Data! Data! Data! I can’t make bricks without clay!’ These words, uttered by Sir Arthur Conan Doyle (author of the Sherlock Holmes books) are truer today than ever; even at its most granular level data is binary, objective and concrete; data don’t offer an opinion or come in different shades of the truth.
However, actually delivering information is only effective and will only deliver true unequivocal insight if it’s easily grasped by its audience without ambiguity. In my experience, too often the emphasis is placed on data discovery, resulting in the correct method of visualising the data being neglected.
In practice and as is well known, decision makers are deluged with information and squeezed for time. It’s also well known that what those decision makers need is at-a-glance data to inform those decisions. However, what isn’t as well known or understood is how a poorly presented piece of key information can completely undermine its currency and impact.
It is estimated that by 2020 there will be nearly as many digital bits in the world as there are stars in the universe, some 44 trillion gigabytes!1 Therefore, is it any wonder why C-level executives are hungry to harness the power of business intelligence (BI) to make sense of their companies’ data and gain a competitive advantage? Lee Warren makes the case for good data visualisation.
Examples could be poor choice of colours, inconsistent sizing or misalignment of graphics, all of which affect accessibility and make a dashboard something people are less inclined to look at. Additionally, whilst it is hoped that consideration is given to choosing the right chart for each metric, this is a practice often overlooked.
Knowing to display time-series values in a line chart (the majority of the time – there are some exceptions!), why 3D charts can lose some of the accuracy of how data is presented and when to choose a bullet chart over a spine chart requires expertise and experience.
Data should be presented alongside an intimate understanding of what decision makers need to know. Whilst dashboards are often developed collaboratively and informed by a systematic and comprehensive requirements gathering exercise, the main focus is on what is needed to be delivered and less on the how. Put simply, in a sea of important metrics presented in a dashboard, how does the key single message reach out at its audience?
The impact of misrepresented visuals on decision making and the ripple effect through a business is considerable. To effectively harness the power of BI solutions and empower individuals to effect change, the following interrelated concepts should be adopted across the business:
- Investment – financial and cultural
Appropriate funding should be invested to ensure the right presentation software is made available. Furthermore, time should be granted in the development phase to ensure industry-standard and bestpractice presentation methods are applied. Decision makers within the business should support the adoption of BI and do what they can to encourage topdown buy-in to maximise chances of successful integration.
- Change how a dashboard evolves
Best-practice BI feeds on input from all levels throughout an enterprise. Effective visualisation requires going one step further: deeper collaboration to prototype and workshop visualisation methods to iteratively ask questions such as ‘what do you see when you look at this chart?’ This approach is less about defined metrics and more about the semantic understanding, which, though likely to differ across individuals, must be factored into any frontend development.
- Recognition of the importance of BI across the enterprise
Investing in industry-standard expertise is vital, as is empowering those individuals to deliver a solution that has real impact on the business by supporting the decision making process.
- Embed the solution
A BI solution can only be successful if sufficient time, care, expertise and investment are given to how data are presented back to the end-user.
Incorporating well designed charting to sit on quality data is an important step. However, it’s essential that time is given to train users on the advantages and limitations of what a dashboard presents them with, to lessen the changes of misinterpretation. Dashboard development should not be viewed as a single activity with a clear end point. Instead, it should be something that is developed collaboratively and iteratively, engaging with end-users and adapting its look and feel as necessary to evolve the solution, in a controlled manner.
A BI solution can only be successful if sufficient time, care, expertise and investment are given to how data are presented back to the end-user. Neglecting this core area introduces the risk of ambiguity and threatens the adoption of BI, undermining the pivotal concept of informed decision making, which itself is a key driver in developing any business.
1. EMC Digital Universe report, with Research & Analysis by IDC, April 2014