Report

visualisation:
From concept
to deployment
A KPMG and CGMA Thought Leadership Report
Contents

03 Visualisation – a means of
enhancing business insight
06 Five principles for effective
report visualisation

10 Case study
A global industrial manufacturer

12 How to obtain value from data

16 Getting started and lessons

learned
18 Case study
A large retail insurer

Chartered Global Management Accountant

CGMA is the most widely held management accounting designation in the world. It distinguishes
more than 150,000 accounting and finance professionals who have advanced proficiency in finance,
operations, strategy and management. In the U.S., the vast majority are also CPAs. The CGMA
designation is underpinned by extensive global research to maintain the highest relevance with
employers and develop competencies most in demand. CGMA designation holders qualify through
rigorous education, exam and experience requirements. They must commit to lifelong education
and adhere to a stringent code of ethical conduct. Businesses, governments and nonprofits around
the world trust CGMA designation holders to guide critical decisions that drive strong performance.

Report visualisation: from concept to deployment

An introduction
to report
visualisation

Report visualisation, along with the advancement In today’s modern business environment, the
of multimedia technology over recent years, has volume of data is far larger than ever before.
become a very powerful means of sharing and Therefore the art of report visualisation has become
providing insight around business performance, increasingly important as we become overwhelmed
which can be easily digested by the recipient. by both data and information.
Report visualisation is the process of presenting
report formats that represent data and information In recent years technology providers have embraced
in a pictorial or graphical format that helps the market and developed new tools to assist in the
the recipient to understand the significance of production of high-impact visual reports. In parallel,
the content more easily than if presented in a the quantum leaps forward in desktop and mobile
traditional report format. technology have enabled visualisation to be truly
interactive and dynamic.
The human brain consumes information much
more easily when presented with graphical This paper shares some of the developments around
representations, and so report visualisation report visualisation and the key steps that should
concepts have become a key aspect of report be adopted to harness such capability. Included are
design. The beauty of visual reporting is that ‘Research Insights’ gathered from Chartered Global
patterns, trends and correlations are much Management Accountants (CGMA designation
easier to spot when presented graphically than holders) across a wide range of industries.
in a table of data.

Report visualisation: from concept to deployment 3

Visualisation - a means
of enhancing business
insight

What to measure - data that matters

A key role of the management accountant is to
support effective decision making by presenting
relevant, timely and accurate information in a
manner that enables informed dialogue and
decision making.

For effective decisions to be taken, the information

presented must focus on the right key performance
indicators (KPIs) that drive true business
performance.

Making decisions based on these drivers should

result in focusing performance around delivery
of business strategy.

#ResearchInsight All the interviewees agreed

that presenting analysis of non-financial
data is of much more value in improving
performance than historic financial results.
The role of visualisation in this process is to present
the relevant data in such a way that the message
is quickly absorbed and understood by the
reader so that action can be taken quickly, and
time is spent on decision making rather than
understanding performance.

Operating model
The demand for relevant reporting must also be
satisfied by a robust delivery model that ensures
a high level of reporting quality which can be
absorbed quickly.

4 Report visualisation: from concept to deployment

Reporting quality focuses on two key areas: When deciding on the appropriate way to present
the data, it is often best to start with an idea of what
• Data and master data that are subject to robust the end goal is i.e. what message are you trying to
control which ensures reporting integrity present (which may be dynamic).
• Standards in reporting that drive commonality Once you have defined the purpose of the report,
across the whole organisation and aid then consider the metrics and the dimensionality
assimilation to the consumer. For example, that is required to provide insight.
if ‘red’ is consistently used to display an adverse
variance, then the reader is not having to ask
the question every time. #ResearchInsight Although most business
users may find charts more user-friendly,
Whether reporting is produced at a local business tables of figures are still very useful to
unit level, shared service centre or outsourced,
accountants as reference or supporting data.
these two key areas are fundamental to quality
reporting. Before embarking on creating the report, consider
who the reader is as this will help determine the
How to present data to make a conclusion delivery channel you select. It might be relevant
and take action to create the same report with multiple reporting
output styles to ensure the message is delivered
The overriding principle in deciding how to present
with maximum effectiveness.
data is that one size does not fit all.
Report visualisation will remain a core skill for the
Our research highlighted that there is still a place
management accountant to develop and evolve.
for traditional row and column table-based
The tools available are ever changing and improving
reporting, but this is used mainly for reference data.
as technology continues to develop at a rapid pace.
The use of graphical reporting is much more
commonly used to tell a story but there are also
readers within the organisation that are ‘turned off’
by the use of graphs. This makes it important
to use appropriate media depending on the
target audience.

Report visualisation: from concept to deployment 5

 The five principles
for effective report
visualisation
The following five key principles should always be To investigate the sales variance you may want
considered. to drill down by:

• country
1 Ensure data is optimised
for report visualisation
• product / service
• distribution channel
Once you have identified the right measures that are
to be reported, you will need to identify the relevant • region.
data sources, which might be more than one for This is facilitated by creating a dimensional structure
each report. The key points to consider per data that stores your data against reporting hierarchies.
source are:

• Where is the data source?

2 Apply the relevant
visualisation tool
• Can I physically extract the data?
The principal factor in choosing the relevant report
• Can the data be refreshed in time to make visualisation technique will undoubtedly be driven by
the report relevant? what the user for the report feels comfortable with.

• How can I ensure sustainable integrity over However, there are a few common guiding
the available data? principles that can be used to help match the
visualisation to the requirement:
It is at this point that you should consider whether it
is appropriate to create a data storage layer. Aside • Variance analysis is often depicted by using
from providing a robust data extraction process and waterfall charts, commonly known as bridges.
a single repository of data from multiple sources, The key benefit delivered is that the size of
this can provide a key role in optimising the data for the steps in the bridge are to scale, so your
report visualisation. eye naturally focuses on the significant
movements first
The first benefit is that within this data storage
layer, we can clean the data and provide a common • In providing an overview of a business area,
definition to measures. This is especially important dashboards can provide a relevant summary.
if there are multiple data sources which have all The key to delivering a useful dashboard is
been developed independently without regard for identifying what are the four or five relevant key
master data alignment. drivers that can be reported on one page or
screen. Dashboards provide a summary and often
The second area to consider is the creation lead to a drill down to the next level, which is likely
of structure or hierarchies within the data that to be a different report.
facilitates drill down to the reader. This is especially
important if using tools such as mobile technology • Trend analysis or time-based results are generally
where the user is encouraged to drill to the next shown using line charts. Modern planning
level with the touch of a finger. solutions can also use line charts as the visual
interface to plan for future periods by establishing
For example, imagine a report on an iPad showing a baseline and then allowing the planner to drag
a map of the world featuring a big red variance for the chart, which then models the underlying data
sales in Europe.

6 Report visualisation: from concept to deployment

• Where organisations are built up of business • Colours – the first area to think through with
areas or regions, mapping charts are useful, colours is whether the colours you have used
especially when presented on an interactive make the report attractive to the eye, which
interface such as a tablet engages the reader.

• When comparing data sets that are built up from The second use of colours is to assist in telling
component parts, then a bar chart or side by the story. ‘Red’ and ‘green’ are used universally
side pie chart will bring the message to life to depict ‘positive’ and ‘negative’, so consider
conditional formatting if appropriate. Colours
• Consider the use of summary but insightful
in related charts should always be consistent -
supporting commentary
for example, if a product or channel is shown in
• Finally if you are producing reference material blue in one chart, it should be shown in blue in
only, do not forget traditional row and column the next chart too.
tables. These are still extremely useful, especially
• Finally, although colours enliven a chart, it is not
when combined with conditional formatting.
a good idea to use too many as the chart will
look messy.

3 Choose an appropriate
report layout
• Scaling – make scaling appropriate to the data
and be wary of hiding the story through scaling.
As visualisation skills become more and more If your report contains multiple charts, try not to
important for the management accountant, he mix different scales on the same page, as this
or she must think of designing a report from the can confuse the reader.
perspective of the graphic designer too. Consider
the following:
#ResearchInsight Our research shows that
• Layout – a simple and clean layout will always pastels are not desirable for business
deliver a more impactful report than a report reports, most organisations preferring clean
with too much information. Clear headings and sharp contrasting colours.
should be used that help explain what data is
being presented.

• If the report is ‘to be delivered digitally’, then

the visible area of the page should match the
screen, as scrolling is not desirable.

• Positioning – the reader’s eye naturally starts

to read reports from the left-hand corner of
the page and continues to the right or down.
Therefore, if there is a key message, this
should be positioned towards this area to gain
optimum impact.

The five principles for effective report visualisation:

Ensure data is Apply the relevant Choose an Optimise the Optimise

optimised for report visualisation tool appropriate report reader experience visualisation to
visualisation layout the appropriate
delivery channel

Report visualisation: from concept to deployment 7

8 Report visualisation: from concept to deployment
4 Optimise the reader
experience
#ResearchInsight I tell all of my team that
when they present a new report, a table is
the last resort if you need to get a message
Engaging with the report user is key to ensuring the
message is transmitted effectively from the paper
across.
or screen. When reporting in paper, think about the #ResearchInsight The key learning from
size and weight of the pack and consider dividers to my years in finance has been that a clean
help readers switch between sections. and simple report will always be preferable
For online reporting, there are many more options to a fussy report. Trying to add too much
available to deliver a better user experience: information has the inverse effect in that
the message is lost. Less is definitely more.
• Personalised – personalised reporting is key
to engaging with users. This can be as simple #ResearchInsight Whilst we will always
as offering ‘bookmarks’ or ‘favourites’, or as produce standard reports delivered to the
complicated as the reporting solution making
user, we also have made available a suite
intelligent reporting suggestions based on the
reader’s history or security access.
of self-service personalised reports which
are very popular amongst certain users.
• Intuitive – navigating searching around an The EVPs tend to like the standard suite
online report should be intuitive and as familiar but the VPs use the self serve.
as searching the web or using a smartphone.
Many reporting suites will include home
Optimise visualisation to the
buttons or searches similar to Google to assist
in the familiar user experience. A few reporting 5 appropriate delivery channel
software providers utilise ‘Gamer Influenced Reports can be delivered via many different
Design’ principles in designing a user channels, such as desktops, laptops, tablets, mobile
experience consistent with a video game. devices or paper. It is clear that a mobile phone is
considerably smaller than an A4 sheet of paper and
• Interactive – with touchscreen computers,
therefore logical that in order to produce a powerful
tablets and smartphones more commonly
report, the design will need to change.
being used for business, it is often desirable
to use touch in an interactive report. Whether Many software providers support functionality that
choosing the next report or drilling down to the automatically renders the report and layout based
next level, it is much more intuitive to swipe on the delivery device. This is termed ‘author-once /
than it is to click the mouse. run everywhere’.
• Customised – a certain group of report users, Whilst these tools can help, it is important
may want to customise their own reports, and to consider other aspects, such as the user
this is facilitated by certain software vendors. experience.
They can then create their own user experience,
but care must be taken to ensure the underlying For example, the smaller the screen, the less
data model is robust and does not allow the you can show on the start page. However, you can
misrepresentation of data by custom-built provide drill paths and filter options to dig into the
reports. next level as required.

You should also consider the urgency of the report

requirement. Urgent notifications can be sent to
mobiles but in order to fully analyse the detail, it
might be acceptable to wait until the user is in front
of a computer.

The previous visualisation principles become even

more important when thinking of various delivery
channels. Technology is just the enabler
to appropriately satisfy these.

Report visualisation: from concept to deployment 9

 Case study 1:

A global
industrial
manufacturer

10 Report visualisation: from concept to deployment

 Problem statement:
The organisation had grown significantly through mergers and acquisitions and as
a result had multiple reporting solutions delivering output that satisfied divisional
reporting requirements only. The company had recently formed a group reporting
function tasked with providing the CEO and CFO with the insight to drive the whole
business. Previously, reports had been pulled together in MS Excel from divisional
reporting which led to inconsistencies in data and definitions. These reports were in
a traditional row and column tabular format, which required user interpretation.

Approach:
Working with KPMG, a summary set of KPIs was developed that linked divisional
performance to satisfy the strategic goals of the organisation. These KPIs
were then expanded into more detailed measures that could be reported on
by the divisions to explain variances in performance. A mobile dashboard was
created that started with a single screen reporting the six strategic goals of the
organisation and then enabled the user to swipe into further information on a
detailed measures (swipe left) view or a divisional (swipe right) view. From each
view, drop downs were also created to intuitively navigate between different views.
Conditional formatting and interactive graphs were used to draw the viewer to the
issue quickly.

Benefits:
Providing a consistent view of the group and divisional performance based on
common KPIs was a giant leap forward for the company. The interactive way of
viewing performance was seen by all users as a much more intuitive method than
sifting through paper reports. The conditional formatting along with the graphical
techniques removed the need to analyse the numbers, allowing managers to take
action more quickly. By being on mobile devices, managers now had instant access
to financial performance at their fingertips.

Tools used:
Tableau

Report visualisation: from concept to deployment 11

How to obtain value
from data
Reporting evolution
Over a significant period of time, we have seen
report visualisation evolve from dot matrix printed
reports to application-supported reports to desktop
publishing-based reports and to the present day,
when we can facilitate highly visual, dynamic,
customisable reports on mobile devices.

The concept of report visualisation is not new.

In 1931, Harry Beck designed the London tube
map. Beck devised a map of straight lines running
vertically, horizontally or at 45 degree angles
which was visually easy to understand but is
only directionally correct compared to the
physical locations of the stations.

Fast forward to the modern day, and the same

principle is used by McLaren Formula 1 who depict
each track as an oval, no matter where they are
racing, because this delivers a simplified visual for
the driver to understand. KPMG has teamed up with
McLaren to exploit data analytical and visualisation
methods developed for the racetrack to solve
complex business problems in this area.

So what of the future? KPMG and Imperial

College London are currently developing a
Data Observatory. Core to the observatory is
the visualisation space, which presents data
in an enveloping circular wall of 64 monitors
with 310 degrees of surround vision and a total
resolution of more than 130,000,000 pixels with
a view to developing the most impactful
visualisation concepts.

Tools available in the market

Until recently, visualisation tools were part of large
enterprise-wide Business Intelligence (BI) solutions
requiring significant investment and specialist
IT skills.

In recent years, cloud-based technology has opened

up a new suite of products that can be owned and
developed by finance.

12 Report visualisation: from concept to deployment

Prices Tableau
The price of visualisation solutions has fallen Tableau is the market leader and has been for a
dramatically, in a trend that will continue as few years. It is easy to use and has comprehensive
more new entrants release products. online support. Perhaps Tableau’s biggest
strength is the number of data sources that can
The market leaders be connected via simple wizards, making it very
popular within the finance function.
Gartner’s Magic Quadrant for Business Intelligence
and Analysis Platforms is a globally recognised
report comparing the visualisation vendors.
Qlik
The market has exploded with new entrants QlikView and QlikSense are also easy to use and
and as of 2016, Gartner were ranking 24 separate are stronger where the need is for complex analysis
vendors of which the market leaders are: of data before visualisation. Qlik facilitates this
with an in-memory engine, which allows readers
Microsoft Power BI to see patterns in data better than with traditional
relational databases.
Microsoft’s BI offering is one of the most cost-
effective products on the market, is easy to use
and integrated with the MS Azure cloud offering.
Microsoft has increased the number of data
sources it can integrate with as standard and now
supports social media platforms such as Facebook.

The new breed functionality:

1. 
Data integration - all of the leading solutions offer pre-built data connectors to multiple-source
systems. IT may need to be involved to ensure there are no security issues to avoid systems
talking to each other, but as soon as the connection is made, you can start visualising.

2. Pre-built starter kits are also included within the solutions. These lead to an accelerated
development time and have been built with best practice in mind.

3. User-friendly functionality - combined with online help tutorials, this enables getting up to speed
as fast as possible.

4. Mobile reporting - used to involve complex coding. All of the leading BI providers now enable the
option to publish reports to mobile devices, whether tablets or phones. Whilst the technology
renders to the correct format, care must be taken to ensure the report is relevant to each device.

Report visualisation: from concept to deployment 13

14 Report visualisation: from concept to deployment
Reporting and analytics come together In the context of reporting, cognitive technologies
can anticipate the need for information and provide
Big Data for everybody it before being asked. ‘Google Now’, for example,
recognises repeated actions that a user performs
When Big Data first arrived on the scene, it
on a device and presents information to the user
was driven by the IT programmer developing
based on these patterns.
new technologies like Hadoop in open-source
technology user communities. The leading visualisation tools already have
some cognitive technologies - for example,
Tools such as Qlik, Tableau and Microsoft Power recommending a particular graph based on the
BI are now all able to connect to huge data data set chosen.
sources, including unstructured data, and have the
processing power to look for patterns in the data Predictive analytics
and provide insights. Big Data analysis is now
firmly in the hands of the management accountant. This area uses statistical techniques and machine
learning on large volumes of data in order to make
predictions about future events. More companies
#ResearchInsight Unstructured data has are using these techniques to drive financial
to be organised into structures by looking planning, removing the ‘gut feel’ element
for common patterns in order to derive of budgeting and forecasting.
valuable insight.
Internet of Things (IoT)
Cognitive technology This refers to physical devices that are able
Cognitive technology is where systems are able to to collect and exchange data via the internet,
perform tasks that only humans used to be able to such as a ‘Fitbit’.
do. It covers areas like computer vision, machine These devices are massively expanding the
learning, natural-language processing, speech ability to collect data from areas previously
recognition and robotics. not thought possible.

Application design principles

1. Standard functionality – custom modifications to the coding on reports will lead to difficulties for ongoing
maintenance and are often made on the basis of individual preferences, which may change over time.

2. Templates – use templates from the software providers if you can. These will be based on best-practice
principles. If these templates do not satisfy your organisation's requirements, then build your own
templates centrally and make them available to relevant stakeholders.

3. Design principles – develop design principles to standardise reporting formats and ensure they are easily
accessible from within the system. Corporate colours can be hardcoded within your standard templates.

4. Clear navigation – always think about your user journey. Reports must be easy to interact with or they
will not be adopted.

5. Emperor's New Clothes – just because you can publish to mobile or include the finest 3D graph does not
mean that you should. Always challenge and if the reporting is to be used on a large scale, establish a
Design Authority. Think about reports from the viewpoint of a graphic artist, rather than an accountant.

6. Keep it simple – this is the golden rule of reporting. Even when the functionality keeps growing, the
simplest reports are generally still the best.

Report visualisation: from concept to deployment 15

Getting started
and lessons learned

In designing performance reports, most Only then should time be spent on developing a
organisations become unstuck by not: visually impactful board report incorporating a
combination of graphs, icons, tables and summary
• Focusing at an enterprise level around commentary which articulates the performance
what should be measured and why of the organisation at a glance. To attain this will
require iterative mock–ups, taking into consideration
• Placing enough emphasis on getting
the preferences of different personality types in
data integrity resolved
developing a house style. In parallel, great care and
• Looking at the holistic picture on how reports consideration has to be given to ensuring that an
fit within one performance management agreed report style can be delivered efficiently and
framework. can be automated to reduce production burden.

In essence there is little point in developing a report Be aware that the report is likely to need to evolve
that looks good but drives the wrong focus on from an initial proof of concept into a sustainable
behaviours; to get visualisation right, the following report. Ensure that the design is managed,
building blocks need to be considered. developed and governed accordingly.

Lastly, the report needs to sit within a performance

Start with pilots, Proof of Concepts (PoCs) management framework which is used as a formal
and scale throughout the company mechanism to assess performance consistently
Critically, the best place to start tends to be at board across the organisation.
level (or equivalent) within the organisation. As
By adopting this approach, you have developed
mentioned, initial focus is centred on getting the
a style for the organisation which can be quickly
content right – i.e. the measures that will ensure
evaluated and embraced on a sustainable basis.
resources are enabled and aligned around delivering
the organisation's strategic intent.

Once these key metrics have been defined, they

should become the basis for designing of all
subsequent metrics cascaded down through
the organisation.

#ResearchInsight We are currently mid way

through a project to determine a design for
future reporting. We are using consultants
during this process who help us understand
what we really should measure and report.
For too long, our reports have evolved through
legacy requirements and are out of date.

16 Report visualisation: from concept to deployment

Make sure data governance is in place #ResearchInsight All our reports and data
Once top-level key metrics have been defined,
come from a centralised repository, which
emphasis should quickly gravitate to ensuring delivers a single version of the truth at all
that data models and governance are designed levels of reporting.
and deployed to manage the integrity of agreed
key metrics. This will require careful consideration A tool is not the (single) answer
to be given to:
By now it should be clear that the selection of
• Data definition along with data consistency – a best-of-breed visualisation tool will not on its
i.e. put effort into building a sustainable own provide you with effective reporting.
data model so that report integrity becomes
second nature Consider the following:

• Identifying data gaps and challenges • Ability of the tool to produce and disseminate
associated with key metrics early and put in defined reports and defined functional
place data-migration and mitigation plans to requirements
improve data quality • Ease with which such a tool can work seamlessly
• Developing master data management with legacy data warehouses, transactional
processes and governance models which are systems and data management tools
pragmatic and are not burdensome on the • Ease of use and effort required to deploy
organisation to maintain
• Features and functionality available
• Designing a common data model which as standard (in particular graphic and
is managed at one focal point across the commentary functionality)
organisation, but executed locally in line with
common policy • Cost effectiveness and vendor
development paths
• Maximising the use of data-management and
governance tools to manage data consistency • Productivity benefits as a result of
and integrity tool deployment.

• Ensuring that your data model design is flexible As a result, the recommended approach is to
to allow for changes in the organisation's conduct a formal assessment of a shortlist of
operating model and environment. potential vendor offerings against formally defined
requirements.
By placing such emphasis on data, you are
addressing the integrity of reports as part of the In addition to this, the approach to design, data
design and, as a consequence, ensuring that the management, governance, process and sustainability
organisation can rely on decisions made based on needs to be deployed in parallel to the tool for you to
report content that is presented in a practical and maximise the impact on the organisation and truly
informative way. enable informed decision making.

Report visualisation: from concept to deployment 17

 Case study 2:

A large retail
insurer

18 Report visualisation: from concept to deployment

 Problem statement:
The organisation had a history of unreliable performance against strategic plans.
Their strategic planning had little linkage to operational targets and business
performance reviews were focussed only on the current year. Hence, limited
strategic discussion on actions to take now to achieve the three-year plan. They
had minimal use of rolling forecasts across the divisions, leading to lack of forward-
looking insight. Hence, the three-year plan was always out of date in the second year.

Approach:
Strategic KPIs were identified to manage overall group performance using a
top-down approach. They were developed into a new performance management
framework that linked strategic planning, target setting, business planning,
forecasting and reporting and analysis, using consistent drivers of performance.
A digital dashboard was created in Microsoft Power BI for the board to manage
and review performance against strategic ambition. The digital dashboard now
linked actual performance to effects on the rolling forecast, which reported the
movements to plan using trend-based graphs that clearly showed the impacts on
current and future years.

Benefits:
By having the right data model underneath the dashboards, actual performance
was now linked to a rolling forecast. This meant business decisions were made
that helped deliver results outside of the current year. The digital dashboard, and
specifically the use of forward-looking trend graphs, allowed instant visualisation
of the effect of decisions across all years, which is difficult to obtain in traditional
tabular reports. Management is now able to react much faster to performance
variations and take corrective actions that can deliver the three-year plan as well as
current year.

Tools used:
Microsoft Power BI

Report visualisation: from concept to deployment 19

Report authors

John O’Mahony
Director
KPMG

Andy Carfax
Principal Advisor
KPMG

aicpa.org
aicpaglobal.com
cgma.org
cimaglobal.com
kpmg.com/uk

July 2017

© 2017 KPMG LLP, a UK limited liability partnership and a member firm of the KPMG network of independent member firms affiliated with KPMG International Cooperative (“KPMG International”), a Swiss entity.
All rights reserved. The KPMG name, logo are registered trademarks or trademarks of KPMG International.

CGMA and Chartered Global Management Accountant are trademarks of the Association of International Certified Professional Accountants and are registered in the United States and other jurisdictions.
The design mark is a trademark of the Association of International Certified Professional Accountants.