Customer Interview Questions That Find Your Messaging
Customers who try the product stay, but most visitors leave. Are we describing it in the wrong words?
First Round Review · Matt Lerner
Published Updated
When a product that works still converts poorly, the words are often the cause. Visitors scan for a match with goals and struggles already in their heads. Matt Lerner's method finds those words by interviewing recent buyers about how they came to buy, turns their goals into headlines that finish the sentence Now you can, checks that strangers understand them in five seconds, and only then tests them in ads.
When the words are the reason buyers leave
Your product works and customers who try it stay. Yet ads barely pay for themselves, the website converts a few percent of visitors, and nobody but the founder can sell it well.
Matt Lerner, who ran B2B growth teams at PayPal and later invested at 500 Startups, argues in First Round Review that the cause is often a handful of words. He calls the fix language-market fit. It means describing the product in words that match goals and struggles your buyers already carry.
His opening example is one headline change. "Accelerate your learning. Develop your thinking through online discussion" converted 3 percent of visitors. "Organize, share and take notes on web pages, PDFs and academic papers easily, so you can learn faster, remember more and get things done" converted 30 percent. Same product, same traffic.
Why exact words matter
Lerner's explanation starts with how people read a site they do not know. They do not read it. They scan, looking for a match with something already in their heads, and decide in an instant whether to stay.
Psychologists call this goal-directed attention. The visitor holds a target, a goal or a struggle, and pattern-matches against it before conscious thought begins. Words that match get more attention. Words that do not, however accurate, are passed over.
A company everyone knows can get away with abstract language. A company nobody knows cannot. It has to use the buyer's words, and the only reliable way to learn them is to ask.
Interview the people who just bought
The research is an interview. Lerner uses the jobs-to-be-done technique developed by Clayton Christensen and Bob Moesta, which treats a purchase as hiring a product to make progress on something.
Find someone who recently bought your product. Before launch, find someone who bought a competitor's product, built a workaround, or asked around for alternatives. What matters is that they recently spent time and money on the problem you solve.
Ask them to walk through the purchase from the beginning, as if you were filming a documentary. When did they first notice the struggle? Where did they look for help? What did they try first? What will this purchase let them do, and why does that matter?
Two habits make the interview work.
- When they use a vague word such as better, flexible, or all-in-one, stop and ask what it means to them.
- When they describe a decision, ask what prompted it on that particular day.
Write down their exact words. After about five interviews, patterns appear. One founder in the article ran his transcripts through a text analysis tool, pulled out the verbs customers used most, and built his 30 percent headline from them.
Turn their words into headlines
Take the goals and struggles you heard and write them as headlines that finish the sentence "Now you can." If your headline instead finishes "Our product is," you are using your words, not theirs.
Lerner's warnings are specific. Avoid platitudes such as all-in-one, faster, revolutionary, and convenient, because buyers cannot tell what they mean and discount them as marketing. Stay specific even when the audience is broad. One compliance startup selling to very different industries found its winning lines were concrete, such as "I'm not ready for an audit."
Check that strangers understand
A test can show which headline wins. It cannot show why. So check comprehension first.
Show each headline to someone for five seconds, then take it away. Ask what it said, what they think the product does, and what it would do for them. If they can only repeat the words back, they did not understand them. Test each headline on four people, revise, and repeat. Four rounds fit in an afternoon.
Only then test the finalists quantitatively. Lerner suggests running the same simple ad with different headlines, because landing page tests take too long without heavy traffic.
What this means for a founder-led company
Expect the result to sound plain. Lerner warns that marketers will call it boring. That is the point. A company without decades of awareness has to be understood before it can be admired.
The interviews are the part to start this week. Five conversations with recent buyers, recorded in their own words, give you language no brainstorm will produce.