Seibel says founders should be able to state the problem in one sentence, define a narrow first group to help, and check whether the problem can actually be solved. Poppy, an on-demand babysitting service, found that infant care required a level of trust and skill that its pool of flexible babysitters might not provide.
Building Product
Speaker
Michael Seibel, CEO of Y Combinator and cofounder of Justin.tv, Twitch, and Socialcam, explains how to identify a product’s customer and problem, test solutions, and build a focused development process in a YC lecture.
Seibel urges founders to identify who has the problem, how often it occurs, how intense it is, and whether those customers are reachable. Car buyers may shop only every seven years, while dealerships need to sell every day; that difference can reveal who gets more frequent value from a product. He also recommends charging early when customers will eventually pay, since free users may not have an urgent need.
Seibel says to build quickly, give the product to customers, and check that each part addresses the original problem. Early users should be desperate for a solution, not merely friends or impressive companies; Socialcam cofounder Michael’s friend Steve, then Reddit’s CEO, admitted he was glad to delete the app after it was sold. Seibel also recommends identifying customers who exploit a service or give feedback without having the problem.
Seibel recommends event based analytics, beginning with five to ten key actions and defining measurement as part of each product spec. He contrasts Justin.tv’s slow, argument driven, three month cycles with a process that starts from a company KPI, brainstorms ideas, sorts them by effort, writes a spec, and ships every week or two. At Socialcam, the team tracked DAU through new users, retention, and new content.
Seibel distinguishes a pivot, which changes the customer or problem, from iteration, which changes the solution. He says finding product market fit can take years, and points to the first iPhone’s missing 3G, limited carrier support, poor battery life, and lack of an App Store as evidence that successful products improve over time. When Justin.tv began talking to gamers, small changes such as fixing lag showed users the team was building for them.
Takeaways
- State the first customer’s problem in one clear sentence, then narrow the initial use case.
- Check whether the problem is frequent, urgent, and solvable with available resources.
- Get feedback from people who urgently need the product; friends and investors may not be representative users.
- Track a few important user actions from the first release and connect product work to a company KPI.
- Use short build cycles to test solutions, and change the customer or problem only when evidence calls for it.
Mentioned in this talk
Search inside this talk
What was claimed 14
- ceo ofMichael Seibel CEO of
Y Combinator
CEO“Michael Seibel, the CEO of Y Combinator”
YouTube Verified2 sources· Talk
- built
“Justin.tv, which later became Twitch”
0:00inWorking at Big Tech Companies Can Be a TrapYouTube Verified3 sources· Talk
- foundedMichael Seibel founded Justin.tvco-founder
“I was co-founder of a company called Justin.tv”
0:00inWorking at Big Tech Companies Can Be a TrapYouTube Verified3 sources· Talk
- featurePoppy: Uber-like model for babysitting
“Poppy is essentially Uber for babysitting.”
YouTube Unverified1 source· Talk
- featurePoppy: Helps parents find babysitters
“They make it really easy for parents who need babysitters to find one.”
YouTube Unverified1 source· Talk
- built
“My company was doing video sharing with friends”
YouTube Unverified1 source· Talk
- feature
Google Analytics: Reports website visitors, page views, and referral sources
“how many people came to your website today, how many pages they viewed, and where they came from”
YouTube Unverified1 source· Talk
- featureMixpanel: Event-based product analytics that tracks user actions
“For all of those things, you want an event-based metrics product.”
YouTube Unverified1 source· Talk






