teach, don't entertain

last weekend i wrote a design brief for my own app full of habit loops. streams, points, leaderboards, confetti when you close a deal. i even wrote a whole post about it called build the habit, not the dashboard.

friday afternoon a guest took it apart, and he didn't even know he was doing it.

he runs a math learning platform. it was his second time on the show. in his booking form, before the first episode, he'd written something i couldn't stop thinking about. his company puts learning effectiveness and efficiency ahead of engagement. the goal isn't to entertain. it's to teach.

and consumer edtech that does the opposite, he wrote, is making a deal with the devil that undercuts the value it delivers over time.

we used his line as the episode title. teach, don't entertain.

so i asked him straight.

so engagement-first edtech is a deal with the devil, you said on the last episode, right? so can you walk me through the exact trade a founder makes

his answer was a gym.

imagine you walk into a gym and tell the trainer you don't like to sweat. you don't want to exert yourself. and the trainer says fine, we'll just play some games, at least you'll keep coming. is that better than nothing? sure. a little. but your body doesn't change, because change takes effort and your body fights to stay the same.

learning is the same. watch a video on astrophysics and you'll feel like you learned something. you didn't. you became familiar with it. the knowing only shows up when you pull it out of your own head, solve the problem, struggle a bit.

so the trade is this. you lower the bar for what the customer has to do, so they'll stay. they stay. and they don't get the thing they came for.

that's why i think it's important not to put the engagement before effort.

he pointed to one of the most famous language apps in the world as the example. a fabulous business. streaks, achievements, all of it nailed, a ton of money made. and he isn't sure it's the best way to get people to make progress.

i sat there thinking about my confetti.

so let me push back on myself honestly, because i think we're both right and the line between us is the whole lesson.

habits get people back. that's real. a tool nobody opens can't help anyone, and i still believe that. but if engagement becomes the thing you're selling, instead of the thing that gets people to what you're selling, you've started a slow leak in your own company.

his customer pays for a student who can do algebra. my customer pays for conversations that turn into clients and capital. if my app gets people opening it every day and the pipeline doesn't move, the leaderboard is lying to both of us.

so the test i'm taking from him is simple. every habit loop has to point at the outcome. the confetti goes off when a deal closes, not when you open the app. the streak counts real conversations, not logins. the reward is for effort that moves something.

engagement as a tool. never engagement as the product.

this matters more now than it did five years ago. ai makes it cheap to build anything sticky. you can bolt streaks and points onto any product in an afternoon. which means more founders will be tempted to win on time spent instead of value delivered. and customers will figure out the difference eventually. they always do.

long term, the companies that last are the ones whose customers can point to what changed for them. not how many days in a row they showed up.

more on building with ai without fooling yourself is in the ai book.

so look at the number you check first every morning... is it measuring what your customer came for, or just whether they came back?