Should I build this feature or talk to customers first?

The three-question test

Before you start a feature, write down answers to these. If any answer starts with "I think" rather than "they told me" or "I watched them," go and talk to people first.

• 1. Who exactly is this for? Name specific people.
• 2. What do they do today without it? The workaround tells you how much the problem costs them.
• 3. What changes for them when it exists? If you can't say, they won't notice it shipped.

When building is the right call

Sometimes building is the fastest way to learn. If a paying customer is blocked without it, if the feature is small enough to ship in days, or if showing something concrete will get you better answers than describing it, build. Decide in advance what result would tell you it worked.

Why founders build instead

Building feels like progress and talking to customers feels like risk. A new feature can't reject you; a customer can. That's exactly why the conversation is usually the more valuable hour.

Common questions

How many customers should I talk to before building a feature?

Five people who have the problem is usually enough to see a pattern. If the five disagree, the feature isn't clear yet.

What if I hate talking to customers?

Keep it short and specific. Ask about the last time the problem happened and what they did. You're listening.

Should I build features that users request?

Treat requests as clues to an underlying problem. Ask what they were trying to do when they wanted it. The fix is often smaller or different.