Trust me, I know what I’m doing

I used to be a developer. Oldskool. When we had to solve a problem, we’d dive in head first. Nothing but code. Code code code. We knew what we were doing. We knew what people wanted. Right?

Today, that’s pretty much unimaginable (except here and there, but here and there probably aren’t doing good business). UX (User Experience) design has changed the way we look at solving (digital) problems. We involve the user. First!

I spent 10 years as a developer (2003–2013). I only saw user involvement twice. My first experience was at Monumental Games, UK. We were developing MotoGP 09/10 for Xbox and PlayStation. Now, with game development, things get complex quickly. And you need a QA team to find the thousands of bugs that naturally occur in an unfinished game (I’m not saying this can’t be partially improved). The second case was at iChoosr, Belgium. We rented one of those UX places: one-way mirrored glass, microphones and cameras, screen recording, eye-tracking, the works! We had to figure out whether users actually understood and could use our web application.

Note that both of these occurred AFTER the majority of the development was done. The focus was usability. Not usefulness. Now, that’s better than no user involvement, but it still isn’t good enough.

In 2013 I started a PhD at the Human-Computer Interaction group of KU Leuven, Belgium. My goal? To create dashboards to help students become more aware of their activities, help them reflect and gather insights regarding their process and progress. So how did I go about that?

By diving into code head first…

Why would I create something without knowing what students needed?

Time to catch up

Putting data to good use often goes like this: hire a bunch of very smart people and let them loose on your data. Gone are the days (probably not but let’s imagine) where we have Report Ninjas make the craziest interactive spreadsheets you’ve ever seen and never thought possible in Excel. No, today data scientists will dig through your data and find patterns, or machine learning experts will create models that predict the future. Data visualizations and dashboards are often an afterthought, resulting in over-complicated dashboards on large displays in the office of the support or business team, who most likely don’t even look at it. (I’m oversimplifying, sure).

To create usable, but more importantly, useful and valuable applications, you need to involve the user. And we see this happen more and more across industries. However, when designing data visualizations, or dashboards, it seems some of us are still lagging behind.

Do NOT let the data speak for itself

I spent over six years researching data visualization design from a user perspective. While it is important to have a general idea of what data you have before thinking of solutions, the first question you should ask yourself is “what does the user need”. Because, if the user doesn’t need it, why bother anyway? You could save yourself a lot of trouble.

If you wish to improve your data solutions, or explore new paths, take small steps. Just take them in the right order. Here’s a list to get you started (this UX approach works for any digital solution by the way):

  1. Start a dialogue:
    Explore the potential of your data by talking to other teams (business, marketing,…), organize brainstorm sessions, get the conversation going!
  2. Involve customers:
    If your data has any public/customer value, set up surveys on your website, analyze your customer behaviors,… ask them what they need. You’ll be surprised at what you discover.
  3. Sketch:
    Do NOT start implementing! Explore ideas on whiteboards, paper, whatever works fastest.
  4. Get feedback:
    Take your ideas back to your potential users. Ask them for feedback, explore their needs further based on your sketches. Try to understand what you did well, where you can improve, and what paths you should ditch.
  5. Prototype:
    Create a visual representation of your idea. Do not develop, but help your users “visualize” the potential of the visualization (pun intended) by creating realistic mock-ups. If you still haven’t gotten the required budget for your project yet, this is perfect to take your results to management. Convince them it is worth pursuing (don’t forget to include the results of step 1–4).
  6. Iterate:
    Involve your customers in the design process. Go back and forth between design and evaluation. Gradually improve your solution. If customers are hard to reach, turn to your colleagues. Any feedback outside that of your own is useful.
  7. Implement:
    Make it real. Hand it over to the development team. But don’t stop there. Keep your users close. Learn from them (through tracking, interviews,…)

The entire process takes time. But every step has valuable, tangible and actionable results. I’ve been hired by companies to figure out their data potential and had results within days! New solutions they never even considered that turned quickly into new projects.

Now, you can hire someone like me who will go through these steps with you (to all of you on another continent: yes this is perfectly feasible as remote work too). My experience on the matter does speed up the process, I’m pretty good at designing data visualisation prototypes and I’m always excited to help companies discover the potential their data holds!

But! If you want to take matters into your own hands, I would advise checking this course I created. It will walk you through all the steps. It’s actually a great guide if you already have a project in mind, as it is a project-based course.

And if you happen to be near Brussels next week, I’ll be giving a talk on how this approach resulted in a couple of successful dashboards in both the field of Education and Unemployment.


I’m a Freelance Data Visualization x UX Researcher/Adviser/Designer. Want to chat or hire me, come say hi at!