The Cost of Saying Yes

Founder friends, clients and people I see on X are all doing a version of the same thing. They have an idea, and within a week the product is live with customers everywhere: an iOS app, a Mac app, a Windows app, a Chrome extension. The building part has got ridiculously easy. Deciding what to build hasn’t.

Limited resources used to make a lot of those decisions for us. Years ago, at a blockchain startup, we were building a custodial wallet for Ethereum that felt normal to use. We spent a long time just figuring out the flow: what someone would see, what would happen next and how the whole experience would work. We were bootstrapped, so even bringing in a designer was a question of whether we had the runway.

There were always more features than we could build, so we had to prioritise them ruthlessly. Sometimes an idea was dropped simply because it would take too long or cost too much. It wasn’t an intelligent filter, and I’m sure we left good ideas unexplored, but it forced us to choose.

Now every surface you launch on is another place where customers interact with the product, and another stream of data coming back. I’ve seen it play out: a few customers on Windows ask for features that only make sense on Windows, others on the Mac want something else, and not all of them are profitable. There is no universal set of features, and you get pulled in different directions. Being available on four platforms isn’t the same as reaching four audiences, either.

Which platform do you back? Which customers? Which requests are signal and which are noise?

AI can help you analyse all that data, but more data isn’t better evidence. If you didn’t set a north star before you started the experiments, the numbers tell you what people did, not which of them has a problem painful enough to solve. The real limit is the time you have to talk to customers and find out. The cost of building used to force those decisions up front. Now we can put them off until after we’ve built and tested everything.

None of this is a nugget of wisdom I’ve suddenly come across. Operate from first principles: find the painkiller problems you notice and solve them extremely well for the customers who have them, rather than solving every problem for everybody. Talk to those customers and you can tune the product ridiculously well. Knowing that is easy. Following it is harder when everyone around you is shipping, shipping, shipping, and it feels like you’ll be left behind if you don’t.

So when building is no longer the constraint, why do we still act as though shipping more things is progress?