It’s all about first impressions

Today, I watched a customer try a prototype of a mobile app. He was navigating through a list of something. (It’s a secret what the app is, so I’m being intentionally vague. “Something.” Woo, mysterious!)

He successfully got through the list. No problem. Instantly.

Then he said “That’s too hard. People won’t figure it out.” He proceeded to demonstrate the difficulty of the gesture. It wasn’t natural, it was undiscoverable, and no one should ever be expected to do it.

But… he did.

We all saw it on video. Look, now I’m rewinding the video. There he is, doing it again, without thinking. I might even call it effortless.

What just happened? A textbook example of reaction vs. feedback. In this case, his reaction was to just flip through the list without thinking — and it worked! His feedback, on the other hand, was an honest attempt to be helpful. It just didn’t match reality.

This distinction between reaction and feedback — introduced to me by my good friend John Zeratsky — is important. See, when companies are reluctant to talk to their customers, they often mistakenly repeat this quote by Henry Ford:

“If I had asked people what they wanted, they would have said faster horses.”

Yo, check this out:

A. The quote is B.S. Henry Ford never said that.

B. Talking to customers is actually an excellent (and cheap) way to get data.

C. But… Yeah, people don’t always know what they want.

D. So, when doing customer research, remember this:

Reactions > feedback

Reactions are hard to fake. Feedback is hard to give. It’s easy to talk yourself into an opinion when you’re trying to be helpful. (I do this myself all the time.)

Be wary of feedback that contradicts reaction. Lean in close and watch how people use it — before they think.


If you liked this post, check out my book Sprint, about the Google Ventures design sprint method. It’s a book about surprising ideas: that the biggest challenges require less time, not more; that individuals produce better solutions than teams; and that you can test anything in one week by building a realistic façade.