Data visualization: making the right decision at scale

Visualizing your data the right way is incredibly important for working data-driven within your organization. Data Strategy Leader Louis de Roo explains why in the blog below.

Let me start with a disclaimer: writing about data visualization is a bit like macramé-ing about politics: you might manage to get your ideas across, but you'll still be expressing yourself in the wrong medium. This piece would of course have been much better with a brilliant set of visualizations to reinforce my story. The reason I'm writing about it anyway is that, like most of us, I'm not all that good at visualizing data myself. Okay, maybe I've picked up a bit more of the underlying principles, and I've learned to recognize good and bad examples: but I'm still definitely no specialist.

Is visualizing data an unnecessary luxury?

For most of us, visualizing data is a nice extra. Having grown up with Excel and being familiar with the data in front of us, it can seem like a somewhat unnecessary luxury to spend time and energy "making our data pretty." The numbers don't lie, right?

The uncomfortable truth is that there's a bit more to it than that. Indeed, numbers don't lie. But making numbers lie is a lot easier than it seems. We often even do it unintentionally. And you don't need to be a shady accountant to make certain information disappear from the numbers.

Why look at data?

Before I launch into a tirade against the company of deception that sneakily manipulates us by showing or not showing certain data, it's good to go back to the why for a moment. Because why do we actually want to look at data? The basic idea behind data-driven working is that we can steer our organization by means of data. The follow-up question is perhaps even more important: how? Steering your company with data is one thing, but do you want to be the only one who understands the data? Or would you rather have your organization have access to the relevant data across the board, and be able to make the necessary decisions based on it?

Data-driven culture

In a truly data-driven culture, everyone in the organization has access to the data that's relevant, and everyone is able to make the right decisions based on that data. But it's precisely in that "being able to" that the sting lies: because where you effortlessly keep an overview of hectares of Excel lists, your employees may find that a bit more difficult. Simply because they don't look at the data daily, or have a somewhat less sharp sense of what the crucial values are that your process should be steered by. Of course, with a lot of effort and time you could drill them on exactly what to look out for, but is that really the fastest way to a more effective approach? No company has only trained data analysts on staff, and besides: the work still needs to get done. Preferably without too much delay.

Clear message

By playing with "what" you show your people, you can influence how they'll react. In this era of fake news and conspiracy theories, that's not news. But even within your own organization, it's a simple fact that most employees will act faster based on a clear, unambiguous message than based on a complex set of data. Making data and the decisions that follow from it accessible and translatable is therefore an important step, and one that can lead to a much wider application of your insights than you could have achieved on your own. Clear messages, clear actions, and clear appreciation of what's good or not good enough: visualization ensures that not only the sender but also the receiver understands what the data is telling them. With good data visualization, you reduce the need for other information transfer in your operational processes, process steps connect better with each other, and everyone works from the same source data toward the same goal.

Visualization standard

An additional benefit is that choosing a good visualization for your data also forces you to think about your own values: what do you consider good and bad, what's important or insignificant? By choosing a stronger logic when visualizing your data, you'll also encounter exceptions in your process, which you can then start doing something about. So a visualization standard is more than just a nice "coat" for your data: it forces you toward more sharply substantiated choices in your business processes, toward clear measurements against your goals, toward targeted actions and responsibilities.

If you, as a manager, don't want to start every workday handing out a stack of orders to your employees, good data visualization is indispensable. Then it's just a matter of admitting you're not good at it — and seeking help.

Written by Louis de Roo
Data Strategy Leader