WHY DATA VISUALIZATION IS IMPORTANT
IN DELIVERING ACTIONABLE INSIGHT
Iain Curtain
Edgewater Ranzal
Agenda
• Introduction
• Design examples
• Understanding the customer
• What makes good design
• How to apply this to reporting
• Summary and examples
• Questions
About
Customer Sample
Design Examples
- Lots of images
- Attention not drawn
to specific items
- Too much color
- Visuals used when
charts or data could
convey messages
better
- Waste of valuable
screen real estate
Design Examples
- 3-D rendering
makes it harder to
interpret the values
- Forecasted Units
and Dollars lines
connect the regions,
when they are
distinct data points
- Dollars and Year
Ago Dollars are
stacked, when they
are also distinct
values
- Excessive use of
the word ‘Region’
Design Examples
- Recent example of
visuals over
substance and
meaning
- Very difficult to
determine
correlation between
circle sizes
Design Examples
- Dark images and
background distract
the viewer
- Cannot determine
trends with only 2
data points
- Confusing as 2010
values are not
circles
- Visuals used when
charts or data could
convey messages
better
Customer Maturity
Foundation
Operations
Mgmt Reports
Dashboards
Scorecards
Customers are on a
reporting journey and
determining what their
requirements and
future plans are
important in
understanding where
they are, what their
needs are and where
they think they are
going.
You must satisfy the
pre-requisites before
climbing up the
pyramid.
Customer Maturity
Area 1 Area 2 Area 3
Main
tain
cu
rrent
busin
ess
opera
tions
Are
as to g
row
Consider 3 different
part of the organization
at different stages of
maturity:
- Finance = Area 1
- Logistics = Area 2
- Manufacturing =
Area 3
User Requirements
• Talk to users
• Listen to users
• Ask why, why, why?
Good design concepts
• Memory limits
• Encoding data for rapid perception
• Gestalt principles of perception
Memory Limits
• Iconic memory – visual cues, pre-conscious
/pre-attentive processing
• Short term memory – conscious
processing, 3-9 chunks only
• Long term memory
Data Encoding
1723957695026398027384956012
9847536970898726547867925019
2005928976548102985079827158
0297456478597069873940588698
5726327189506972915069871256
2783789
- How many 3’s can
you find?
- As there is no
encoding of data,
we process
sequentially –
attentive processing
- very slow!
Data Encoding
1723957695026398027384956012
9847536970898726547867925019
2005928976548102985079827158
0297456478597069873940588698
5726327189506972915069871256
2783789
- 7 is the correct
answer
- Much easier to see
when data is in a
different color, the
same goes for
bolding, size, shape
and orientation
changes as well
Gestalt Principles
• Proximity
• Similarity
• Enclosure
Here we see:
- 2 groups rather than
7 blogs
- 2 different sets
within the groups
- 2 further groups
within the groups
Gestalt Principles
• Closure
• Continuity
• Connection
Our minds:
- Close
- Continue
- Link
even though these can
be seen as discrete
items.
Applying concepts
• Focus on the value add you are showing by organising
and minimising the data shown.
• Arrange information in a way that makes sense, making
sure that the important data stands out.
Edward R Tufte
• Tufte provided lots of thought around how we view and
perceive data
Data Ink Ratio
• Key concepts: reduce non data ink from graphics, focus
on the values
• Reduce graphic paraphernalia (chartjunk)
Chartjunk
• Which of these has clutter
and unnecessary items?
• Which is easier to see the
data?
“We can easily see patterns
presented in certain ways, but if they
are presented in other ways they
become invisible.”
Colin Ware
Information Visualization -
Perception for Design, 2000
Colin Ware
Colin Ware
• Color
• Form
We distinguish the
items if different in
terms of:
- Color – either hue or
intensity
- Form – can also be
size, shape,
orientation
Colin Ware
• Position
• Motion
We distinguish the
items if different in
terms of:
- Position
- Motion:
flashing/moving
should only be used
for real time data or
issues requiring
immediate attention
Stephen Few
• Combined previous theories and melded with current
designs and dashboard and communication ideas
Stephen Few
This simple example
shows how we can’t
easily compute size
variations in area.
The large circle is 16
time larger.
Pie and area charts
should not be used as
we cannot quickly
recognize the
differences.
How does this help us?
• Simplify – reduce the data presented
• Simplify – concentrate on important
information
• Simplify – remove unnecessary color and
distractions
Examples
Top example is very
difficult to read and
interpret numbers.
Bottom report is
cleaner and can easily
see the items without
distracting border and
shading.
Examples
Top example is very
distracting and difficult
to focus on areas that
need attention.
Bottom report is much
cleaner and easier to
see items needing
attention.
Design Examples
Much easier to
compare the different
market capitalizations
when presented as a
bar chart.
Also much easier to
see the best/worse if
ordered.
Design Examples
Design Examples
128.5
1,500.0
4.373.5
165.845.020.4 25.2 2.9
2005 2009 2010
0
200
400
600
800
1,000
1,200
1,400
1,600U
S$ M
illio
ns
UK
Germany
North America
128.5 73.5 20.4
1,500.0
165.825.24.3 45.0 2.9
UK Germany North America
0
200
400
600
800
1,000
1,200
1,400
1,600
US
$ M
illio
ns
2005
2009
2010
In charting the previous
graphic, it’s obvious
that there is no
relationship between
the years or countries.
Maybe a table would
have been better?
“Graphical excellence is that which
gives the viewer the greatest
number of ideas in the shortest time,
with the least ink in the smallest
space.”
Edward R Tufte
The Visual Display of
Quantitative Information, 1983
Iain Curtain
Consultant
http://thinsliced.wordpress.com/
202 683 0874
www.ranzal.com