The 2007 post-apocalyptic horror movie I Am Legend, starring Will Smith, had a bizarre ending. Humanity has been destroyed irreparably by a virus that turns humans into strange vampire creatures that fear sunlight. Robert Neville is a researcher who spends his days alone, attempting to study these creatures and find a way to turn them back into humans through diseases. However, experiment after experiment ends in failure, with the subjects he captured dead on his table.
After spending the entire movie attempting to save humanity, his home is overrun by these Darkseekers, who find his basement laboratory. A young woman and girl he rescued previous are in the house, and Will Smith uses a grenade to blow himself and the Darkseekers to smithereens, saving two of the last humans.
However, for those that read the book, this ending was bizarre and meaningless. The reversal is that the Darkseekers aren’t the monsters, Neville is. He abducts them while they sleep, takes them to his basement, and kills them. To them, he’s the monster. This ending is far more nuanced, and in the book the Darkseekers take his final subject from him while he realizes that his walls are covered in the photos of the hundreds of subjects who died at his hands.
Why was this changed?
Focus groups. The focus groups that they ran the movie through didn’t like the ending, so they changed it to something far less impactful, less groundbreaking. The end result? Another standard, boring action movie that was forgotten a few months after release.
When it comes to user research in UX, I see many young designers obsessed with the idea of performing user research. Not just demographic research, but actual design A/B testing and focus groups.
What I suspect is that for most of these young designers, research is a crutch that they use to hide from allegations that their designs suck. “I haven’t done the research,” is a common phrase, or “I mostly focus on research.”
Can you program in R or Python? Do you have a data science degree? Are you running your research in blind studies and outputting the data visualization in Tableau or PyViz? Are you maintaining a clean, uncorrupted study throughout with controls for demographics and context? Is the data empirical? Are you running the same tests multiple times to ensure the results are consistent?
We all know the answer. True user research is done through data science, and data science is done by data scientists, not a bootcamper with Figma anxiety. Too many young designers are hiding behind this obsession with research for a single reason: they are terrified of designing.
UX is the promised land of non-technical six figure tech jobs for many people. Just search “Break into UX” on YouTube to drown in videos telling people how to network to find their job, how they don’t need to know how to design, how their portfolio should mainly focus on their process.
I recently had to contract a UX designer for an app I was about to release. I decided to open the requirements to include juniors.
Every portfolio I got was page after page of meaningless sticky notes, user flows, invented personas, and other extraneous bullshit just to reach the end and realize it was another SAAS landing page or mobile checkout. So many of them did pages of pages of completely invented research, tiny N values (4 people?) incredibly colored or leading form questions, all for an amateurish and straight up ugly final product.
You’re not going to break new ground with your research if the thing you’re working on isn’t groundbreaking. Creating a new social media platform or doing a design audit of Reddit isn’t going to require extensive user research.
The harsh truth is you’re not educated, prepared, or equipped to do real user research. A typeform survey and some A/B screenshots isn’t just ineffective, it’s an abject waste of time. Good design creates good UX. When your background color, font color and typeface are working in harmony, suddenly vision impaired people have no trouble using it. When you have adequate white space, suddenly the users are able to find the buttons they need to click. When you have great information hierarchy, suddenly you don’t need to spend hours screwing around with Google sheets to justify your salary and instead can focus on creating.
User research is sometimes necessary. Vitally so. For the Artemis rockets that are about to take off for the moon, I hope they’re doing extensive user research to ensure the airlock buttons are very clear about their purpose. For the new Riot MMO that’s coming out in 5–6 years, I hope they’re doing heavy user research to understand how players are going to manage their inventory, mounts, and 25–30 abilities. For the SpaceX mars colonies, I hope to god they’re keeping the oxygen tank controls far away from the intercoms.
But for your Netflix clone, your Tinder for gamers, your Reddit for Pharmacists, just stick with the basics. Be the catalyst. Get it done, iterate on feedback. Stop wasting time frontloading research on things that simple user heuristics cover.
