Why Five Customers Can Tell You More Than 1000 Survey Respondents—The Power of Small Data

Decades of usability research—from software to physical products—have consistently shown that you don't need massive surveys to improve your design; you need meaningful, close-up observation of a handful of real users. When a team at a tech company was preparing to launch a new service, they debated if customer testing would be worth it. Some argued for sending out online surveys to hundreds of people. Others proposed in-person interviews with just five carefully chosen individuals.

The latter group won out. During a day of one-on-one sessions, customer after customer encountered the same obstacle on a key screen. One struggled with an ambiguous button, another hesitated with a confusing prompt, and by the third interview, the solution was clear: the issue was obvious and fixable. The company made rapid improvements before launch, saving them untold embarrassment and frustration.

The core behavioral principle at work is diminishing returns—after five users, each new interview reveals fewer unique issues. The richest, most actionable data comes from personal observation, not sheer volume. Teams that believe in 'small data' waste fewer resources, iterate faster, and end up with better products.

Identify five people who genuinely represent your target user and get their honest, unfiltered feedback on your prototype. Watch closely; most major problems and highlights will emerge quickly and clearly. Take note, be ready to change things, and don’t fool yourself into thinking you need an avalanche of data. This kind of testing is faster, cheaper, and infinitely more revealing than any survey—so set up your first round of interviews this week. You’ll be amazed at how much you learn before you spend big.