Brian Balfour

For startups, you don't have the luxury to spread your chips across platforms. You have to choose one and go all in. Scarce resources, scarce attention. Higher risk, higher reward.

When evaluating a new distribution platform to bet on: (1) retention + depth of engagement, (2) user quality and monetizability, (3) the value exchange they're offering developers, (4) pure scale. Then immediately plan your exit before the platform closes.

The cycles are getting shorter and shorter. You have less time to ride each new platform before it closes. Whatever window existed for Facebook or Google, the next one will be smaller.

ChatGPT has the "smile curve" — retention that goes down, then comes back up higher over time. I've seen this in very few products in my career. Slack had it. It's extremely rare and almost always predicts escape velocity.

Most CEOs are incredibly disconnected from what's actually happening with AI adoption inside their companies. They announce the decree. They think it's spreading. It's not. Almost 90% of the time, it's one or two people using the new tool.

For AI platforms, the better signal is retention and depth of engagement — not MAU or signups. Google and Facebook both won not by being the biggest, but by having far better retention than their competitors.