For maximum impact follow these simple principles and guidelines when creating DASHBOARDS.

The supporting excel file provides examples of how leading MVP’s use these principles on a daily
basis.

Data distortion / misrepresentation

D A chart may be accurate but if the underpinning data is out of context it’s pointless.
Smoothed lines are great for identifying trends but can be misinterpreted as actuals.

D Aesthetics
A Use easy to read fonts, sizes and formats for text inputs. Ensure text is proper (and not
in upper or lower case). Keep axis label alignment horizontal as it’s easier to read.

Simplicity
S Keep charts as simple as possible. Inclusion of unnecessary information pollutes key
messages and adds clutter. Ensure the chart is the focus so the key messages are clear.

Harmony
H If you have multiple charts in your dashboard ensure they have a similar look and feel.
Avoid mixing formal & informal charts. Avoid overpowering colours / themes styles.

Best chart selection

B Chart selection depends upon the type of data you need to visualise and the key message(s)
you want to get across. eg tracking trends & relationship : Pie Column Chart Variations

Overpowering 3D effects
O They may look funky but they should be avoided. They distract from the simplicity and they
distort data visually. As a result they are prone to confusion and incorrect interpretation.

Audience
A Who is your audience and what key messages are important to them? This determines
which style you use (formal or informal) and the level of detail you contain within it.

Removal of non-essential items

R The more clutter you add to a chart (gradient backgrounds, full bodied axis lines, labels
etc) the harder the chart will be to read. They also have severe printing limitations.

Dynamic / interactive charts

D Great for providing your audience with the ability to control the content within a chart.
To prevent confusion ensure controls are keep simple and intuitive to control .

Sensible scale and axis selection

S Your scale / axis selection can have make a significant difference to the visual effect of a
chart. Incorrect selection or categorisation will distort trends and misrepresent the
data.
ASHBOARDS.
es on a daily

pointless.
as actuals.

per (and not

to read.

ollutes key
ges are clear.

ook and feel.

mes styles.

he key message(s)
hart Variations

mplicity and they

interpretation.

s determines
n within it.

is lines, labels
mitations.

within a chart.
trol .

ual effect of a
esent the
The majority of the example charts used in this guide were sourced directly from, or variants of John Walkenbac
Many thanks to John for giving permission for their inclusion.

All other sources are referenced accordingly.

r variants of John Walkenbach's book 'Excel 2007 Charts'.
 1
Month
Apr
Calls Per Hour
30
Target
80
Calls Per Hour
120
May 40 80
Jun 45 80 100
Jul 32 80 80
Aug 40 80
Sep 55 80 60
Oct 61 80 40
Nov 60 80
20
Dec 80 80
Jan 100 80 -
Feb 100 80 Dec Jan Fe
Mar 32 80 Calls Per Hour Target

Both charts are technically accurate, however, chart

Month Sales
3
Apr 0 Line Chart with Straight Line
May 8 10
Jun 0 8
Jul 0 6
Aug 9 4
Sep 0 2
Oct 0 0
Nov 7 -2
Dec 5 -4
Jan 0 -6
Feb 0 -8
Mar 5 -10
Apr May Jun Jul Aug Sep Oct Nov Dec Jan

Chart 4 has smoothed lines so distorts the appearan

5
Team Qtr 1 Qtr 2 % Change % Qtr 2 Contribution
60%

40%
Team A 1,500 1,555 3.7% 61.5% 60%
Team B 800 900 12.5% 35.6%
Team C 50 75 50.0% 3.0%
40%
Qtr 2 Total 2,530

20%

0%
Team A

Using percentage change can be misleading. Team C m

 2
Per Hour Calls Per Hour
120

100

80

60

40

20

-
Jan Feb Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar

Hour Target Calls Per Hour Target

accurate, however, chart '1' is completely out of context so should be avoided.

4
th Straight Line Line Chart with Smoothed Line
10
8
6
4
2
0
-2
-4
-6
-8
-10
p Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar

so distorts the appearance ie it looks like there are negative values

Team Sales
60%

40%
 Team Sales
60%

40%

20%

0%
Team A Team B Team C

be misleading. Team C may have the greatest qtrly change 50% change but they only contributed to 3% of sales
 6
Month Region 1 Region 2
Apr 1093 786 REGIONAL
May 1283 873
Jun 1101 659 1,800
Jul 1235 711
Aug 1457 739 1,600
Sep 1574 745
1,400
1,200
1,000
800
600
400
200
0
Apr

Source: Accoun

Appreciate this is an extreme example of w

7
When ever possible avoid altering the alignment of
your axis labels.

If you have long axis labels which can not be

shortened, try one of Andy Pope's solutions on the
following link :

http://www.andypope.info/charts/staggeredlabels.htm

If that doesn't work consider changing the chart type

eg make the column chart a bar chart.

These labels were quite long. Instead of ch

 R E G I O N A L P E R F O R M A N C E ( fi r s t h a l f )
1,800 Note: This text is too small to be
legible. But that's OK because
nobody will read it anyway. This
1,600 text is too small to be legible. But
that's OK because nobody will read it
anyway. This text is too small to be
legible. But that's OK because
1,400 nobody will read it anyway.

1,200
1,000
800
Region 1 Region 2
600
400
200
0
Apr May Jun Jul Aug Sep

Source: Accounting Department

his is an extreme example of what not to do, but it really emphasises the point about consistent formatting

s were quite long. Instead of changing the alignment of the text you can stagger the text
alf)

ion 1 Region 2
 8
Day Call Volume Daily Call Volumes
300
Mon 100
250
Tue 125 250
225
Wed 150 200
200
Thu 175 175

Fri 200 150

125
Sat 225
100
Sun 250 100

50

0
Mon Tue 150
Wed Thu Fri Sat Sun

Chart 8 is full of unnecessary colour and clutter which dis

On one of Chandoo's blogs Jon Peltier explains …

A good dashboard report is not defined by a fancy colour scheme. It is defined by the information it
when it appears, it really means something.
 9
Call Volumes
Daily Call Volumes
250 300
225
200 250
Call
175 Vol ume
200

150

100

Thu Fri Sat Sun

50
Mon Tue Wed Thu Fri Sat Sun
0

y colour and clutter which distracts from the key message.

defined by the information it conveys, its clarity, its comprehensiveness, its succinctness. It’s best to use colour sparingly, so
mes

Fri Sat Sun

olour sparingly, so
Chandoo, an Excel MVP recently arranged a dashboard creation competition. There are
some cracking examples from lots of different people that demonstrate great use of simple
colours and consistent themes. The attached images just one example, but you can see all
submissions on the link below.

http://chandoo.org/wp/2010/01/04/sales-dashboards/

There's lots of other fantastic material on Chandoo's site, however, the section on design
principles is particularly relevant to this section.

http://chandoo.org/wp/management-dashboards-excel/

Excel 2007 has some great theme colours pre built in the fill options.
The following ad-in from Andy Pope gives 2007 users extra pattern selection options.
http://www.andypope.info/charts/patternfills.htm
 1
0

esign
 11

Scotland Wales
Qtr 3
Qtr 4
Qtr 4 Qtr 3

Qtr 2 Qtr 1

Qtr 2

England Ireland
Qtr 4 Qtr 4
Q
Qtr 3
Qtr 1

Qtr 3
Qtr 2
Qtr 2

QTR Scotland Wales England Ireland

Qtr 1 1,082 1,093 1,099 1,076
Qtr 2 544 566 597 619
Qtr 3 183 217 241 450
Qtr 4 87 308 387 698

There are lots of great reference sites that give examples on chart selectio
http://chandoo.org/wp/2009/12/18/charts-to-compare-targets/
 12

Wales Performance by Country

1,200
Qtr 4 1,000
800
Qtr 1
600
400
r2 200
-
Qtr 1 Qtr 2 Qtr 3 Qtr 4
Scotland Wales England Ireland

Ireland
Performance by Country
3,000
Qtr 4
Qtr 1 2,500 Qtr
2,000 4
Qtr
1,500 3
Qtr
Qtr 2 1,000 2
500
-
Scotland Wales England Ireland

If you are trying to compare performance the pie charts in '11'

A far better option would be one of the column charts in '12'

es on chart selection, for example, Chandoo's top 10 charting ideas to compare actual values
e-targets/
 y Performance by Country
1,200
1,000
Qtr 1
800 Qtr 2
600 Qtr 3
Qtr 4
400
200
3 Qtr 4
-
Ireland Scotland Wales England Ireland

ry

Qtr
4
Qtr
3
Qtr
2

land

e the pie charts in '11' are completely ineffective.

column charts in '12'

mpare actual values with targets …

 13

Country % Sales
Scotland 22%
Wales 28%
England 28%
Ireland 22%

Scotland; 22%

The slices in this pie chart make it look li

14
Qtr Scot Wal Eng Ire
Qtr 1 34% 23% 24% 10%
Qtr 2 27% 29% 24% 26%
Qtr 3 28% 23% 26% 31% 35%
Qtr 4 11% 26% 27% 33%
30%
25%
20%
15%
10%
5%
0%
Qtr 1
Qtr

This chart is hard to make accurate com

15

35%
30%
 Qtr Scot Wal Eng Ire
Qtr 1 34% 23% 24% 10%
35%
Qtr 2 27% 29% 24% 26%
30%
Qtr 3 28% 23% 26% 31%
25%
Qtr 4 11% 26% 27% 33%
20%
15%
10%
5%
0%
Qtr 1

Adding perspective tends to obscure dat

 Sales Per Country
Wales; 28%

England; 28%
Scotland; 22%

Ireland; 22%

n this pie chart make it look like Wales sold more than England yet their achievement is the same

5%
0%
5%
0%
5%
0%
Wal
5%
Ire
0%
Qtr 1 Eng
Qtr 2
Qtr 3 Scot
Qtr 4

s hard to make accurate compassions

35%
30%
 35%
30%
25%
20%
15%
Ire
10%
5% Eng
0%
Wal
Qtr 1
Qtr 2 Scot
Qtr 3
Qtr 4

spective tends to obscure data points

 16
100
Qtr 4 Sales Take a D
90

Qtr Sales 80
Qtr 1 95
Qtr 2 90 70
Qtr 3 85 60
Qtr 4 10
50

40

30

20

10

0
Qtr 1 Qtr 2 Qt

Make sure the key message is right for your audience !

17
Qtr Sales Customer Complaints Dec
Apr 2,483 3,000
May 2,046
Jun 1,726
Jul 1,606 2,500
Aug 1,528
Sep 1,520 2,000
Oct 1,459
Nov 1,371
1,500
Dec 1,324
Jan 1,203
Feb 1,071 1,000
Mar 941

500

0
Apr May Jun Jul Aug Sep Oct Nov
Great message and use of clip art but you wouldn't want it in the m
Qtr 4 Sales Take a Dive

Qtr 2 Qtr 3 Qtr 4

your audience !

er Complaints Decline

Aug Sep Oct Nov Dec Jan Feb Mar

ou wouldn't want it in the middle of a formal dashboard
 18
Median Home Prices 1989 - 2006San Diego vs. Phoen
Adjusted for inflation
$700,000

$600,000

$500,000

$400,000

$300,000

$200,000

$100,000

$0
1989 1994 1999 2004

Phoenix San Diego

The John Walkenbach examples above demonstrate the benefits of

principle of "data Ink" maximization. Edward R Tufte has written bo
 19
n Diego vs. Phoenix Median Home Prices

2004

1989
Adjusted for inflation 1994 1999

ate the benefits of removing non essential elements. Chart '19' solves printing limitation is
ufte has written books on this principle and Jon Peltier templates are other great examples
 $700 K

San $600 K
Diego

$500 K

$400 K

$300 K

Phoenix $200 K

$100 K

$0 K
1999 2004

rinting limitation issues and is in line with the

her great examples of this line of thinking.
 There are countless sites and books that provide brilliant tutorials on how to make your charts
and dashboards dynamic.

The links below are just a few examples of my favourite, and most used reference sites and
resources that relate to dynamic functionality.

PHD's training section on creating dynamic charts

http://chandoo.org/wp/tag/dynamic-charts/

Contextures has lots of really usefuly tutorials on dynamic ranges

http://contextures.com/tiptech.html

Andy Pope's tutorials eg using spinners and dynamic date ranges

http://www.andypope.info/charts.htm

Using dynamic chart source data

http://peltiertech.com/WordPress/dynamic-chart-source-data/

John Walkenbach has numerous reference books to stimulate and inspire your dynamic solutions.
This one is particularly brilliant for charting tips such as dynamic ranges etc

http://www.amazon.co.uk/Excel-2007-Charts-Spreadsheets-Bookshelf/dp/0470044004/ref=sr_1_

There are additonal on-line dynamic tips within John's speadsheet tips section
http://spreadsheetpage.com/index.php/site/tips
 20
on how to make your charts

used reference sites and

your dynamic solutions.

helf/dp/0470044004/ref=sr_1_1?ie=UTF8&s=books&qid=1265103720&sr=8-1
 2
Month Sales
1
Apr 11,291
May 11,688
Sales Remained Flat Throughout the
Jun 12,096 14,000
Jul 12,021
12,000
Aug 12,132
Sep 12,470 10,000
Oct 12,893
8,000
Nov 12,780
Dec 12,954 6,000
Jan 12,584 4,000
Feb 12,687
2,000
Mar 12,973
-
Apr May Jun Jul Aug Sep Oct Nov Dec Jan F

If you don't use a sensible value axis on your chart key me

23
Net Income: 37 Years of Gr
Year Net Income 8,000,000
1970 56,890 7,000,000
1980 1,837,130 6,000,000
1990 3,870,981
5,000,000
2000 5,197,329
2005 6,697,992 4,000,000
2006 7,008,873 3,000,000
2007 7,298,326 2,000,000
1,000,000
0
1970 1980 1990 2000 2005 2006 20

Chart '23' paints a misleading picture as standard categor

25

3,000

2,500
 3,000
Month Qty
Apr 95 2,500
May 293
Jun 1,032 2,000
Jul 1,293
Aug 1,456
1,500
Sep 1,183
Oct 1,392
Nov 1,582 1,000
Dec 2,083
Jan 1,893 500
Feb 2,594
Mar 2,899
-
Apr May Jun Jul Aug Sep Oct N

3,000
2,500
2,000
1,500
1,000
500
-
Apr May Jun Jul Aug Sep Oct N

Remarkably these graphs are from the same data set yet
Advisor Score
1 21
2 28 Bin
3 27 <=20
4 19 20-40
5 50 40-60
6 11 60-80
7 50 80-100
8 80
9 37
10 46
11 40 Bin
12 54 <=10
13 21 10-30
14 90 30-40
15 25 40-50
16 96 50-60
17 17 60-100
18 31
19 22
20 53

Advisor Score
Apr 785 28
May 832
Jun 974
JRT Travel Se
Jul 850 Results of June P
Aug 759 5,000
Sep 4,890 4,900
Oct 689
4,800
Nov 725
Dec 840 4,700
Jan 785 4,600
Feb 881 4,500
Mar 754 1,000
900
800
700
600
500
400
300
200
 900
800
700
600
500
400
300
200
100
0
Apr May Jun Jul Aug Sep

29

These examples demonstrate methods of visualising a bro

 22
lat Throughout the Year Sales are on the rise!
13,500

13,000

12,500

12,000

11,500

11,000
g Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar

ue axis on your chart key messages are likely to be misleading or under/over stated

24
me: 37 Years of Growth Net Income: 37 Years of Growth
8,000,000
7,000,000
6,000,000
5,000,000
4,000,000
3,000,000
2,000,000
1,000,000
0
2000 2005 2006 2007 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010

picture as standard category labels are used. Chart '24' is better as the axis type is set to the dates

3,000

2,500
 3,000

2,500

2,000

1,500

1,000
Jul Aug Sep Oct Nov Dec Jan Feb Mar

500

-
t
g
n

b
c
r

Oc
Ap

De
Au

Fe
Ju

Jul Aug Sep Oct Nov Dec Jan Feb Mar

rom the same data set yet the initial interpretation would be different for each chart.
 26
Score Distribution
Bin Frequency 8
20 3
6
40 3
60 7 4
80 4 2
100 4 0
<=20 20-40 40-60 60-80 80-100

27
Bin Frequency
10 0
Score Distribution
30 4 10
40 1 8
50 5 6
60 2 4
100 8 2
0
<=10 10-30 30-40 40-50 50-60 60-100

Unequal bin sizes in a histogram can misrepresent the data

JRT Travel Services

Results of June Promotion
In John Walkenbach's book 'Excel
2007 Charts' there are a couple of
methods. This one uses two
overlying charts.
 Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar

Andy Pope also demonstrates

another method that incorporates
an additional series and data label
to recreate the effect.

http://www.andypope.info/charts
/brokencolumn.htm

methods of visualising a broken axis. Very handy is your data has large ranges.