2 Notes on User Feedback

October 16, 20202 min read

1) we don't know what we want

I had a road trip last weekend, and it was my first time driving a huge SUV (Chevrolet Tahoe).

On highway, I was distracted for a split second and the SUV swerved slightly towards another lane. To my surprise, the steering wheel smartly steered the car back to my lane without my control.

Turns out the SUV has a cool new feature called Lane Keeping Assist (LKA) that has sensors at the side to make sure you are driving within your lane.

And when a car in front of me slowed down abruptly, the Adaptive Cruise Control feature automatically applied brake to keep a safe distance.


I've been renting cars for the past 10 years.

10 years ago, if Chevrolet were to ask me what I want in an ideal car, I could only think of making cars more fuel-efficient.

2 years ago, If Chevrolet were to ask me if I need a potential feature similar to Lane Keep Assist and Adaptive Cruise Control, I'd be skeptical - does it really work? is it really necessary?

But now that I have tried it in my hands, I love it.


We often don't know what we want until we can see, touch, and use something.

That means:

A - If you are asking potential customers about an idea, you need to be careful not to taking their feedbacks at face value.

I recently read The Mom Test - how to get real insights from customers by framing questions in a way even our mom can't lie to us. You can read my summary here.

What is unmistakable from customer feedbacks is the pain points they face currently. E.g. if Chevrolet were to ask whether I dooze off or lose attention sometimes in a long drive, the answer would be a clear yes.

B - The best way to know if your idea works is to build a prototype and put it in customer's hands.

Even if you have done great user interview and shown that the problem exists, it doesn't mean your solution will solve the problem.

There is no shortcut to getting your idea validated other than building wireframe / mockup / prototype and getting them in the hands of a user.

2) how Lunchclub.ai pivoted with user feedback

Lunchclub.ai is a cool platform that uses AI to match you with people for a 1-on-1 video chat for networking purpose. I've been using it since Covid and I've met interesting people there.

It was initially founded as a prototype project that built a simple reminder bot that reminded people to reconnect with friends they hadn’t spoken to in a while. But user feedback taught them something else - people were more interested in connecting with new people than people they already knew.

So they pivoted to matching you with new person for 1-on-1 coffee meetup (in person). For each city, they only start matching when there's 500 people signed up in interest list.

When Covid happened, they pivoted again to virtual 1-on-1, and the product exploded as the liquidity of the platform is much higher without the limit of geography.

Source: https://consumerstartups.substack.com/p/-newsletter-24-find-your-wozniak

Thank you for reading

If you'd like a monthly email on new posts, please pop your email here.

If you like this article, please click on the meaningless heart icon below.

0 loves

© 2020 yinhow

built with

Gatsby Logo