Hey everyone. It's been a while since I've written, mostly because I haven't had time to write the longer form essays that started this newsletter. But I've been missing writing. As a solution, I've started writing shorter-form "screenshot essays" on my X account, mostly focused on startup advice (based on my experiences as a startup founder and investor/advisor to other startups). These are much shorter posts, but I'm going to share a few of them here from time to time in case some of you find them interesting. Enjoy.
There’s a saying in startups: being early is the same as being wrong. If you’re right but you show up to the market early, it doesn’t matter – no one is ready to try your product. Shipping too late is just as bad, but for other reasons – especially if you’re doing so to overly prepare for GA in a closed beta.
First off, holding back on shipping to GA so you can optimize in a closed beta almost always leads to imperfect signals. Everyone in your closed beta is by definition biased. They proactively chose to join your small beta and jump through all the necessary hoops of signing up for it: filling out a Typeform, installing TestFlight, joining your Discord, and maybe even giving you feedback if they’re really pumped about what you’re building. They are extremely qualified, and this will show up in their engagement metrics and feedback. You will think your product is better than it actually is.
Second, because your closed beta is made up of highly qualified users, the feedback loop around user needs and product roadmap is heavily skewed. You end up building features that power users want – not the general population. The problem with this is that power users almost always want different things than the rest of the world, and you’re trying to get really big – not simply cater to a small, niche audience.
Third, and possibly worst of all: by the time you do finally get to GA, you’ve built a load of features specifically designed for power users – and now you have to support those features forever (or until you decide to pivot). You’ve signed yourself up for an eternal backlog of product and technical debt you must continue to pay off in perpetuity, thus slowing down your sprints on the stuff that actually matters. In other words, you are now simply a slow product team, maintaining features that can’t find and sustain PMF.
Conventional startup wisdom suggests you should perfect your product in beta before you go to GA. There’s a fear that you’ll “burn through your potential users” on an imperfect product. There are nearly 8 billion people in the world. You’ll be lucky if you can get a few thousand people to try your product in the first week, regardless of whether your product is great or terrible. I say skip right to GA, ship constantly (at least every week), and start learning on real-world data ASAP.