back at the factory

I'm back. The collab that ate February and March is done. I didn't enjoy any of it. I did learn how to work in a new environment, which is a real thing, even if it wasn't the reward I was after.

During those two months I couldn't focus on my own apps. Same story as the last post: splitting attention across too many things drains the fun out of everything. It was exactly as true the second time.

Early this year I also made a decision I hadn't written down yet. Stop building the fun-idea apps. Focus on boring apps. Utilities, daily-use, things that print ad revenue instead of things I want to show friends. That decision took a while to stick. Now it has.

This week I submitted two apps. Ten more are building. I'm building one right now, as I write this.

What actually changed is the factory. I've been building nst, the Nante Studio CLI. It scaffolds an Android app with ads, push notifications, and a content pipeline wired up on day one. Still rough in plenty of places, but it prints apps. Building the CLI itself is its own kind of fun.

The worries are real. App reviews on the store are slower than I'd like. And I'm not sure yet that any of these apps have the retention to justify spending on Google Ads — the whole factory model only works if paid acquisition is net positive, and I don't have the data to know that yet. I guess I'll find out by deploying and running.

Where this is pointed: master AdMob, Android, Google Ads, Firebase. Get paid by Google for building things Google's ad network wants to serve. It almost feels like being employed by Google, except I pick what to build and nobody approves my pull requests.