Wunderman Thompson London
8 months ago
If you’ve been glued to the data headlines lately, you’d have noticed massive consolidation in the visualisation space. Google snapped up Looker and Salesforce acquired Tableaux in huge deals, each of which was inked on the premise that improving data visibility, context, and relevance will drive better decisions and greater value in the digital economy.
Don’t get me wrong, these platforms can do great things today. But throughout the last 20 years, we’ve heard the same thing many times: companies promise to bring data-driven insight to the masses that will enable us to make more informed decisions. So, what’s different this time? Data visualisation done right does democratise data, establish context, and business relevance. It can deliver real insight to non-data types.
For a simple example, imagine you’re running a digital advertising campaign to drive sign-ups for a new insurance product. You’re targeting five segments with a variety of automated ads spanning display, web, and email. Data visualisation enables operations to monitor and tune channel performance (e.g., frequency, cadence, and spend) and informs the CMO of your campaign’s contribution to sales objectives and brand awareness.
While it may seem like we ought to be doing already, it’s a practice more honoured in the breach than the observance. Most companies’ data is not set up to be nimble, insight isn’t delivered in real-time, and it’s hard, quite literally, to see where adjustments should be made. Data visualisation has the potential to change this for a huge range of business applications - but not if we don’t get a few things right first.
Get good data. No one should have to explain this problem in 2019, but it’s surprising how many companies are awash in data that is out of date, incorrect, and duplicative. If you visualise bad data, you make bad decisions. The first step, long before worrying about data visualisation, is making sure you do the hard work to ensure a high level of data integrity and accuracy. Unless we get the fundamentals right, the fancy stuff doesn’t matter.
Make sure you have actionable KPIs. If you want to move the needle, it’s important to know what you can affect and why. Again, this is very basic, but your data visualisation will be useless if it does not target KPIs that matter - and that you can change. A vehicle manufacturer may want to be perceived as safe or innovative, but that’s simply not going to happen if you don’t know how “safe” or “innovative” can be quantitatively discerned from customer engagements, no matter how much data you throw at it. Instead, such a company should do the homework of identifying measurable goals and use data visualisation to activate and achieve them across the enterprise.
Understand the audience. There is no one-size-fits-all data visualisation. A middle manager certainly needs different insights from an executive. And some time-starved executives want only the big picture, while more micromanaging execs like to dig in. Developing the right views requires identifying a level of detail and context that is meaningful, actionable, and aligned with personal and corporate goals. This can be as much art as science. This is a good time for most organisations to bring in an outside resource with experience setting up data visualisation. Seasoned outsiders likely already made plenty of mistakes - and know how to avoid them.
Invest in storytelling. There is a big difference between a data scientist and a data storyteller. The former is an increasingly common and sought-after role. The latter is a specialist who is a cross between a scientist and an artist, able to take ones and zeros and turn them into a compelling narrative that resonates with a particular audience. If the goal is to democratise insight to drive better decisions, you certainly need people who can create stories with contextual relevance for your audience—and it’s an added benefit if you have a few folks who are strong visual designers. Your insights will drive a lot more decisions if the visualisation is clean, organised, relevant, and, dare we say, nice to look at.
In other words, the road to great data visualisation draws on the vast foundation of data engineering that has been honed over the last few decades. However, great visualisation is not solely the result of adding one more powerful tool to your data lake or integrated platform. Rather, it is the combination of a solid data foundation, measurement discipline, and the mastery of storytelling that enables better decisions and breakthrough results.
David Butt is consulting director, London, at Acceleration, a Wunderman Thompson company
Scott Molitor is consulting director, North America, at Acceleration, a Wunderman Thompson companyWunderman Thompson London, 8 months ago