Heck, maybe even secretly subscribe to a few similar keywords automatically, and in a week or two tell me how many notifications I've missed out on without a paid plan 👀

Another option here would be to show me some sample keywords and results to give me a sense of what I should be thinking about as keywords for myself.

There's probably an upsell opportunity for you here too, to recommend similar keywords to ones that I've entered...

Ha! Some feedback from my onboarding:

- it'd be a great idea to guide me to set up notifications _before_ keywords, so my first result ends up in my inbox/slack/discord
- when setting up keywords, seeing a recent result or two would help me feel more confident about what I'll get in the future

Sure as heck will! @cmgriffing.bsky.social this is 🔥

Y'all need to go check out messijo.com - it's such a tidy tool for keeping an eye out for mentions across networks.

mikebifulco.com/newsletter/o...

You'd have already known this if you were getting 💌 Tiny Improvements in your inbox each week, of course. Read the rest and subscribe at the link below 👇

mikebifulco.com/newsletter/s...

It's a notion that has stuck around for ages. It's simple, easy to remember, and easy to test.

But here's the thing: counting clicks **sucks** as a usability metric.

You may have heard of the "3-Click Rule" - this is the idea that users should be able to find what they need to do in 3 clicks.

Your users won't notice steps they're taking if each one is obvious and friction-free. They do notice when they’re lost, frustrated, or overwhelmed.

The result has been a stable and performant app, which has been much easier to maintain. We are happily on Expo 53 now, and I don't think we'll have much problems upgrading to expo 54 when it goes live in the next few weeks.

I ended up rewriting most of the app that had to do with navigation and routing, and performance improved dramatically.

We also had a good number of useEffect calls fighting each other, and several poorly scoped contexts that are now gone.

Seeing some activity on this original post, so I figured I would clarify based on what I've learned about react native since I posted this:

The vast majority of our problems were due to a fundamental misunderstanding of how react native's routing works.

mikebifulco.com/newsletter/o...

Buying a new house? If you've got ~25 pick from in your market, look at 5 and then put in an offer on the next great one that comes along.

Today's newsletter is the direct result of one of these rabbit holes: the Optimal Stopping Problem gives us some really easy rules to decide how many samples to evaluate when trying to balance time vs quality.

I really enjoy learning about statistics rules that feel like they shouldn't work: There's loads of examples in stats that seem _too simple_ or go against intuition completely, and yet -- math.