Choose Your Customer

Rahul Vohra raised Superhuman's product-market fit score from 22% to 33% without shipping a single feature; all he changed was who he counted as his target market.

The score came from a startup survey: ask users how they would feel if they could no longer use it, and score the share who answer "very disappointed." Scored across everyone, Vohra's came back at 22%. He used those users' own words about themselves and the value they got from Superhuman to define the target customer.

Then he ran the math again counting only people who fit that description, and the score was 33%.

The new customer description created an almost uncomfortable level of detail. Their target was now someone who lives in their inbox, reads 100 to 200 emails on a normal day, typically sends 15 to 40 emails back, but as many as 80 on a busy day.

Counting fewer people might sound like gaming the numbers, but it came with a real decision: Vohra picked that one type of customer to build for, and said “no” to feature requests from everyone else.

You can copy Vohra's move without running a survey. Write down exactly who your product is for, in one sentence, and let that sentence decide which feature requests you turn down and which you pursue.

P.S. AI has made shipping features easier than ever, but that's only revealed the real cost of trying to please everyone: losing focus.