Avoiding bad data by asking the right questions
Three types of bad data from customer conversations: compliments, fluff, ideas.
This post is part of my summary of The Mom Test, a practical book for better customer conversations.
Types of bad data from customer conversations
With the exception of industry experts who have built very similar businesses, opinions are worthless.
You want facts and commitments, not compliments.
Deflect those like Shaq does with bad vibes.
Say no to compliments
We discussed "chasing compliments" on my very fake T-shirt business example.
Again, good humans are supportive, they want to see you succeed, and they will tell you good things.
The best way to escape the misinformation of compliments is to avoid them completely by not mentioning your idea.
The book's example of a bad vs. good conversation is genius [page 18.]
I won't copy-paste the whole thing, though. Get the book.
When compliments come your way, use this:
Back to PJ fake t-shirt biz
PJ: And it's like expensive black t-shirts but better!
Potential Customer: That's really cool. I love it! (compliment, yikes)
PJ: Whoops, sorry about that, I got excited and started pitching. What was the last piece that got into your wardrobe? What's your habit for shopping clothes? (That compliment made me suspicious. Let's deflect it and get to facts and specifics.)
Say no to fluff
When someone starts talking about what they “always” or “usually” or “never” or “would” do, they are giving you generic and hypothetical fluff.
Ask good questions that obey The Mom Test to anchor them back to specifics in the past.
Remember, asking good questions is as easy as 1, 2, 3.
Say no to ideas
Ideas are cheap and rarely thought through when coming from customer conversations.
There's more value in understanding the core desire behind the idea.
I codify desires using Reforge's map from Experimentation Deep Dive.
A different take from Michael Margolis
Rookie researchers usually fall for leading questions, those that suggest a particular answer and contain information the examiner is looking to have confirmed.
Mentioned in the book Sprint: How to Solve Big Problems and Test New Ideas in Just Five Days, Michael’s advice to avoid leading questions comes down to just two rules:
DON’T ask multiple choice or “yes/no” questions.
❌ Would you? Do you? Is it?
DO ask “Five Ws and One H” questions.
✅ Who? What? Where? When? Why? How
Deflect hypotheticals and anchor on recent past facts.
Your new challenge
Learning that your beliefs are wrong is frustrating, but it's progress.
It's bringing you ever closer to the truth of the real problem.
You want the truth, not a gold star.