A button can look perfectly consistent in Figma and still create confusion. Which variant fits this task? What happens when its label becomes longer in another language? How does it behave when someone navigates with a keyboard? These questions changed how I thought about design systems. Creating a component is only part of the work.
A green uptime dashboard looks reassuring. But it can be wrong in the way that matters most. A website can return 200 OK all day and still be broken for the people trying to use it. The homepage loads. The monitor stays green. The server responds. Meanwhile, the contact form has stopped sending. Checkout will
AI products are often shown through their best-case scenario: Give the system an input, let the model process it, and get an accurate result in seconds, but production systems are rarely that clean. A scanned insurance form may be slightly rotated. A field may be handwritten. A passport image may be blurred. Different versions of
Confidence arrives before comprehension. A chart that is easy to read feels like a chart you have understood. This week I spoke to a friend whose startup will have to deal with this problem in the next year or so. We discussed how their users might find it difficult to understand the app’s data, and,
A fan watches their football team lose for ten months, through rain and heartbreak and the bottom half of the table, and then renews the season ticket in May, quite happily. That same person has shopped at the same supermarket for years, but when their coffee goes up by fifty pence they are across the