User testing consists of observing real people interacting with a product or prototype to identify friction points, clear paths, and stumbling blocks—well before freezing the design.

Jakob Nielsen popularized a beautifully counter-intuitive reality: you can detect the vast majority of usability issues with just five users. There is no need for a massive statistical sample to see that users are missing a crucial button; instead, we need to observe a small group deeply and early. A test is not an opinion poll (“do you like this?”) but a keen observation of behavior (“can you achieve this?”).

This is where the decisive tension lies, and it is entirely in the mindset of the tester. You can test to validate—seeking confirmation that your design is perfect—or you can test to learn, actively hunting for what is falling short. The first approach is a trap: it leads to leading questions, rationalizing failures (“they just aren't used to it yet”), and hearing only what reassures us. A test conducted just for reassurance is wasted effort. Instead, an exceptional test aims to prove us wrong as early as possible: it welcomes hiccups now to prevent costly failures later.

Hence the discipline: observing without dropping hints, welcoming user struggles as valuable gifts, and distinguishing what people say from what they actually do. User testing only drives true differentiation if we are genuinely excited to discover where we got it wrong.

Our philosophy: test early, with a compact group, in real-world scenarios, and with a healthy appetite for being contradicted. A test that reveals nothing new was likely designed just to prove you right. It is always better to have a brilliant failure during testing than a costly flop in production.

Where the magic also happens: prototyping and iterations, usability, cognitive biases (specifically confirmation bias), ethnography, and design thinking.

References: J. Nielsen, "Why You Only Need to Test with 5 Users" for usability testing.