.
sales) combined with a <em>line chart</em> representing another measure (e.g. employees) is a typical example.
</p>
<figure><img src="img/co-4.1-2.png"
alt="Figure CO 4.1-2: Overlay chart with columns and lines using the same valueaxis">
<figcaption>Figure CO 4.1-2: Overlay chart with columns and lines using the same valueaxis</figcaption>
</figure>
<p>Sometimes, the same value axis is used as well. A <em>column chart</em> representing a measure (e.g. sales)
combined with a <em>line chart</em> representing the same measure (e.g. industry average) is a typical example
for such an <em>overlay chart</em>.</p>
<figure><img src="img/co-4.1-3.png" alt="Figure CO 4.1-3: Overlay column chart with integratedvariances">
<figcaption>Figure CO 4.1-3: Overlay column chart with integratedvariances</figcaption>
</figure>
<p>Column or bar charts with <em>integrated variances</em> (variances displayed within the columns or bars) are
other
typical example for <em>overlay charts</em> using the same value axis (see the last two figures).</p>
<p>Compared to two-tier charts, this presentation of two data series uses much less space. The disadvantages,
though, are twofold: First, it is difficult to label the data of both the primary and secondary chart. Second,
the development over time (horizontal axis) respectively the structure (vertical axis) of the primary chart is
difficult to see.</p>
<figure><img src="img/co-4.1-4.png" alt="Figure CO 4.1-4: Overlay bar chart with integratedvariances">
<figcaption>Figure CO 4.1-4: Overlay bar chart with integratedvariances</figcaption>
</figure>
<p>Suggestion: If there is enough space, use multi-tier charts instead.</p>
<h3 id="co-42-show-multi-tier-charts">CO 4.2 Show multi-tier charts</h3>
<figure><img src="img/co-4.2.png" alt="Figure CO 4.2: Show multi-tier charts">
<figcaption>Figure CO 4.2: Show multi-tier charts</figcaption>
</figure>
<p><em>Use multi-tier charts</em> to increase information density by adding additional tiers to the same category
axis for analyses on the same basic data. Multi-tier charts are most frequently used for displaying variances
along with the basic values, see Figure CO 4.2.</p>
<figure><img src="img/co-4.2-1.png" alt="Figure CO 4.2-1: Horizontal multi-tier charts">
<figcaption>Figure CO 4.2-1: Horizontal multi-tier charts</figcaption>
</figure>
<p>In a <em>two-tier chart</em>, a <em>secondary chart</em> is shifted in parallel to the category axis of
the <em>primary chart</em>. For horizontal charts the secondary chart appears above the primary chart, for
vertical charts the secondary chart appears to the right of the primary chart.</p>
<p>In both cases, the <em>category axes</em> of the primary charts are reduplicated in the secondary charts, usually
having a different semantic scenario design.</p>
<figure><img src="img/co-4.2-2.png" alt="Figure CO 4.2-2: Vertical multi-tier chart">
<figcaption>Figure CO 4.2-2: Vertical multi-tier chart</figcaption>
</figure>
<p>Both the primary and the secondary charts have their own value axes. Value axes showing the same currency or the
same physical unit should be scaled identically.</p>
<p>In a <em>three-tier chart</em> a third chart appears above a horizontal or to the right of a vertical two-tier
chart. In special cases, more than three tiers can be combined.</p>
<p>Improve the interpretation of a primary chart showing grouped bars for actual and plan data by adding variances.
In the second and third figure a secondary chart with absolute variances and a tertiary pin chart with relative
variances are combined.</p>
<h3 id="co-43-show-extended-charts">CO 4.3 Show extended charts</h3>
<figure><img src="img/co-4.3.png" alt="Figure CO 4.3: Show extended charts">
<figcaption>Figure CO 4.3: Show extended charts</figcaption>
</figure>
<p>An <em>extended chart</em>, arranges additional charts next to the primary chart by virtually extending the
category axis. This way of increasing information density often is used when displaying context information such
as market averages or competitor figures, see Figure CO 4.3.</p>
<figure><img src="img/co-4.3-1.png" alt="Figure CO 4.3-1: Horizontal extended chart">
<figcaption>Figure CO 4.3-1: Horizontal extended chart</figcaption>
</figure>
<p>For horizontal charts, additional charts appear to the left or right of the primary chart, for vertical charts,
above or below. In both cases, position the <em>category axes</em> of the additional charts on a virtual
extension of the category axes of the primary chart.</p>
<p>In an extended chart, use the same value axis for both the primary and the additional charts.</p>
<p>Improve the interpretation of a primary chart by adding extended charts showing the same values from a different
perspective. In the figure on the left, a secondary <em>grouped column chart</em> at the right hand side shows
the monthly average.</p>
<h3 id="co-44-embed-chart-elements-in-tables">CO 4.4 Embed chart elements in tables</h3>
<figure><img src="img/co-4.4.png" alt="Figure CO 4.4: Embed chart elements in tables">
<figcaption>Figure CO 4.4: Embed chart elements in tables</figcaption>
</figure>
<p>Increase the information density of tables by using <em>chart elements</em>, see Figure CO 4.4. Bars, warning
dots, sparklines, and traffic lights are the predominant chart element types in tables.</p>
<p><strong>Table bars</strong></p>
<p><em>Table bars</em> are bar charts integrated into tables. The categories of these bar charts must correspond to
the rows of a table. Both single bar charts with single bars or pins and waterfall bar charts are powerful means
to visualize the absolute figures and variances in tables. Most recommendations concerning vertical chart types
can be applied to <em>table bars</em>.</p>
<p><strong>Warning dots</strong></p>
<p>Not to be confused with <em>traffic lights</em>, <em>warning dots</em> can be a good solution in highlighting
important negative, positive, or questionable parts of a table. It is important to use only very few warning
dots in one table.</p>
<p><strong>Sparklines</strong></p>
<p>Omit <em>sparklines</em> if not scaled properly. Individually scaled sparklines can be misleading because small
fluctuations in a series of other small fluctuations look the same as big fluctuations in a series of big
fluctuations. However, sparklines with proper scaling (e.g. indexed) can be helpful.</p>
<p><strong>Traffic lights</strong></p>
<p><em>Traffic lights</em> contain little information, as they represent no more than three (red, green, yellow)
states. Use them only if there is no more information to be conveyed than those two or three states (e.g. “yes”
or “no”). In all other cases, replace traffic lights with more suitable means of representation, such
as <em>table bars</em>.</p>
<h3 id="co-45-embed-explanations">CO 4.5 Embed explanations</h3>
<figure><img src="img/co-4.5.png" alt="Figure CO 4.5: Embed explanations">
<figcaption>Figure CO 4.5: Embed explanations</figcaption>
</figure>
<p>Both the density of information and the level of comprehension increase when explanations are embedded into
charts and tables (this applies to written reports and handouts only). When the explanation refers directly to
the visual presentation in question, it helps to establish a connection and speeds up comprehension, see Figure
CO 4.5.</p>
<h3 id="co-5-add-objects">CO 5 Add objects</h3>
<p>Reports and presentation material consist of one or more <em>pages</em>. The content of one page can be viewed
together without referring to other content, e.g. flipping to other pages.</p>
<p>Reports and presentation material often arrange more than one chart on one page. While this increases information
density and fosters a higher level of comparability, it presents a design challenge: A uniform notation concept
and consistent scaling are even more important than on pages with single charts.</p>
<p>The most important types of pages with multiple objects are small multiples and multi-charts (including <em>ratio
trees</em>).</p>
<h3 id="co-51-show-small-multiples">CO 5.1 Show small multiples</h3>
<figure><img src="img/co-5.1.png" alt="Figure CO 5.1: Show small multiples">
<figcaption>Figure CO 5.1: Show small multiples</figcaption>
</figure>
<p>Substantially improve the comprehension of complex relationships by displaying charts of the same type and the
same scale on the same page. These charts are called <em>small multiples</em>, see Figure CO 5.1.</p>
<p>Typical applications are charts with different countries, products, or projects placed next to each other. Of
course, there is an upper limit to the number of charts on one page, depending mainly on the page- and font-size
used.</p>
<figure><img src="img/co-5.1-1.png" alt="Figure CO 5.1-1: Screen page with small multiples">
<figcaption>Figure CO 5.1-1: Screen page with small multiples</figcaption>
</figure>
<p>Showing <em>small multiples</em> is a good way to compare a set of up to around 25 charts. Instead of exceeding
this number on one page, a new chart called “Others” containing the accumulation of all other elements could be
a solution.</p>
<p>As mentioned in the chapter “CHECK – Ensure visual integrity”, all small multiples must use the identical scale.
</p>
<p>Working with <em>small multiples</em> can be difficult if certain charts show significantly bigger values than
others. Using a different scale for a chart with bigger values is not a feasible option, increase the size of
this chart instead.</p>
<h3 id="co-52-show-related-charts-on-one-page">CO 5.2 Show related charts on one page</h3>
<figure><img src="img/co-5.2.png" alt="Figure CO 5.2: Show related charts on one page">
<figcaption>Figure CO 5.2: Show related charts on one page</figcaption>
</figure>
<p>Different from small multiples, <em>related charts cover different topics (different measures) on one page.</em>
They mostly use different scales, too. This arrangement of charts on one page is sometimes
called <em>multi-charts</em>. But the term <em>multi-charts</em> fails to underline the fact that these charts
must have a useful relationship. It does not make sense to arrange several, completely unrelated charts on one
page.</p>
<p>This approach offers high data density and a higher level of comparability – but it can be a demanding visual and
technical challenge as a uniform notation concept, clear terms, and an understandable scaling prove even more
important (see Figure CO 5.2).</p>
<figure><img src="img/co-5.2-1.png" alt="Figure CO 5.2-1: Page showing a ratio tree">
<figcaption>Figure CO 5.2-1: Page showing a ratio tree</figcaption>
</figure>
<p>Consistent scaling of <em>multi-charts</em> can be difficult. Sometimes different scales for the same unit or
measure are inevitable. In this case, clearly indicate the use of a different scale by an appropriate mean, e.g.
scaling indicators.</p>
<p><em>Ratio trees</em> are multi-charts showing root causes. Use ratio trees to prove or explain a specific issue.
Pointing out the assumptions and root causes of variances or temporal evolvements improves understanding and is
more convincing, too. In general, the <em>ratio</em> is broken down into its components (mostly from left to
right). Thus individual charts, preferably identical size, are arranged in a tree shape structure.</p>
<p>Consistent scaling of <em>ratio trees</em> can be difficult. Sometimes different scales for the same unit or
measure are inevitable. In this case, clearly indicate the use of a different scale by an appropriate mean, e.g.
scaling indicators.</p>
<p>A typical example of a page showing a <em>ratio tree</em> is the “Return on asset” tree.</p>
<h3 id="co-53-show-chart-table-combinations">CO 5.3 Show chart-table combinations</h3>
<p>Combining charts and tables on a page is not to be confused with the integration of chart elements in tables.</p>
<p><em>Chart-table combinations</em> cover situations where a separate chart is added to a page with a table or vice
versa. In general, such a combination is very useful if both objects display supplementary data. Tables simply
listing the numbers of a chart are superfluous in most cases (see also UN 2.3 “<a
href="#un-23-unify-the-position-of-legends-and-labels%E2%80%9D">Unify the position of legends and
labels</a>.</p>
<h3 id="co-54-show-charts-and-tables-in-text-pages">CO 5.4 Show charts and tables in text pages</h3>
<p>Embedding illuminating charts and tables in the text of a written report helps the reader understanding the
message.</p>
<p>Always position charts and tables in close proximity to the phrase carrying the message, which the chart or table
supports.</p>
<p>Text pages should contain a title element like other pages. Also use a title – and, if possible, a message – for
every chart and table embedded in a text page.</p>
<h2 id="check--ensure-visual-integrity">CHECK – Ensure visual integrity</h2>
<p>CHECK covers all aspects of ensuring visual integrity in reports and presentations.</p>
<p><em>Ensuring visual integrity</em> means that reports and presentations present information in the most truthful
and the most easily understood way by avoiding misleading visuals.</p>
<p>This chapter covers avoiding manipulated axes and visualization elements, using the same scales, and showing data
adjustments.</p>
<ol>
<li><a href="#ch-1-avoid-manipulated-axes">Avoid manipulated axes</a></li>
<li><a href="#ch-2-avoid-manipulated-visualization-elements">Avoid manipulated visualization elements</a></li>
<li><a href="#ch-3-avoid-misleading-representations">Avoid misleading representations</a></li>
<li><a href="#ch-4-use-the-same-scales">Use the same scales</a></li>
<li><a href="#ch-5-show-data-adjustments">Show data adjustments</a></li>
</ol>
<h3 id="ch-1-avoid-manipulated-axes">CH 1 Avoid manipulated axes</h3>
<p>Charts serve as a means to visually compare numerical values. Manipulated axes defeat this purpose of explaining
actual interrelations.</p>
<h3 id="ch-11-avoid-truncated-axes">CH 1.1 Avoid truncated axes</h3>
<figure><img src="img/ch-1.1.png" alt="Figure CH 1.1: Avoid truncated axes">
<figcaption>Figure CH 1.1: Avoid truncated axes</figcaption>
</figure>
<p>Charts with value axes not starting at zero (“cut” axes) are not “wrong”in and of themselves, but the message to
be visually conveyed then does not correspond to the numerical values upon which the chart is based. Therefore,
value axes should generally start at zero, see Figure CH 1.1.</p>
<p>One exception to this rule exists: charts with indexed data (e.g. if the value for the index period is set to
100%) with only small variances from 100%. Here “zooming in” on the variances could be of greater value than
indicating the absolute values (starting at zero). In this case, position the category labels at the 100% line
in order to avoid misinterpretations.</p>
<h3 id="ch-12-avoid-logarithmic-axes">CH 1.2 Avoid logarithmic axes</h3>
<figure><img src="img/ch-1.2.png" alt="Figure CH 1.2: Avoid logarithmic axes">
<figcaption>Figure CH 1.2: Avoid logarithmic axes</figcaption>
</figure>
<p><em>Avoid logarithmic scales</em> because they do not allow the visual comparison of values, see Figure CH 1.2.
In business, very few applications for logarithmic axes exist (e.g. comparing growth rates of different stocks
in percent).</p>
<h3 id="ch-13-avoid-different-class-sizes">CH 1.3 Avoid different class sizes</h3>
<figure><img src="img/ch-1.3.png" alt="Figure CH 1.3: Avoid different class sizes">
<figcaption>Figure CH 1.3: Avoid different class sizes</figcaption>
</figure>
<p>If the categories represent ordered classes of elements (e.g. age classes) as used for the visualization of
distributions in histograms, use class sizes of identical width (e.g. ten years). Otherwise, true visual
comparability is impossible, see Figure CH 1.3.</p>
<h3 id="ch-2-avoid-manipulated-visualization-elements">CH 2 Avoid manipulated visualization elements</h3>
<p>Displaying values differing by orders of magnitude can be challenging. Use creative solutions instead of clipping
visualization elements or cutting value axes.</p>
<h3 id="ch-21-avoid-clipped-visualization-elements">CH 2.1 Avoid clipped visualization elements</h3>
<figure><img src="img/ch-2.1.png" alt="Figure CH 2.1: Avoid clipped visualization elements">
<figcaption>Figure CH 2.1: Avoid clipped visualization elements</figcaption>
</figure>
<p>Similar to “cut” axes, clipped visualization elements such as broken columns make visual comparisons impossible,
see Figure CH 2.1.</p>
<h3 id="ch-22-use-creative-solutions-for-challenging-scaling-issues">CH 2.2 Use creative solutions for challenging
scaling issues</h3>
<figure><img src="img/ch-2.2.png" alt="Figure CH 2.2: Use creative solutions for challenging scalingissues">
<figcaption>Figure CH 2.2: Use creative solutions for challenging scalingissues</figcaption>
</figure>
<p>Creative visualization elements can be used to compare extreme values, e.g., displaying data in two-dimensional
or even three-dimensional visualization elements allows the comparison of values differing by orders of
magnitude, see Figure CH 2.2.</p>
<p>This rule must be clearly separated from the rules of section CH 3 “<a
href="#ch-3-avoid-misleading-representations">Avoid misleading representations</a>” where area and volume
visualizations are used improperly.</p>
<h3 id="ch-3-avoid-misleading-representations">CH 3 Avoid misleading representations</h3>
<p>Graphical representations are misleading if the visual impression for the observer differs from the underlying
values.</p>
<h3 id="ch-31-use-correct-area-comparisons-prefer-linear-ones">CH 3.1 Use correct area comparisons, prefer linear
ones</h3>
<figure><img src="img/ch-3.1.png" alt="Figure CH 3.1: Use correct area comparisons, prefer linearones">
<figcaption>Figure CH 3.1: Use correct area comparisons, prefer linearones</figcaption>
</figure>
<p>Using two-dimensional representations (areas of circles, icons, or emblems) for the visualization of data is only
valid, if the size of these areas corresponds to the underlying values. The visual perception will be misleading
if the diameters of circles or the heights of icons represent the values, see Figure CH 3.1.</p>
<h3 id="ch-32-use-correct-volume-comparisons-prefer-linear-ones">CH 3.2 Use correct volume comparisons, prefer
linear ones</h3>
<figure><img src="img/ch-3.2.png" alt="Figure CH 3.2: Use correct volume comparisons, prefer linearones">
<figcaption>Figure CH 3.2: Use correct volume comparisons, prefer linearones</figcaption>
</figure>
<p>Similar to areas, the visual perception will be misleading, if the (one-dimensional) diameters or the
(two-dimensional) areas of three-dimensional visualization elements (spheres, cubes, etc.) represent the values,
see Figure CH 3.2. Even if their volumes represent the values, it is hard to perceive them properly. Prefer
linear comparisons instead.</p>
<h3 id="ch-33-avoid-misleading-colored-areas-in-maps">CH 3.3 Avoid misleading colored areas in maps</h3>
<figure><img src="img/ch-3.3.png" alt="Figure CH 3.3: Avoid misleading colored areas in maps">
<figcaption>Figure CH 3.3: Avoid misleading colored areas in maps</figcaption>
</figure>
<p>Different colored areas can be helpful to visualize the precipitation per square meter or the population density.
However, do not use colored areas for the visualization of non-area-related figures such as market shares or
return on sales. Position columns or bars of identical scale in maps instead. By the way, pie charts also work
well here (an exception to the EX 2.1 “<a href="#ex-21-replace-pie-and-ring-charts">Replace pie...”</a> because
they can be placed precisely at one point, like a city (see Figure CH 3.3).</p>
<h3 id="ch-4-use-the-same-scales">CH 4 Use the same scales</h3>
<p>Proper visual comparison requires the usage of identical scales for identical measure units, or – if this is not
possible – a clear indication of the difference. If possible, use a consistent scaling concept for the complete
report or presentation material.</p>
<h3 id="ch-41-use-identical-scale-for-the-same-unit">CH 4.1 Use identical scale for the same unit</h3>
<figure><img src="img/ch-4.1.png" alt="Figure CH 4.1: Use identical scale for the same unit">
<figcaption>Figure CH 4.1: Use identical scale for the same unit</figcaption>
</figure>
<p>If presenting more than one chart of the same unit on one page, use the identical scale for these charts, see
Figure CH 4.1. In extreme situations identical scales might not be desirable. In these exceptional cases the use
of scaling indicators (see<a href="#ch-43-use-scaling-indicators-if-necessary">CH 4.3</a> and<a
href="#un-52-unify-scaling-indicators">UN 5.2</a>) can be helpful.</p>
<h3 id="ch-42-size-charts-to-given-data">CH 4.2 Size charts to given data</h3>
<figure><img src="img/ch-4.2.png" alt="Figure CH 4.2: Size charts to given data">
<figcaption>Figure CH 4.2: Size charts to given data</figcaption>
</figure>
<p>Using identical scales in multiple charts can be demanding if the values in the charts differ by orders of
magnitude. A good solution is adapting the chart size to the given data, see Figure CH 4.2.</p>
<h3 id="ch-43-use-scaling-indicators-if-necessary">CH 4.3 Use scaling indicators if necessary</h3>
<figure><img src="img/ch-4.3.png" alt="Figure CH 4.3: Use scaling indicators if necessary">
<figcaption>Figure CH 4.3: Use scaling indicators if necessary</figcaption>
</figure>
<p>There are several ways to overcome challenging scaling problems. <em>Scaling indicators</em>, such as <em>scaling
lines</em> and <em>scaling areas</em> indicating the same numerical height (typically a power of 10) in all
charts are helpful to assist in comparing multiple charts (of the same unit) with different scales, see Figure
CH 4.3.</p>
<p>This guide suggests a <em>semantic design</em> for scaling lines and scaling areas, see UN 5.2 “<a
href="#un-52-unify-scaling-indicators">Unify scaling indicators</a>”.</p>
<h3 id="ch-44-use-outlier-indicators-if-necessary">CH 4.4 Use outlier indicators if necessary</h3>
<figure><img src="img/ch-4.4.png" alt="Figure CH 4.4: Use outlier indicators if necessary">
<figcaption>Figure CH 4.4: Use outlier indicators if necessary</figcaption>
</figure>
<p>Certain values that are very big in comparison to other values are called outliers. If an outlier is not
important for business, e.g. a big relative variance of a small value, then it is not appropriate to scale the
whole chart to this outlier. Therefore, use <em>outlier indicators</em> for unimportant outliers, see Figure CH
4.4.</p>
<p>This guide suggests a <em>semantic design</em> for outlier indicators, see UN 5.3 “<a
href="#un-53-unify-outlier-indicators">Unify outlier indicators</a>”.</p>
<h3 id="ch-45-use-magnifying-glasses">CH 4.5 Use magnifying glasses</h3>
<p>Another way to assist in scaling problems is to use “<em>magnifying glasses</em>” for zooming in on a part of a
chart with a bigger scale. Use an appropriate visualization element to mark the part of a chart to be zoomed in
and to link it to a second chart displaying the zoomed part on a bigger scale.</p>
<h3 id="ch-5-show-data-adjustments">CH 5 Show data adjustments</h3>
<p>For longer time series, currency and inflationary effects can bias the visual impression, hiding the real
development of business.</p>
<h3 id="ch-51-show-the-impact-of-inflation">CH 5.1 Show the impact of inflation</h3>
<figure><img src="img/ch-5.1.png" alt="Figure CH 5.1: Show the impact of inflation">
<figcaption>Figure CH 5.1: Show the impact of inflation</figcaption>
</figure>
<p>Making inflation effects transparent helps avoid misinterpretations of time series visualizations, see Figure CH
5.1.</p>
<h3 id="ch-52-show-the-currency-impact">CH 5.2 Show the currency impact</h3>
<figure><img src="img/ch-5.2.png" alt="Figure CH 5.2: Show the currency impact">
<figcaption>Figure CH 5.2: Show the currency impact</figcaption>
</figure>
<p>Similar to inflation effects, the adjustment of currency effects can help to avoid misinterpretations, see Figure
CH 5.2.</p>
<p>→]()</p>
<h2 id="unify--apply-semantic-notation">UNIFY – Apply semantic notation</h2>
<p>UNIFY covers all aspects of applying semantic notation in reports and presentations.</p>
<p><em>Applying semantic notation</em> means that reports and presentations follow this governing
principle:<em>Similar content should be visualized in a similar manner;</em> what looks the same should also
mean the same. On the flip side: If the content is not the same, it should not look the same.</p>
<p>In many specialized disciplines such as engineering, music, and architecture, <em>semantic notation
standards</em>
are a matter of course. The world of business communication lacks such notation standards, one of the main
reasons management reports are sometimes hard to understand. For example, no common agreement on the meaning of
various style elements such as titles, line markers, axes, highlighting indicators, etc. used in business charts
exists yet.</p>
<p>This chapter covers semantic rules for all important and frequently recurring aspects of meaning in the context
of business communication, such as terminology (e.g. words, abbreviations, number formats), descriptions (e.g.
messages, titles, legends), dimensions (e.g. measures, scenarios, time periods), analyses (e.g. comparisons and
variances), and indicators for highlighting, scaling and other purposes.</p>
<ol>
<li><a href="#un-1-unify-terminology">Unify terminology</a></li>
<li><a href="#un-2-unify-descriptions">Unify descriptions</a></li>
<li><a href="#un-3-unify-dimensions">Unify dimensions</a></li>
<li><a href="#un-4-unify-analyses">Unify analyses</a></li>
<li><a href="#un-5-unify-indicators">Unify indicators</a></li>
</ol>
<h3 id="un-1-unify-terminology">UN 1 Unify terminology</h3>
<p><em>Terms</em> are the non-visual part of business communication. Unified <em>terms and abbreviations</em> as
well
as unified formats for <em>numbers, units and dates</em> accelerate understanding.</p>
<h3 id="un-11-unify-terms-and-abbreviations">UN 1.1 Unify terms and abbreviations</h3>
<figure><img src="img/un-1.1.png" alt="Figure UN 1.1: Unify terms and abbreviations">
<figcaption>Figure UN 1.1: Unify terms and abbreviations</figcaption>
</figure>
<p>The standardization of terms and abbreviations in reports and presentations is achieved by using an unambiguous
language (see SA 4.2 “<a href="#sa-42-use-precise-words">Speak with precise words</a>“) and by unifying the
usage of terms (glossary).</p>
<p>Unify, compile and explain all terms and abbreviations used in reports and presentations in a clearly
arranged <em>glossary</em> including abbreviations and definitions, see Figure UN 1.1.</p>
<p>A glossary with terms and abbreviations in more than one language might be necessary in order to avoid different
translations.</p>
<p>Often the names of business measures are too long for charts and tables. Use abbreviations instead. It might be a
good solution to define <em>short abbreviations</em> (e.g. to <em>A/R</em> for <em>Accounts Receivable</em> be
used
in table <em>column</em> headers) and <em>long abbreviations</em> (e.g. <em>Acc. Receiv.</em> to be used in
table <em>row</em> headers).</p>
<p>Unified terms and abbreviations for the notation of scenarios and time periods are covered in the respective
sections.</p>
<h3 id="un-12-unify-numbers-units-and-dates">UN 1.2 Unify numbers, units, and dates</h3>
<figure><img src="img/un-1.2.png" alt="Figure UN 1.2: Unify numbers, units, and dates">
<figcaption>Figure UN 1.2: Unify numbers, units, and dates</figcaption>
</figure>
<p>The uniform use of formats for numbers, units and dates will enhance legibility, see Figure UN 1.2.</p>
<p><strong>Numbers</strong></p>
<p>Different languages and countries use different <em>number formats</em>, e.g. 1.234.567,00 (D); 1,234,567.00
(USA); 1’234’567,00 or 1’234’567.00 (CH).</p>
<p>It is important to unify the number formats in all reports and presentations. The <em>International System of
Units (SI)</em> as described in “ISO 80000-1” recommends the following notation:</p>
<ul>
<li>Thousand delimiter: 1 234 (blank space)</li>
<li>Decimal sign: 1,23 or 1.23 (SI allows both versions)</li>
</ul>
<p>Do not use long numbers in order to avoid distraction and to concentrate on the essentials, see also SI 5.2 “<a
href="#si-52-avoid-long-numbers">Avoid long numbers</a>”. Use <em>currency prefixes</em> and <em>metric
prefixes</em> to limit the number of digits to a maximum of three in charts and four in tables.</p>
<p>The most common formats for <em>negative values</em> are “-123” and “(123)”. Use the same format for all negative
values.</p>
<p><em>Positive values</em> do not have a plus sign, unless they represent variances.</p>
<p><strong>Currencies</strong></p>
<p>Use the standard <em>currency abbreviations</em> based on<a href="http://en.wikipedia.org/wiki/ISO_4217">ISO
4217</a>. ISO 4217 provides a set of currency abbreviations using three-letter acronyms such as EUR, CHF,
USD, and GBP. The use of special currency symbols such as €, $, and ₤ is not recommended if a report includes
many different currencies.</p>
<p>Use “metric prefixes” in combination with the currency units for monetary values expressed in thousands or
millions. Use lower case characters to differentiate the prefixes from the currency abbreviations and use single
digit metric prefixes to save space, such as “k” for thousand, “m” for million and “b” for billion. The
following shows the correct use of currency metric prefixes with EUR:</p>
<p>1 bEUR = 1 000 mEUR = 1 000 000 kEUR = 1 000 000 000 EUR</p>
<p>(The metric prefixes for physical units are “M” for “millions”and “G” for “billions”. Nevertheless, this guide
suggests using “m” and“b” for currency metric prefixes, as “mEUR” and “bEUR” which is easier to understand than
MEUR and GEUR.)</p>
<p><strong>Physical units</strong></p>
<p>For <em>physical units</em> use the<a
href="http://en.wikipedia.org/wiki/International_System_of_Units">International System of Units (SI)</a>
such as kg, t, m, km, etc.</p>
<p>In the case of non-monetary values expressed in thousands or millions, use<a
href="http://en.wikipedia.org/wiki/Metric_prefixes">metric prefixes</a> suggested by the <em>International
System of Units</em> such as “G” for billion, “M” for million, and “k” for thousand.</p>
<p><strong>Dates</strong></p>
<p><em>Dates</em> are best displayed using<a href="http://en.wikipedia.org/wiki/Iso_date">ISO 8601</a>, an
international standard covering the exchange of date and time-related data: YYYY-MM-DD, e.g. “2015-12-31”.</p>
<p>Other significant notation principles regarding time-related aspects will be dealt in UN 3.3 “<a
href="#un-33-unify-time-periods-and-points-of-time">Unify time periods</a>”.</p>
<h3 id="un-2-unify-descriptions">UN 2 Unify descriptions</h3>
<p><em>Descriptions</em> are textual elements that describe the visual elements in reports and presentations
facilitating comprehension. The following suggests unified layouts for every kind of <em>descriptions</em>.</p>
<h3 id="un-21-unify-messages">UN 2.1 Unify messages</h3>
<figure><img src="img/un-2.1.png" alt="Figure UN 2.1: Unify messages">
<figcaption>Figure UN 2.1: Unify messages</figcaption>
</figure>
<p>The <em>message</em> the author intends to convey to the reader or audience is best recognized, if the position
and the layout of the message is always the same, see Figure UN 2.1.</p>
<figure><img src="img/un-2.1-1.png" alt="Figure UN 2.1-1: Message text line">
<figcaption>Figure UN 2.1-1: Message text line</figcaption>
</figure>
<p>The <em>notation</em> of messages should be two text lines at the top of a report or presentation page, either a)
above the title (see figure on the left) or b) right of the title. Position b) is not structured as clearly as
position a) but it helps saving valuable vertical space especially on pages in landscape format.</p>
<h3 id="un-22-unify-titles-and-subtitles">UN 2.2 Unify titles and subtitles</h3>
<figure><img src="img/un-2.2.png" alt="Figure UN 2.2: Unify titles and subtitles">
<figcaption>Figure UN 2.2: Unify titles and subtitles</figcaption>
</figure>
<p><em>Titles</em> identify the content of pages and their objects in its entirety, omitting nothing necessary to
understand the content, see Figure UN 2.2. In contrast to messages, titles do not contain any evaluating
aspects, such as interpretations, conclusions, or propositions.</p>
<p>If there is more than one object on a page, use <em>page titles</em> for entire pages, slides, or screens
and <em>sub titles</em> for different objects on a page.</p>
<p><strong>Page titles</strong></p>
<p><em>Page titles</em> identify the content of a page. In general, three lines suffice to completely describe the
content of a page:</p>
<p><strong>Title line 1: Reporting unit</strong></p>
<p>Element(s) of a structure dimension representing the object of the report, typically a legal entity, an
organization unit, a line of business, a project, etc. or combinations thereof, e.g.</p>
<ul>
<li>ABC Corporation</li>
<li>ABC Corporation and its main competitors</li>
<li>ABC Corporation, European division</li>
<li>ABC Corporation, European division, Project B</li>
</ul>
<p>Add filter information if the elements are not exhaustive, e.g.</p>
<ul>
<li>International Chocolate Corporation, top ten clients</li>
<li>International Chocolate Corporation, divisions with negative EBIT in 2016</li>
</ul>
<p>If title line 1 becomes too long its content can be split into two lines, e.g.</p>
<blockquote>
International Chocolate Corporation, European division<br>
Top ten clients
</blockquote>
<p><strong>Title line 2: Business measure(s)</strong></p>
<p>Element(s) of the measure dimension such as sales, profit, and shipment. Business measures are measured either in
currency units (e.g. EUR, USD) or in physical units (e.g. #, kg, t). Use metric prefixes (e.g. k, m, b) where
appropriate. Measures are written in bold font, their units are written in regular font. Examples are:</p>
<ul>
<li><strong>Net sales</strong> in mEUR</li>
<li><strong>Net sales</strong> in mEUR,<strong>margin</strong> in %</li>
<li><strong>Headcount</strong> in #</li>
</ul>
<p>Use a suiting name for a <em>set of measures</em>, if more than two measures have to be presented on one page.
Examples are:</p>
<ul>
<li><strong>Income statement</strong> in kEUR</li>
<li><strong>ROI tree</strong> in mEUR</li>
<li><strong>Balanced scorecard</strong></li>
<li><strong>Product market portfolio</strong></li>
</ul>
<p>Use footnotes if parts of the measures are redundant or if parts of the measures are of minor importance for
understanding. Examples are:</p>
<ul>
<li><strong>Net sales</strong> in mEUR (without intercompany sales) – simpler:<strong>Net sales*</strong> in
mEUR</li>
<li><strong>Operating margin</strong> in mEUR (non-IFRS) – simpler:<strong>Operating margin*</strong> in mEUR
</li>
</ul>
<p><em>Additional information</em> about the way presenting the content can help to understand better the respective
page. They might concern structure dimensions, e.g.</p>
<ul>
<li><strong>Profit</strong> in mUSD, by products</li>
<li><strong>Net sales</strong> in kEUR, by products and by countries</li>
</ul>
<p>or they might be analytical annotations, e.g.</p>
<ul>
<li><strong>Net sales</strong> in mEUR and<strong>profit</strong>in mEUR, sorted by net sales (↓)</li>
<li><strong>Full time equivalents</strong> in #, indexed (2012 = 100%)</li>
<li><strong>Gross margin</strong> in kUSD, top ten</li>
</ul>
<p>or even combinations of structure dimensions and analytical annotations, e.g.</p>
<ul>
<li><strong>Net sales</strong> in mEUR, by countries, sorted by net sales (↓)</li>
<li><strong>Full time equivalents</strong> in #, by offices, indexed (2012 = 100%)</li>
</ul>
<p>If title line 2 becomes too long its content can be split into two lines, e.g.:</p>
<blockquote>
<strong>Full time equivalents</strong> in #, by offices<br>
Indexed (2012 = 100%)
</blockquote>
<blockquote>
<strong>Net sales</strong> in mEUR, by countries<br>
Sorted by net sales (↓)
</blockquote>
<p><strong>Title line 3: Time period(s) scenario(s), and variance(s)</strong></p>
<figure><img src="img/un-2.2-1.png" alt="Figure UN 2.2-1: Title lines">
<figcaption>Figure UN 2.2-1: Title lines</figcaption>
</figure>
<p>Element(s) of the time dimension (e.g. years, months), of the <em>scenario</em> dimension (e.g. actual, plan),
and
variances (e.g. ΔPL, ΔPL%) if necessary.</p>
<p>In general, elements of the <em>time</em> dimension (e.g. 2016, 2016-Q1) are necessary for understanding.
Elements
of the <em>scenario</em> dimension (e.g. AC, PL, FC) and <em>variances</em> are added if they help to understand
the page content faster. If only actual values are presented, the attribute AC can be omitted.</p>
<p>Display the time element first if both time and scenario elements are shown in title line 3.</p>
<p>Use “&amp;” (ampersand sign) when title elements together make up a time series, e.g. “AC&amp;PL” (without
blanks) if the first 8 months of a year present AC values and the last 4 months present PL values.</p>
<p>Use “and” when different elements are presented for all time periods, e.g. “AC and PY” if all 12 months of a year
present both AC and PY values.</p>
<p>Examples of alternative arrangements in <em>title line 3</em> are:</p>
<ul>
<li>2017-Q1</li>
<li>2016-03..2017-02</li>
<li>2017 AC and PL</li>
<li>2017 AC&amp;FC and PY</li>
<li>2016 AC, 2017..2021 PL, or: 2014..21 AC&amp;PL, or: 2014..21</li>
<li>2016 AC and PL and ΔPL, or: 2016 AC and PL, or: 2016</li>
</ul>
<p>Keep it clear and easy to understand – too many elements tend to be confusing. In many cases the information
depicted in column headers of tables and legends of data series in charts are sufficient and give better and
quicker insight than long texts in <em>title line 3</em>. In any case, rules for abbreviating time periods and
dates as well as the rules for abbreviating scenarios and variances must be followed.</p>
<p>In general, position <em>page titles</em> at the very upper left corner of a page, directly underneath the
message
(if a message exists). Alternatively, position them at the same height as the message if there is not enough
space – preferably on the left hand side of the message.</p>
<p>Here are some typical examples of <em>page titles</em>:</p>
<blockquote>
Chocolate Corp.<br>
<strong>Gross profit</strong> in mUSD<br>
2016
</blockquote>
<blockquote>
Construction Inc., Division EMEA<br>
<strong>Net sales</strong> in mEUR, <strong>profit margin</strong> in %<br>
_2016-Q3 (AC, PL)
</blockquote>
<blockquote>
Beverage Corporation<br>
<strong>Product market portfolio</strong><br>
2015 and 2016
</blockquote>
<blockquote>
Milk & Cheese Corp.<br>
<strong>Shipments</strong> in t, by product, by country<br>
2016-W01..10
</blockquote>
<p><strong>Subtitles</strong></p>
<figure><img src="img/un-2.2-2.png" alt="Figure UN 2.2-2: Subtitles">
<figcaption>Figure UN 2.2-2: Subtitles</figcaption>
</figure>
<p><em>Subtitles</em> identify either page segments or objects (e.g. charts and tables) within a page with multiple
objects. They complement the identification information already given in the page title. Subtitles display
identifiers that differ from object to object on a page. Put identifiers that are identical for all objects of a
page in the page title and not in the subtitles.</p>
<p>In most cases, one line is sufficient for subtitles because different elements of only one dimension have to be
identified. Typical examples are:</p>
<blockquote>
<strong>Revenue</strong> in mEUR<br>
Apples<br>
2014..2016
</blockquote>
<blockquote>
<strong>Sales</strong> in SKU<br>
Pears<br>
2016-Q1..Q4
</blockquote>
<blockquote>
<strong>Avg. price</strong> in EUR/SKU<br>
Oranges<br>
2016-10..12
</blockquote>
<p><em>Subtitles</em> are positioned above the respective objects (charts, tables, etc.) which they identify –
either centered or left-aligned.</p>
<p><strong>Titles on screen pages</strong></p>
<p>Unlike titles on printed pages, the layout of titles on screen pages can depend on the device (responsive
design). For small devices in landscape format e.g. writing the three title lines in one line separated by a “|”
(pipe sign) is a valid solution.</p>
<p>Titles on screen pages can also mutually interfere with interactive navigation objects such as drop-down boxes
for selection and check boxes for filtering. These navigation objects can contain redundant title information,
but they cannot replace the title or parts of it. Hide these navigation objects when they are not in use or when
the screen page is being printed.</p>
<h3 id="un-23-unify-the-position-of-legends-and-labels">UN 2.3 Unify the position of legends and labels</h3>
<figure><img src="img/un-2.3.png" alt="Figure UN 2.3: Unify the position of legends and labels">
<figcaption>Figure UN 2.3: Unify the position of legends and labels</figcaption>
</figure>
<p>A standardized notation of <em>legends</em> and <em>labels</em> will improve legibility and speed up
comprehension
of charts, see Figure UN 2.3.</p>
<p><strong>Legends</strong></p>
<p><em>Legends</em> (also called “<em>data series labels</em>”) identify data series.</p>
<p>If possible, integrate legends into charts, not positioned externally. Write legends horizontally for better
legibility.</p>
<p>Legends for single column charts and single bar charts are best integrated into the title.</p>
<figure><img src="img/un-2.3-1.png" alt="Figure UN 2.3-1: Legends of a stacked column chart">
<figcaption>Figure UN 2.3-1: Legends of a stacked column chart</figcaption>
</figure>
<p>In stacked column charts, position legends either to the left of the leftmost column or to the right of the
rightmost column.</p>
<figure><img src="img/un-2.3-2.png" alt="Figure UN 2.3-2: Legends of a stacked bar chart">
<figcaption>Figure UN 2.3-2: Legends of a stacked bar chart</figcaption>
</figure>
<p>Center legends of stacked bar charts above the top bar.</p>
<figure><img src="img/un-2.3-3.png" alt="Figure UN 2.3-3: Legend with assisting line">
<figcaption>Figure UN 2.3-3: Legend with assisting line</figcaption>
</figure>
<p><em>Assisting lines</em> can help to assign the legends to the correct <em>visualization elements</em>. In
grouped
column charts and grouped bar charts, assisting lines can also help to assign the legends to the
correct <em>visualization elements</em>.</p>
<figure><img src="img/un-2.3-4.png" alt="Figure UN 2.3-4: Legends of a line chart">
<figcaption>Figure UN 2.3-4: Legends of a line chart</figcaption>
</figure>
<p>In line charts, position legends either to the right of the line end or close to the course of the line.</p>
<figure><img src="img/un-2.3-5.png" alt="Figure UN 2.3-5: Legends in a chart with two value axes">
<figcaption>Figure UN 2.3-5: Legends in a chart with two value axes</figcaption>
</figure>
<p>For charts with two value axes, externally positioned legends next to symbols can be a good choice. When helpful,
integrate these legends into the chart by positioning them next to typical points or bubbles.</p>
<p><strong>Labels</strong></p>
<p><em>Labels</em> (more precise:<em>data labels</em>) assign the data values to the respective visualization
elements.</p>
<p>Omit labels of small visualization elements, use labels with not more than three digits, and avoid unnecessary
and distracting labels (see also the SIMPLIFY rules SI 5 “<a href="#si-5-avoid-distracting-details">Avoid
distracting details</a>”).</p>
<p>Write labels horizontally for better legibility.</p>
<p>Position labels next to their visualization elements. If this is not possible, use lines connecting the labels to
the correct visualization elements.</p>
<figure><img src="img/un-2.3-6.png" alt="Figure UN 2.3-6: Labels in a column chart">
<figcaption>Figure UN 2.3-6: Labels in a column chart</figcaption>
</figure>
<p>In charts with horizontal category axes, position labels above or below the visualization elements, see the first
and second figure. In stacked columns, either center labels in the data points (if the data points are large
enough) or position them outside of the data points.</p>
<figure><img src="img/un-2.3-7.png" alt="Figure UN 2.3-7: Labels in a line chart">
<figcaption>Figure UN 2.3-7: Labels in a line chart</figcaption>
</figure>
<figure><img src="img/un-2.3-8.png" alt="Figure UN 2.3-8: Labels in a chart with vertical categoryaxis">
<figcaption>Figure UN 2.3-8: Labels in a chart with vertical categoryaxis</figcaption>
</figure>
<p>In charts with vertical category axes, position labels right or left of the visualization elements. In stacked
bars, either center labels in the data points (if the data points are large enough) or position them outside of
the data points.</p>
<figure><img src="img/un-2.3-9.png" alt="Figure UN 2.3-9: Labels in a chart with two value axes">
<figcaption>Figure UN 2.3-9: Labels in a chart with two value axes</figcaption>
</figure>
<p>In charts with two value axes, position labels next to the visualization elements (above or below, right or
left). Large bubble visualization elements labels can also have centered labels.</p>
<h3 id="un-24-unify-comments">UN 2.4 Unify comments</h3>
<figure><img src="img/un-2.4-1.png" alt="Figure UN 2.4-1: Unify comments">
<figcaption>Figure UN 2.4-1: Unify comments</figcaption>
</figure>
<p>Mainly in static reports, <em>comments</em> detail other elements (e.g. definitions of data series) and objects
such as charts and tables. Sometimes comments also refer to complete pages.</p>
<p>The level of comprehension increases when comments refer directly to the visual representation. Therefore,
comments on an object (e.g. chart) are integrated into that object when possible. Comment elements should be
linked to the content of tables, charts, etc. through comment references.</p>
<h3 id="un-25-unify-footnotes">UN 2.5 Unify footnotes</h3>
<figure><img src="img/un-2.5-1.png" alt="Figure UN 2.5-1: Unify footnotes">
<figcaption>Figure UN 2.5-1: Unify footnotes</figcaption>
</figure>
<p><em>Footnotes</em>, a special type of comments, provide general explanations, explanations of abbreviations, and
information that increases the credibility of the content such as the sources or the dates of retrieval and
printing. They can be omitted from slides projected on the wall, but must be included in written material.</p>
<p>Position footnotes at the bottom of a page.</p>
<h3 id="un-3-unify-dimensions">UN 3 Unify dimensions</h3>
<p>Data in reports and presentations can be viewed from various perspectives called <em>dimensions</em>. For
example,
all business measures, such as sales, profit, margin, etc., constitute a measure dimension, all months,
quarters, years, etc., a time dimension.</p>
<p>Identifying dimensions via uniform visualization will help to understand reports and presentations.</p>
<p>This section suggests visualization standards for measures, scenarios, time periods, and structure dimensions.
</p>
<h3 id="un-31-unify-measures">UN 3.1 Unify measures</h3>
<figure><img src="img/un-3.1.png" alt="Figure UN 3.1: Unify measures">
<figcaption>Figure UN 3.1: Unify measures</figcaption>
</figure>
<p>Business <em>measures</em> such as sales, profit, margin, etc. describe, report, and calculate business
situations. A standardized notation will help to comprehend the specific characteristics of measures, e.g.
whether they are basic measures or calculated ratios of measures, whether they represent value or volume
figures, flow or stock figures, or whether they have a positive or negative impact, see Figure UN 3.1.</p>
<p>The <em>visualization</em> of business measures is presented here. Their <em>definition</em>, generally given in
accounting manuals or similar documentation, is not discussed here.</p>
<p><strong>Basic measures and ratios</strong></p>
<p><em>Basic measures</em> such as “export sales” are directly derived from business processes. <em>Ratios</em> such
as “return on sales” are quotients of two basic measures.</p>
<p><strong>Basic measures</strong></p>
<p>Basic measures have either <em>currency units</em> (e.g. EUR) or <em>physical units</em> (e.g. kg). They are
neither shares of something (percentages) nor quotients of two measures.</p>
<figure><img src="img/un-3.1-1.png" alt="Figure UN 3.1-1: Monthly basic measures in a column chart">
<figcaption>Figure UN 3.1-1: Monthly basic measures in a column chart</figcaption>
</figure>
<p>Use 2/3 of the category width for the column width in <em>column charts</em> and the bar width in <em>bar
charts</em> to visualize basic measures.</p>
<figure><img src="img/un-3.1-2.png" alt="Figure UN 3.1-2: Monthly basic measures in a line chart">
<figcaption>Figure UN 3.1-2: Monthly basic measures in a line chart</figcaption>
</figure>
<p>Use thick lines for representing basic measures in <em>line charts</em>.</p>
<p><strong>Ratios</strong></p>
<p><em>Ratios</em> are quotients of two basic measures such as “return on sales”. In practice, few denominators
exist: “Sales”, “units sold”, “headcount”, and “capital”constitute the majority of all business ratios.</p>
<p>If both the enumerator and denominator have the same unit the resulting ratio has no unit. It is expressed
in <em>percent</em> (e.g. “profit in % of sales”).</p>
<p>In addition, if both enumerator and denominator have the same basic measure (e.g. “headcount”), it is called
a <em>share</em>(e.g. “gender share”).</p>
<figure><img src="img/un-3.1-3.png" alt="Figure UN 3.1-3: Monthly ratios in a column chart">
<figcaption>Figure UN 3.1-3: Monthly ratios in a column chart</figcaption>
</figure>
<p>The width of both bars and columns representing <em>ratios</em> is 1/3 of the category width, i.e. 50% of the
width of bars and columns representing <em>basic measures</em>.</p>
<figure><img src="img/un-3.1-4.png" alt="Figure UN 3.1-4: Monthly ratios in a line chart">
<figcaption>Figure UN 3.1-4: Monthly ratios in a line chart</figcaption>
</figure>
<p>Represent ratios in <em>line charts</em> with thin lines (50% of thick lines).</p>
<p><strong>Value and volume</strong></p>
<p><em>Value</em> measures such as “profit” and “capital” have currency units. <em>Volume</em> measures such as
“shipment” and “headcount” have physical units.</p>
<p><strong>Flow and stock</strong></p>
<p><em>Flow</em> measures like “net sales” relate to a certain <em>time period</em> such as months or
years. <em>Stock</em> measures like “inventory” relate to a certain <em>fixed date</em>, such as December 31st
of
2015 (at midnight).</p>
<p><strong>Positive, negative, and neutral impact</strong></p>
<p>An increase of a <em>positive measure</em> such as “profit” or “sales” positively impacts the organization’s
result.</p>
<p>An increase of a <em>negative measure</em> such as “cost” or “waste” negatively impacts the organization’s
result.
</p>
<p>An increase of <em>neutral measures</em> such as “market size” or “investment” has no direct impact to the
organization’s result.</p>
<h3 id="un-32-unify-scenarios">UN 3.2 Unify scenarios</h3>
<figure><img src="img/un-3.2.png" alt="Figure UN 3.2: Unify scenarios">
<figcaption>Figure UN 3.2: Unify scenarios</figcaption>
</figure>
<p><em>Scenarios</em>(also called data categories, data types, or versions) represent different layers of a business
model. Typical scenarios are “Actual”, “Previous year”, “Plan”, “Budget”, and“Forecast”. In special
cases <em>benchmarks</em> such as competitor data or market averages are also called scenarios.</p>
<p>Often comparisons and variances between different scenarios are presented to provide business insights.</p>
<p>There are two basic types of scenarios:</p>
<ul>
<li>
<p><strong>Actual scenarios</strong> refer to <em>measured</em> data about things that already happened in
present or past time periods. These data might not be perfectly correct because of difficulties with
systems, unclear definitions, and false data acquisition – but they are as correct as possible. The
terms we use most often for scenarios of this type are ‘Actual’ and ‘Previous year’.</p>
</li>
<li>
<p><strong>Planned scenarios</strong> refer to <em>fictitious</em> (not materialized) data. The terms we use
most often for scenarios of this type are ‘Plan’ and ‘Budget’.</p>
</li>
</ul>
<p>In-between those two basic scenario types there is a third one:</p>
<ul>
<li><strong>Forecasted scenarios</strong> refer to <em>expected</em> data which are strictly speaking fictitious
but already taking into account measured data. A typical example forexpected data is the sales forecast
based on the measured order entry. Forecasted scenarios represent a higher level of certainty than scenarios
with planned data but are not completely materialized yet. The term we use most often for scenarios of this
type is ‘Forecast’.</li>
</ul>
<p>When analyzing charts and tables, it is very important to quickly recognize whether you look at measured,
expected, or fictitious data. Readers can visually recognize these scenario types by looking at the <em>area
fill</em> of a visualization element without having to read the labels. Typical chart visualization elements
such as bars, columns, line chart markers, scenario triangles, etc. carry the semantic scenario notation.</p>
<p>In charts presenting variances, their <em>axes</em> carry the semantic scenario notation in order to show the
respective reference scenario (see<a href="#un-41-unify-scenario-analyses">UN 4.1</a>).</p>
<p>In charts with stacked columns, stacked areas, and charts with multiple lines or areas, the application of this
semantic scenario notation can become a challenge. In these cases, applying the semantic notation to the axis
instead of the columns etc. is a valid option.</p>
<p><strong>Actual scenarios: measured data</strong></p>
<figure><img src="img/un-3.2-1.png" alt="Figure UN 3.2-1: Visualization of measured data">
<figcaption>Figure UN 3.2-1: Visualization of measured data</figcaption>
</figure>
<p>Scenarios with measured data are identified by a solid dark (e.g. black or dark gray) fill for the areas of the
respective visualization elements.</p>
<p>If measured data of recent periods (“Actual”) are compared with measured data from earlier periods (e.g.
“Previous year”,“Previous month’”, “Month YoY”) the areas representing the earlier periods are identified by a
lighter solid fill (e.g. light gray).</p>
<p>The suggested two-letter codes for the most important measured data scenarios are “AC” for “Actual” and “PY” for
“Previous Year”.</p>
<p><strong>Planned scenarios: fictitious data</strong></p>
<figure><img src="img/un-3.2-2.png" alt="Figure UN 3.2-2: Visualization of fictitious data">
<figcaption>Figure UN 3.2-2: Visualization of fictitious data</figcaption>
</figure>
<p>Scenarios with fictitious data are identified by bordered (outlined, framed) areas of the respective
visualization elements. The areas within these borders literally “fill up when materializing”, e.g. when
changing from fictitious data to measured data.</p>
<p>The suggested two-letter codes for the two most important fictitious data scenarios are “PL” for “Plan” and “BU”
for“Budget”.</p>
<p><strong>Forecasted scenarios: expected data</strong></p>
<figure><img src="img/un-3.2-3.png" alt="Figure UN 3.2-3: Visualization of expected data">
<figcaption>Figure UN 3.2-3: Visualization of expected data</figcaption>
</figure>
<p>Expected data is strictly speaking fictitious, so they are also identified by bordered (outlined, framed) areas.
However, as it is based on measured data, the area fill of the respective visualization elements becomes
hatched. The color of the dark stripes correspond to the color of measured data (e.g. dark gray).</p>
<p>The suggested two-letter code for the most important expected data scenario is “FC” for “Forecast”.</p>
<h3 id="un-33-unify-time-periods-and-points-of-time">UN 3.3 Unify time periods and points of time</h3>
<figure><img src="img/un-3.3.png" alt="Figure UN 3.3: Unify time periods and points of time">
<figcaption>Figure UN 3.3: Unify time periods and points of time</figcaption>
</figure>
<p>Using standard notations for <em>time periods</em> (for flow measures) and <em>points of time</em> (for stock
measures) is important as they are frequently applied to all forms of business communication. This requires
standard notations for the visual direction of time, time period and points of time abbreviations and – in
charts with horizontal time axes – category widths, see Figure UN 3.3.</p>
<p><strong>Visual direction of time periods</strong></p>
<figure><img src="img/un-3.3-1.png" alt="Figure UN 3.3-1: Visual direction of time periods">
<figcaption>Figure UN 3.3-1: Visual direction of time periods</figcaption>
</figure>
<p>As opposed to structural comparisons, horizontal axes visualize data series over time. In tables, present data
series over time in columns. In both cases time moves from left to right.</p>
<p><strong>Time period and points of time abbreviations</strong></p>
<figure><img src="img/un-3.3-2.png" alt="Figure UN 3.3-2: Time period and points of timeabbreviations">
<figcaption>Figure UN 3.3-2: Time period and points of timeabbreviations</figcaption>
</figure>
<p>For a better understanding, use unified <em>abbreviations for time periods and points in time</em>. ISO 8601
recommends the following pattern for time periods: YYYY-MM-DD (e.g. 2017-05-13) for its unambiguousness and easy
sorting. The <em>abbreviations</em> in the figure.</p>
<p>In some countries or organizations other abbreviations such as Oct 2017, Q2 2017, W07 2017 are common. They can
also be used as long as they are used consistently.</p>
<p>A “.” (full-stop) before the period name indicates the <em>first day</em> of a time period, e.g. “.2017” for the
first day of 2017 or “.Jun” for the first day of June.</p>
<p>Likewise, append “.” (full-stop) to the period name to visualize the <em>last day</em> of a time period, e.g.
“2017.” for the last day of 2017 or “Jun.” for the last day of June.</p>
<p>The sign “..” (two full-stops) indicates a <em>time span,</em> e.g. “Jan..Mar” (without blanks) for “from January
to March.”N.B.: Use two dots instead of three dots (“ellipsis”).</p>
<p><strong>Category widths</strong></p>
<figure><img src="img/un-3.3-3.png" alt="Figure UN 3.3-3: Category widths">
<figcaption>Figure UN 3.3-3: Category widths</figcaption>
</figure>
<p>When helpful, differentiate different types of time periods with different <em>category widths</em> according to
this rule: the longer the period the wider the category segments on the category axis.</p>
<p>It might be necessary to use rather <em>wide</em> category segments to label stacked columns or
rather <em>narrow</em> category segments due to restricted dashboard space. In any case, if certain period types
have been allocated certain category widths, this allocation should be the same for the entire report or
presentation.</p>
<h3 id="un-34-unify-structure-dimensions-use-vertical-direction">UN 3.4 Unify structure dimensions, use vertical
direction</h3>
<figure><img src="img/un-3.4.png" alt="Figure UN 3.4: Unify structure dimensions, use verticaldirection">
<figcaption>Figure UN 3.4: Unify structure dimensions, use verticaldirection</figcaption>
</figure>
<p><em>Structure dimensions</em> are all dimensions that are not measures, scenarios, or time periods. In many
cases, the following structure dimensions are used: regions, organization units, products, customers, channels,
and accounts.</p>
<p>Display structures always in vertical direction. Use custom symbols if it is helpful to differentiate structure
dimensions, see Figure UN 3.4.</p>
<h3 id="un-4-unify-analyses">UN 4 Unify analyses</h3>
<p><em>Analyses</em> are carried out in order to understand certain business situations, e.g. finding the greatest
variances from a plan or calculating the monthly average.</p>
<p>This section comprises analyses regarding different dimensions such as scenario analyses, time series analyses,
and structure analyses. A section covering different adjustment analyses is added.</p>
<h3 id="un-41-unify-scenario-analyses">UN 4.1 Unify scenario analyses</h3>
<figure><img src="img/un-4.1.png" alt="Figure UN 4.1: Unify scenario analyses">
<figcaption>Figure UN 4.1: Unify scenario analyses</figcaption>
</figure>
<p><em>Analyze scenarios</em> by comparing them and by calculating their absolute and relative variances. Notation
standards for scenario analyses cover the labelling of variances and the semantic design of chart elements such
as columns, bars, and axes, see Figure UN 4.1.</p>
<p><strong>Scenario comparisons</strong></p>
<figure><img src="img/un-4.1-1.png" alt="Figure UN 4.1-1: Scenario comparisons">
<figcaption>Figure UN 4.1-1: Scenario comparisons</figcaption>
</figure>
<p><em>Scenario comparisons</em> place the data of different scenarios next to each other, for example actual data
next to previous year or budget data. This is relevant for both charts and tables. In tables, scenarios usually
are shown in columns.</p>
<p>Scenarios can be compared in an absolute or relative way:</p>
<p>Absolute variance = primary scenario – reference scenario</p>
<p>Relative variance = absolute variance / reference scenario</p>
<figure><img src="img/un-4.1-2.png" alt="Figure UN 4.1-2: Column charts with scenario comparisons">
<figcaption>Figure UN 4.1-2: Column charts with scenario comparisons</figcaption>
</figure>
<p>Arrange scenarios of <em>different time periods</em> (mainly years) in temporal ascending order either from left
to right (horizontal axes) or from above to below (vertical axes), e.g. PY (= AC 2014), FC 2015, PL 2016.</p>
<p>No rule governs the sequence of scenarios referring to the <em>identical time period</em> – e.g. PL 2015, FC
2015,
AC 2015, but the selected sequence should be kept the same in all charts and tables.</p>
<p><em>Scenario comparisons</em> are visualized either by grouping columns or bars (e.g. overlapping columns of PY
and AC or overlapping bars of PL and AC), or with <em>scenario triangles</em> using the respective area coding
(e.g. solid light color for PY) to represent the reference scenario. <em>Scenario triangles</em> can also be
added to overlapped bars or columns in order to show a third scenario.</p>
<p>The scenarios AC and FC stand in the foreground of other scenarios in grouped columns or bars.</p>
<p><strong>Absolute variances</strong></p>
<p>An <em>absolute variance</em> is the difference between two values of one measure from different scenarios.</p>
<p>The sign “Δ” represents the absolute variance as a prefix to the subtrahend of the respective difference, i.e.
“ΔPL” for the absolute difference “AC minus PL” (AC-PL) or – if FC is compared to PL – “FC minus PL” (FC-PL).
</p>
<p>The most common <em>absolute variances</em> are the following:</p>
<ul>
<li>
<p><strong>Plan variance:</strong> “ΔPL” for AC-PL or FC-PL (when comparing FC to PL)</p>
</li>
<li>
<p><strong>Previous year variance:</strong> “ΔPY” for AC-PY or FC-PY (when comparing FC to PY)</p>
</li>
</ul>
<p>If it is not clear whether AC or FC is compared to plan in ΔPL or ΔPY, use the following notation:</p>
<ul>
<li><strong>Plan variance:</strong> “AC-PL” and “FC-PL”</li>
<li><strong>Previous year variance</strong>: “AC-PY” and “FC-PY”</li>
</ul>
<p><em>Positive absolute variances</em> (as well as positive percent variances) have a “+” to emphasis this aspect:
“+13” always means a <em>variance</em> of 13, “13” means any absolute value of 13.</p>
<p>If absolute variances are displayed in columns or bars (“variance columns” or “variance bars”), these variance
columns or bars have the same width and the same scale as the corresponding base value columns or bars.</p>
<figure><img src="img/un-4.1-3.png" alt="Figure UN 4.1-3: Colors for displaying variances">
<figcaption>Figure UN 4.1-3: Colors for displaying variances</figcaption>
</figure>
<p>Variance bars and columns representing a <em>positive impact</em> on business issues (mainly result) are colored
green, those representing a <em>negative impact</em> red, see figure on the left. Variance bars and columns
representing a <em>neutral impact</em> are colored medium gray. If no color is available, replace red with dark
gray, green with light gray. For readers with color deficiency, replace green with blue-green.</p>
<p>If it is helpful, numbers in tables representing variances are colored in the same way.</p>
<p><strong>Note</strong>: These colors for positive, negative, or neutral variances must not be confused with red
and green “traffic lights” (see also EX 2.5 “<a href="#ex-25-replace-traffic-lights">Replace traffic
lights</a>”).</p>
<figure><img src="img/un-4.1-4.png" alt="Figure UN 4.1-4: Bar charts with absolute variances">
<figcaption>Figure UN 4.1-4: Bar charts with absolute variances</figcaption>
</figure>
<p>In order to visualize the <em>scenario to be analyzed</em> (minuend), apply scenario notation to the fill of the
variance columns or bars, e.g. <em>solid</em> green or red fill for AC and <em>hatched</em> green or red fill
for
FC. If in special cases the minuend is PL (e.g. variance of plan versus average) the variance columns and bars
are <em>outlined</em> green or red.</p>
<p>Position data labels for variance columns and bars always <em>outside</em> of these visualization elements. These
labels’ position aligns with the direction of positive or negative increase, i.e. the label of a positive
variance (green) in a variance column is positioned above the column; the label of a negative variance (red) on
the left hand side outside of the bar.</p>
<p>In order to visualize the <em>reference scenario</em> (subtrahend) of an absolute variance (in general PY, PL, or
BU), apply scenario notation to the axis: For absolute variances to PY the axis is colored solid light, for
absolute variances to PL or BU the axis takes an outline shape (two parallel lines).</p>
<p>Treat variances of ratios, e.g. percent values (profit on sales) in a special way: Absolute variances of percent
values are called <em>percent points</em>, e.g. AC 50% – PL 40% = +10pp.</p>
<p><strong>Relative variances</strong></p>
<p>A <em>relative variance</em> is an absolute variance as a percentage of the subtrahend of the absolute variance.
</p>
<p>For the textual notation of relative variances, use the sign “Δ”as a prefix to the subtrahend and the sign “%” as
appendix, e.g. ΔPL% for the relative variance (AC-PL)/PL*100.</p>
<p>The most common <em>relative variances</em> are the following:</p>
<ul>
<li>
<p><strong>Plan variance</strong>: “ΔPL%” for (AC-PL)/PL*100 or (FC-PL)/PL*100 (when comparing FC to PL)</p>
</li>
<li>
<p><strong>Previous year variance</strong>: “ΔPY%” for (AC-PY)/PY*100 or (FC-PY)/PY*100 (when comparing FC
to PY)</p>
</li>
</ul>
<p>Display “n.a.” (not available) if the calculated relative variance cannot be interpreted, as is often the case
when a positive value is compared to a negative reference value (denominator):</p>
<pre><code>Profit AC = 30
Profit PL = -30
ΔPL = +60
ΔPL% = 60 /
-30 = -200% => n.a.</code></pre>
<p>Use the following notation, if it is not clear whether AC or FC is compared to Plan:</p>
<ul>
<li><strong>Plan variance</strong>: “(AC-PL)%” and “(FC-PL)%”</li>
<li><strong>Previous year variance</strong>: “(AC-PY)%” and “(FC-PY)%”</li>
</ul>
<p><em>Positive relative variances</em> (as well as positive absolute variances) have a “+”-to emphasize this
aspect: “+13%”always means a <em>variance</em> of 13%, “13%” means any kind of percentage such as ratio or a
share.</p>
<figure><img src="img/un-4.1-5.png" alt="Figure UN 4.1-5: Columns charts with relative variances">
<figcaption>Figure UN 4.1-5: Columns charts with relative variances</figcaption>
</figure>
<p>Relative variances are displayed in thin columns (vertical pins) or thin bars (horizontal pins), see the two
figures on the left.</p>
<p>Pins representing a <em>positive impact</em> on business issues (mainly result) are colored green, those
representing a <em>negative impact</em> red. Pins representing a <em>neutral impact</em> on business issues are
colored medium gray. If no color is available, replace red with dark gray, green with light gray. For readers
with color deficiency, replace green with blue-green.</p>
<figure><img src="img/un-4.1-6.png" alt="Figure UN 4.1-6: Bar charts with relative variances">
<figcaption>Figure UN 4.1-6: Bar charts with relative variances</figcaption>
</figure>
<p>The labels of pins and the numbers representing variances in tables can be colored in the same way.</p>
<p><strong>Note</strong>: These colors for positive, negative, or neutral variances must not be confused with red
and green “traffic lights” (see also EX 2.5 “<a href="#ex-25-replace-traffic-lights">Replace traffic
lights</a>”).</p>
<p>Position data labels of pins outside the pin in the direction of the positive or negative increase, e.g. position
the label of a horizontal pin depicting “sales growth in %” (green) on the right hand side of the pin, position
the label of a vertical pin depicting “cost decrease in %” (green) below the pin.</p>
<p>Add head markers to the pins to visualize the <em>scenario to be analyzed</em> (minuend). Apply the scenario
notation to the fill of the heads, e.g. solid dark fill for AC and hatched fill for FC.</p>
<p>Apply the scenario notation to the axis in order to visualize the <em>reference scenario</em> for a relative
variance (in general PY, PL, or BU): For relative variances to PY fill the axis solid light, for relative
variances to PL or BU the axis takes an outline shape (two parallel lines).</p>
<p>Treat relative variances of percent values the same way as relative variances of absolute values, e.g. (AC 50% –
PL 40%) / PL 40% * 100 = +25%.</p>
<h3 id="un-42-unify-time-series-analyses">UN 4.2 Unify time series analyses</h3>
<figure><img src="img/un-4.2.png" alt="Figure UN 4.2: Unify time series analyses">
<figcaption>Figure UN 4.2: Unify time series analyses</figcaption>
</figure>
<p>Notation for <em>time series analyses</em> covers year-to-date analyses, moving analyses, and temporal indexing,
see Figure UN 4.2.</p>
<p><strong>Year-to-date analyses</strong></p>
<figure><img src="img/un-4.2-1.png" alt="Figure UN 4.2-1: Year-to-date analyses">
<figcaption>Figure UN 4.2-1: Year-to-date analyses</figcaption>
</figure>
<p><em>Year-to-date analyses</em> (YTD) refer to the period from the beginning of the year to the present (<em>YTD
time span</em>). The beginning of the year is not necessarily January 1. Some companies have fiscal years
beginning at other dates.</p>
<p>Where helpful, visualize analyses showing YTD values by prefixing an underscore to the <em>time period name</em>,
e.g. “_Jun 2015”or “_Jun∅” respectively. Optionally, add the first period of the YTD time span, e.g.
“January_June 2015”. In charts, add the underscores as a prefix at the left hand side of the end of the columns
or at the upper side of the end of bars.</p>
<p>Year-to-date operations cover accumulation of values, calculation of averages, and picking of last date values.
</p>
<p><strong>YTD accumulation</strong></p>
<p>In this context, <em>accumulation</em> means totaling successive time period values from the beginning of a
calendar year or fiscal year to the present. In this stricter sense, accumulation applies only to flow measures,
such as sales or costs.</p>
<p>If it is helpful, visualize analyses showing <em>YTD accumulation</em> with the underscore prefix (without
additional notation) e.g. “_Jun 2015”.</p>
<p><strong>YTD average</strong></p>
<p>In this context, the <em>average</em> is calculated by dividing the <em>YTD accumulation</em> by the number of
periods in the <em>YTD time span</em>. YTD average applies to both <em>flow</em> and <em>stock measures</em>.
</p>
<p>If it is helpful, visualize analyses showing <em>YTD averages</em> with the underscore prefix and an appended
“∅”sign, e.g. “_Jun 2015∅”.</p>
<p><strong>Last date value</strong></p>
<p>A special YTD analyses for stock measures is picking the <em>value of the last date</em> in the <em>YTD time
span</em>.</p>
<p>If it is helpful, visualize analyses showing <em>last date values</em> with the underscore prefix and an appended
full-stop, e.g. “_Jun 2015.”.</p>
<p><strong>Year-to-go analyses</strong></p>
<p>By analogy to year-to-date analyses, <em>year-to-go analyses</em> (YTG) refer to the period from the presence
(now) to the end of the (fiscal) year.</p>
<p>Where helpful, visualize analyses showing YTG values by appending an underscore to the <em>time period name</em>,
e.g.“Jun-2015_”.</p>
<p><strong>Moving analyses</strong></p>
<figure><img src="img/un-4.2-2.png" alt="Figure UN 4.2-2: Moving analysis labeling in a column chart">
<figcaption>Figure UN 4.2-2: Moving analysis labeling in a column chart</figcaption>
</figure>
<p>In general, <em>moving analyses</em> refer to the period of the previous twelve months.</p>
<p>If it is helpful, visualize moving analyses by prefixing the <em>time period name</em> with a tilde, e.g. “~Jun
2015” or “~Jun∅” respectively. In charts, add the tilde as a prefix at the left hand side of the end of columns
or the upper side of the end of bars.</p>
<p>Similar to year-to-date operations, moving analyses cover accumulation of values (<em>moving annual total</em>
MAT), calculation of averages (<em>moving annual average</em> MAA), and picking of last date values.</p>
<p>The visualization concept for <em>accumulation of values</em>, <em>calculation of averages</em>, and <em>picking
of
last date values</em> is identical to the visualization concept of year-to-date analyses – the tilde simply
replaces the underscore.</p>
<p><strong>Temporal indexing</strong></p>
<figure><img src="img/un-4.2-3.png" alt="Figure UN 4.2-3: Visualizing temporal indexing">
<figcaption>Figure UN 4.2-3: Visualizing temporal indexing</figcaption>
</figure>
<p>Using <em>temporal indexing</em> (indexing a time series), all period values are depicted in relation to the
value
of a chosen reference period (1 or 100%).</p>
<p>To visualize temporal indexing, position a black arrowhead pointing right at the left of the index point. “100%”
or “100”is written left of the arrowhead. If helpful, add an assisting horizontal line.</p>
<h3 id="un-43-unify-structure-analyses">UN 4.3 Unify structure analyses</h3>
<p>Notation for <em>structure analyses</em> covers averaging, ranking, selecting, indexing, and normalizing.</p>
<p><strong>Structural average</strong></p>
<figure><img src="img/un-4.3-1.png" alt="Figure UN 4.3-1: Visualizing structural averaging">
<figcaption>Figure UN 4.3-1: Visualizing structural averaging</figcaption>
</figure>
<p>The term “<em>average</em>” usually refers to the mean of different values. The section time series analyses
described <em>temporal averages</em> (e.g. monthly average of a year). <em>Structural averages</em> (e.g.
average
sales of several subsidiaries) are covered here. Typical structural averages are average by product, average by
country, and average by customer.</p>
<p>Visualize analyses showing structural averages with a “Ø” sign either appended or as a prefix, e.g. “EuropeØ” or
“Ø464”. If needed, add an assisting line.</p>
<p><strong>Ranking</strong></p>
<p><em>Ranking</em> analyses refer to descending or ascending rankings of structure elements. Words can be ranked in
alphabetical order, numbers in numerical order.</p>
<p>If helpful, append an arrow sign to rankings, e.g. “country names↓” or “product sales↑”.</p>
<p><strong>Selecting</strong></p>
<p>The structure analysis <em>selecting</em> is related to the structure analysis <em>ranking</em>, used, in
general,
to determine either maximal (fastest, most expensive) elements or the minimal (slowest, cheapest) elements. Top
ten, last ten, first quartile, last percentile, etc., are common forms of selecting.</p>
<p><strong>Structural indexing</strong></p>
<figure><img src="img/un-4.3-2.png" alt="Figure UN 4.3-2: Visualizing structural indexing">
<figcaption>Figure UN 4.3-2: Visualizing structural indexing</figcaption>
</figure>
<p><em>Structural indexing</em> depicts all element values in relation to the value of a chosen reference element
(=1 or 100%). Typical reference elements are the mean, the maximum, or a specific element in a given structure,
e.g. “Germany = 100%”.</p>
<p>To visualize <em>structural indexing</em>, position a black arrowhead close to the index point. “100%” or “100”,
is written next to the arrowhead. If helpful, add an assisting line.</p>
<p><strong>Structural normalizing</strong></p>
<figure><img src="img/un-4.3-3.png" alt="Figure UN 4.3-3: Visualizing structural normalizing">
<figcaption>Figure UN 4.3-3: Visualizing structural normalizing</figcaption>
</figure>
<p><em>Structural normalizing</em> refers to the comparison of several shares of some whole, e.g. shares of export
to different countries. Indexing and normalizing are similar analyses, indexing refers to one element (e.g. a
selected country), normalizing to the whole of several parts (e.g. country sales in % of Europe sales).</p>
<p>To visualize <em>structural normalizing</em>, add an assisting line representing 100%. Position a black arrowhead
at one end of the assisting line. “100%” or “100”, is written next to the arrowhead.</p>
<h3 id="un-44-unify-adjustment-analyses">UN 4.4 Unify adjustment analyses</h3>
<p><em>Adjustment analyses</em> can offer insight into root causes as they adjust values by neutralizing special
effects. In general, <em>adjustment analyses</em> are used in conjunction with scenario analyses. Here the
values
of one scenario are recalculated with correction factors from another scenario: e.g., adjust AC sales for
currency effects by re-measuring them with the PY exchange rates.</p>
<p>Typical <em>adjustment analyses</em> deal with currency, inflation, and seasonal effects.</p>
<h3 id="un-5-unify-indicators">UN 5 Unify indicators</h3>
<p><em>Indicators</em> in reports and presentations serve different purposes, e.g. highlighting and scaling. Using
the indicator with the same design for the same purpose will help to identify the situation much faster.</p>
<h3 id="un-51-unify-highlighting-indicators">UN 5.1 Unify highlighting indicators</h3>
<figure><img src="img/un-5.1.png" alt="Figure UN 5.1: Unify highlighting indicators">
<figcaption>Figure UN 5.1: Unify highlighting indicators</figcaption>
</figure>
<p>The message to be conveyed should be highlighted on the respective page by appropriate visual
means. <em>Highlighting elements</em> enhance the meaning and importance of other elements. Use highlighting
elements for <em>assisting</em> purposes, for visualizing <em>differences</em> and <em>trends</em>, for
underlining <em>values</em>, for indicating a <em>reference</em>, or for linking <em>comments,</em> see Figure
UN
5.1.</p>
<p><strong>Assisting lines and areas</strong></p>
<figure><img src="img/un-5.1-1.png" alt="Figure UN 5.1-1: Assisting lines">
<figcaption>Figure UN 5.1-1: Assisting lines</figcaption>
</figure>
<p>Use <em>assisting lines</em> for different highlighting purposes, e.g. for showing differences, for separating,
arranging, or grouping data in charts or tables, or for coordinating visualization elements of different charts,
see figure on the left.</p>
<p>Use <em>assisting areas</em> for different highlighting purposes, e.g. for highlighting words in a longer text,
or
for highlighting certain parts of charts or tables.</p>
<p><strong>Difference markers</strong></p>
<figure><img src="img/un-5.1-2.png" alt="Figure UN 5.1-2: Difference markers">
<figcaption>Figure UN 5.1-2: Difference markers</figcaption>
</figure>
<p>Highlight differences in charts by using two parallel assisting lines to project the respective lengths of two
columns or bars to a <em>difference marker</em> highlighting the distance between the two assisting lines.</p>
<p>Position difference markers in a way that they can clearly highlight the respective difference.</p>
<p>Difference markers representing a positive impact on business issues (e.g. profit) are colored green; difference
markers representing a negative impact on business issues (e.g. loss) are colored red. Difference markers
representing neutral impacts on business issues are colored gray.</p>
<p><strong>Trend arrows</strong></p>
<figure><img src="img/un-5.1-3.png" alt="Figure UN 5.1-3: Trend arrow">
<figcaption>Figure UN 5.1-3: Trend arrow</figcaption>
</figure>
<p>Arrows can highlight trends in charts and (seldom) tables, too.</p>
<p><em>Position trend arrows</em> in a way that they can clearly highlight the direction of the trend with the
respective slope. The arrow starts at the first period and ends at the last period included in the calculation
of the respective trend. The arrowhead is pointing in time direction. Adding a number and a designation for the
calculation method (e.g. CAGR: 10.8%) will give additional insight.</p>
<p>Trend arrows representing a positive trend are colored green; trend arrows representing a negative impact on
business issues (e.g. loss) are colored red. Trend arrows representing neutral impacts on business issues are
colored gray.</p>
<p><strong>Highlighting ellipses</strong></p>
<figure><img src="img/un-5.1-4.png" alt="Figure UN 5.1-4: Highlighting ellipse">
<figcaption>Figure UN 5.1-4: Highlighting ellipse</figcaption>
</figure>
<p><em>Use highlighting ellipses</em> to highlight single values. Good reasons for highlighting single values are
e.g.</p>
<ul>
<li>
<p><strong>Highlighting messages</strong>: If the message refers to a specific value in a chart, table or
graph, highlight this value with a blue ellipse.</p>
</li>
<li>
<p><strong>Highlighting additional values</strong>: Sometimes it is helpful to add additional values (e.g.
percent value) in charts or tables. In this case, use a black ellipse.</p>
</li>
</ul>
<p><strong>Reference arrowheads</strong></p>
<figure><img src="img/un-5.1-5.png" alt="Figure UN 5.1-5: Reference arrowhead">
<figcaption>Figure UN 5.1-5: Reference arrowhead</figcaption>
</figure>
<p><em>Use reference arrowheads</em> for highlighting a reference standard. Examples of reference standards are:</p>
<ul>
<li>
<p><strong>Indices</strong>: Either one value (e.g. the value of the year 2010) is set to 100%, or the total
is set to 100% (see sections about time series analyses and Structure analyses).</p>
</li>
<li>
<p><strong>Benchmarks</strong>: Popular benchmarks are market averages, competitors, or best practices.</p>
</li>
</ul>
<p>Position the arrowhead close to the point representing the index or the benchmark. Write the label for the index
(e.g. “100%” or “100”) or for the benchmark (e.g.“Market avg.”) next to the arrowhead. The arrowhead points in
the direction of an imaginary index or benchmark line. If helpful, add an assisting line.</p>
<p><strong>Comment references</strong></p>
<figure><img src="img/un-5.1-6.png" alt="Figure UN 5.1-6: Comment reference">
<figcaption>Figure UN 5.1-6: Comment reference</figcaption>
</figure>
<p>Use <em>comment references</em> in pairs to link comments to the corresponding values or positions in a chart or
a
table.</p>
<p><strong>Variance highlighting indicators</strong></p>
<p>Highlight variances in tables by using visualization elements representing the magnitude of the variance, such as
bars and pins (see also EX 2.5 “<a href="#ex-2%E2%80%945-replace%E2%80%94traffic-lights">Replace traffic
lights</a>”). “Traffic lights” might be useful for highlighting <em>single variances</em> related to the
message or to comments in tables without chart elements, though. Another means for highlighting <em>single
variances</em> are “warning dots” positioned next to the value or text element needing attention.</p>
<p>Indicators highlighting variances representing a <em>positive impact</em> on business issues are colored light
green, those representing a <em>negative impact</em> light red. If no color is available, replace red with dark
gray, green with light gray. For readers with color deficiency, replace green with blue-green.</p>
<p>Use only few variance highlighting indicators per page.</p>
<p><strong>Other highlighting</strong></p>
<p>Add visualization elements for not-valid values, limits, or other relevant phenomena. Standardize and document
these “signals” so that they become an effective means of communication.</p>
<h3 id="un-52-unify-scaling-indicators">UN 5.2 Unify scaling indicators</h3>
<figure><img src="img/un-5.2.png" alt="Figure UN 5.2: Unify scaling indicators">
<figcaption>Figure UN 5.2: Unify scaling indicators</figcaption>
</figure>
<p>Proper <em>scaling</em> is very important for the creation of meaningful charts. Several semantic <em>scaling
indicators</em> exist to deal with in challenging scaling problems. Use <em>scaling lines</em> and
<em>scaling
areas</em> (<em>scaling bars</em>) if necessary, see Figure UN 5.2.
</p>
<p><strong>Scaling lines</strong></p>
<figure><img src="img/un-5.2-1.png" alt="Figure UN 5.2-1: Scaling line">
<figcaption>Figure UN 5.2-1: Scaling line</figcaption>
</figure>
<p>Use scaling lines when comparing multiple charts (with the same unit) having different scales. Position a scaling
line parallel to the category axis at the same numerical height in all charts. If one chart among several other
charts uses a different scale, this fact can easily be identified (in general, the differing scale uses a
multiplier of ten).</p>
<p><strong>Scaling areas</strong></p>
<figure><img src="img/un-5.2-2.png" alt="Figure UN 5.2-2: Scaling area">
<figcaption>Figure UN 5.2-2: Scaling area</figcaption>
</figure>
<p>If helpful, fill the areas between the scaling lines and the category axes with light color. Use different colors
for scaling lines and scaling areas used in order to represent different scales.</p>
<h3 id="un-53-unify-outlier-indicators">UN 5.3 Unify outlier indicators</h3>
<figure><img src="img/un-5.3.png" alt="Figure UN 5.3: Unify outlier indicators">
<figcaption>Figure UN 5.3: Unify outlier indicators</figcaption>
</figure>
<p>Sometimes values (mostly relative variances) can be very big in comparison to other values. If such
an <em>outlier</em> is not important for business, e.g. a big relative variance of a small value, do not scale
the whole chart to this outlier rather visualize unimportant outliers with <em>outlier indicators</em>.</p>
<p>Omit the pin head and add <em>outlier triangles</em> pointing in the direction of growth, see Figure UN 5.3.</p>
<h2 id="thats-all-folks">That's all, folks</h2>
<p>Thanks for reading! Follow @<a href="https://twitter.com/ohmypy">ohmypy</a> on Twitter to keep up with new stuff
🚀</p>
</body>
</html>