PG Canon · Chapter 03

Understand Your Users

The single highest-leverage skill in a startup is understanding your users — what they actually do, not what they say, and not what you assume.

After this lesson

  • Separate what users say from what they do
  • Use Startups in 13 Sentences as a running operating checklist
  • Turn user contact into a repeatable weekly practice
Understand your users. That's the key. The essential task in a startup is to create wealth; the dimension of wealth you have most control over is how much you improve users' lives; and the hardest part of that is knowing what to make for them.
Paul Graham, Startups in 13 Sentences

The one thing that matters most

Of thirteen condensed principles for startups, this is the one PG says he'd keep if he could only keep one. Nearly every other principle — make something people want, launch fast, talk to users, don't scale prematurely — is really a special case of understanding your users well enough to know what they need before they can articulate it.

What comes from years of direct user contact

  1. Users will be polite and say things they don't mean — watch what they do, not what they promise
  2. A close, sustained relationship with a small number of users beats a survey of many
  3. Support and sales conversations are a product research channel, not just a cost center
  4. The users who complain the loudest are often your most valuable source of what to fix next
  5. What a user asks for by name is rarely the simplest fix for the actual job they're trying to get done

Case

Case: the export button that was actually a spreadsheet

Several users of a billing-analytics tool were quietly exporting raw numbers every week and rebuilding the same chart by hand in a separate Google Sheet. None of them had mentioned this in onboarding calls or filed it as a feature request — it just looked like normal Tuesday-afternoon work to them.

Takeaway: People are bad at generalizing their own repeated workaround into a feature request; the workaround feels like their problem to manage, not the product's problem to fix. Watching a screen-share of someone's actual week surfaced a need that months of direct questions had missed entirely.

A user-understanding loop

  1. Watch someone use the product without prompting or explaining anything to them
  2. Ask what they were trying to do, not what they think of the feature
  3. Note every workaround, hesitation, or muttered complaint
  4. Convert one workaround per week into a hypothesis for a fix

Trust the words or the behavior?

A user says they love the product, but their usage tells a different story.

Read the original

Startups in 13 Sentences

PG's condensed operating manual — most of the thirteen are really about understanding users.

This chapter is a working summary. When the idea matters, read the source once — then come back and do the practice.

Read the original

What I've Learned from Users

Lessons pulled from years of direct founder contact with users of YC companies.

This chapter is a working summary. When the idea matters, read the source once — then come back and do the practice.

Practice