Scribd
Upload
142 views

TNV Data Visualization Poster

When designing a dashboard, the principles of using effective data visualization and focusing on what is most important still apply. An efficient dashboard allows users to focus on numbers, relationships, trends and anomalies without distractions. Colour should be used deliberately to clarify information, and visualizations like gauges, pies and graphs without a zero baseline can obscure meaningful comparisons and insights. Highlighting differences from targets is an effective way to show performance at a glance.

Uploaded by

Rao Vedula
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
142 views

TNV Data Visualization Poster

When designing a dashboard, the principles of using effective data visualization and focusing on what is most important still apply. An efficient dashboard allows users to focus on numbers, relationships, trends and anomalies without distractions. Colour should be used deliberately to clarify information, and visualizations like gauges, pies and graphs without a zero baseline can obscure meaningful comparisons and insights. Highlighting differences from targets is an effective way to show performance at a glance.

Uploaded by

Rao Vedula
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

DATA VISUALIZATION.

8 TIPS FOR AN EFFECTIVE


DASHBOARD

DASHBOARDS PROVIDE
INSIGHT INTO THE
PERFORMANCE OF
ORGANIZATIONS AND CAN
THUS IMPROVE THE
DECISION-MAKING
PROCESS.

A dashboard isnt there just to inform you about the


current situation but mainly to answer the question:
What improvements can be made? A successful
dashboard can help you to improve on management
and operating prot. But what makes a dashboard
successful? Correct and timely gures?
Unquestionable! But just having the right gures
isnt enough anymore.
A dashboard stands or falls by the degree in which it
gives you the required insights. A picture is often
worth more than a thousand words. But this also
applies to gures and data. By cleverly displaying
data visually, patterns, trends and exceptions
become visual at a glance.

Therefore effective data visualization is critical for


the success of a dashboard. But how do you make
the right decisions? What works and what doesnt?
With some of our best practices on effective data
visualization we want to help you with these
decisions.
We not only offer a unique experience in dashboard
design but we are also the experts in SAP BusinessObjects BI.
Want to know more? Contact the THE NEXT VIEW!.
[email protected]
www.thenextview.nl

1 2 3 4 5 6 7 8
LESS IS MORE

CHOOSE YOUR COLOURS WITH


CARE

NO MORE GAUGES

START AT ZERO

SHOW THE DIFFERENCE

NO MORE PIES

HIGHLIGHT WHATS
(REALLY) IMPORTANT

YOUR GRAPH FROM


ANOTHER ANGLE

The art of omission; highlight what is


important, leaving out what isnt.

Good use of colours can highlight


and clarify, while poor use will
obscure or conceal information.

Although the use of gauges and


speedometers in dashboards is quite
popular, there are better and more
effective visualizations available that
also take up less space.

With a bar chart, always allow the


vertical axis to start at zero (0) to
prevent graphs from being wrongly
interpreted.

If you want to compare two series


you can also show and highlight the
difference.

The pie chart. Colourful and popular


but not very effective.

Keep a dashboard neutral and


highlight what is important; like the
current position or a different value.

A horizontal bar (graph) is often the


best choice when long labels are
used or when you want to show the
hierarchy.

When designing a dashboard, the old adage, less is


more still applies. An efcient dashboard allows
users to focus on what is really important; numbers,
relationships, trends and anomalies. Decorations
often make a dashboard less readable and thus less
efcient
In the late 1970s, Edward R. Tufte, famous for his
books on data visualization, introduced the term
"data-ink". In data-ink you cannot delete anything
without inevitably changing the message.
Non-data-ink is the term hes using for all other, non
essential, information.
His message? Increase the ratio between data-ink
and non-data-ink; Highlight whats important, leave
out what isnt. This includes for instance unnecessary
3D effects, colour gradients, shadows and other
effects that are often purely decorative.

30

Or can it? In dashboard design aesthetic aspects


also play a role, but it is mainly about how you can
best deliver the message.
In a dashboard colour has multiple functions. You
can use colours to highlight what is important but
also to group what belongs together. Theres often an
unnecessary use of different colours while it does not
have a specic function. The colours add nothing
and provide a less clear graph.
.

USA

UK

Holland

Germany

France

20

By properly choosing the colour, charts will be much


easier to read.

10

JAN

FEB

MAR

80

APR

And one based on the 'less is more' principles:


Europe

20

USA

10

JAN

FEB

MAR

APR

Only the important parts of the graph are highlighted


and the less important ones, such as grid lines, have
a light shade of grey. We believe that the principles of
Tufte still have their value in helping us develop
efcient dashboards.

$1.4 mln

How many times bigger is the number of sold products compared to the ones sold by the competitor?

80

Expenses

One major shortcoming is that they take up a lot of


space while they communicate very little information.
In this case, only the actual value is being displayed
which you could also represent as a single number;
possibly supported by a signal color if the value
crosses a certain threshold.
Also there is nothing to compare it with. Is the current
value better or worse than the one from the previous
month? That is something we are not able to read
from this visualization. Furthermore it is hard to place
the value in perspective since there is no scale
available.

Often actual values are being compared with budget


values. The difference between these two values is
important.
Because they are both set up in time it allows you to
show the development of the individual values. The
differences you need to calculate yourself.

60

200
50

Competitor

Own

The number of own products seems to be approximately twice as big. Seems to be. After all, the length
of the bar denotes the value and the one for own
products is about twice as tall. This graph however
gives a distorted or even deceptive representation of
the truth since the vertical axis (y-axis) does not start
at zero.
If we draw the same graph again but this time with a
correct y-axis, the gap between the two bars
becomes a lot less dramatic.

100

Holland

Germany

France

Jan

Feb Mar Apr May Jun

Another approach is focusing on the difference itself:

50

Not everyone sees colours in the same way. Therefore we use special tools, to view our designs
through the eyes of someone who is (partly)
colour-blind. So we know for sure that the dashboard
is useable for everyone.
Whatever colours you choose, do it deliberately and
apply them consistently.

It becomes immediately clear that the expenses go


up and that it is currently outside the ideal
bandwidth.

-50

This type of graph also takes up less space making it


very well suited for dashboards in which large
numbers of values need to be depicted. .
In short, there are often better alternatives available
than the familiar gauges and speedometers.

Own

An incorrect y-axis can thus result in a


misrepresented picture and erroneous
interpretations. To prevent this from happening, the
y-axis of a bar chart should always start at zero.

Probably not. And what about now?

8545398542
7543798971
6198842715
2275192439

Bar (graphs) are most likely to be displayed vertically


but a horizontal orientation also has its advantages.
Labels sometimes are displayed vertically if they are
too long to be placed next to each other. This does
not improve the readability however.
A horizontal orientation has a practical advantage
that the category labels can also be placed
horizontally and therefore are easier to read. Because
you read such a graph top-down this form is very
suitable for displaying a hierarchy:
0

50

100

200

USA
UK
Germany
Japan
By sorting categories top-down the relationship is
immediately clear.
If you want to show a relationship in time always
choose an orientation from left to right. This is also
the best match with our natural sense of time.

10%

Jan

Feb Mar Apr May Jun

By highlighting negative values it is immediately clear


where results differ from expectations.

Competitor

8545398542
7543798971
6198842715
2275192439

Therefore we use a neutral dashboard and use


highlights to immediately show what is really
important.

20%

-25

20

40%

25

80

40

Without the exact values it is difcult to compare the


slices. Our ability to estimate corners and surfaces is
much less developed than for horizontal or vertical
lines.

30%

Expenses $1.4 mln

300

60

UK

400

In a dashboard you want the most important


information to immediately stand out. An important
visual tool here is to provide sufcient contrast.
Do you immediately see the number of 3s in the
rows?

Which slice is bigger, A or B?

500

70

The following alternative already gives some more


historical context since it displays the last 12 months
as well.

40

30

Often the dashboard of a car is used as a metaphor


when designing a dashboard to manage an
organization. Speedometers like the one below can
often be found in a dashboard:

The purpose of a graph is to communicate information fast and accurate by means of visualization.

However, they are not very effective.

80

40

Here is a typical line graph:

Europe

How to choose the right colours in a dashboard?


Use of colour is often about personal taste; and that
cannot be argued.

The pie chart, who hasnt used it? This graph form is
very popular not only in the media but also in reports
and dashboards. The pie chart has one major
advantage: it immediately shows a "share-of-thewhole" relationship. But is it effective?

This graph can be used both for absolute differences


as well as relative percentage differences. The last
one for example makes it possible for various KPIs to
be compared.

0%

Much clearer, right?


A pie chart looks nice but is one of the least effective
graph forms. There are specic situations where a pie
chart is effective but most likely there are better
alternatives to use.

www.thenextview.nl

You might also like