ISB AI Founders

The product loop

Pick an app you open every single day. When did it last change on you?

Loops, not lines

Every app on your phone runs this loop, forever. Today, so does yours.

Exhibit zero · the app in your pocket

2012: no streaks, no leagues, no owl with a personality. Just short lessons and a login screen. A product with no loop running yet is just... a line. It grew slowly.

Same app, fourteen years of loop turns

None of that existed in 2012. The loop never stops, even 14 years in.

Fourteen years of loop turns bought them this

The owl earned that ad budget one shipped feature at a time.

But every loop starts somewhere

Before Duolingo could loop its way to today, someone shipped an ugly, wrong first guess. Here is the most famous wrong first guess in an app on every one of your phones.

Exhibit A · San Francisco, 2010

Two founders, Kevin Systrom and Mike Krieger, built a check-in app like Foursquare: points for hanging out with friends, plans you could make, badges, AND photo sharing bolted onto the side.

What the logs actually showed

The founders watched real usage, not opinions, and found one feature buried in a cluttered app doing all the work: a photo, a filter, a caption.

The cut

Camera, filter, post, comment. Renamed Instagram, relaunched October 2010: 25,000 signups on day one, over a million users within two months. That is a zoom, not a persevere. Keep that word.

Exhibit B · one more, for good measure

In 2005 the founders built "Tune In Hook Up": upload a video describing your type. Nobody dated. People just uploaded random videos of pets and parties, like this one, the actual first video ever posted to the site.

19 seconds, a guy, an elephant enclosure. The dating site never happened. This did.

Same story, twice

Burbn showed Instagram. A dating site showed YouTube. Neither founder was smarter than you. They just trusted behavior over opinion.

Real people, real behavior

Two teams, same testers

Your turn

Persevere, zoom, or pivot. Every founder faces this exact decision. Today, so do you.

You first

The decision

Option 1 · Persevere

Only choose this if you can show the exact line that proves it. Team StudyBuddy: their gauntlet log showed 4 of 5 strangers understood the flashcard app in under 10 seconds, and 3 asked to keep using it after the test. Verdict: persevere, go deeper on what already works.

Option 2 · Zoom

Team ChoreQuest: a chore app with points, a rewards store, a calendar and a leaderboard. Every tester ignored all of it except the daily checklist and the streak counter. Verdict: zoom, kill the store and the calendar, make the checklist and streak the whole app.

Option 3 · Pivot

Team CafeteriaFix: a lunch pre-order app to kill the line. Testers shrugged at pre-ordering but couldn't stop using the food-rating feature added as an afterthought. Verdict: pivot, the real product is a cafeteria review app, and they already built half of it.

Two teams, same data

Not all confusion is equal

Your core action is the one thing your product must let someone do. A confusion that stops that is a 5. A typo in the footer is a 1. Rank first, fix in that order.

Same log, sorted correctly

"I couldn't find the checkbox" blocks the core action: rank it first. "The owl mascot's arms look weird" is real feedback, and completely irrelevant to Saturday. Rank it last, or not at all.

Exercise 1 · setup

You've been collecting evidence since yesterday afternoon: confused strangers, family objections. Claude is about to cluster all of it into patterns and rank it by pain, exactly like you just watched with ChoreQuest. Get your notes open now.

Copy this prompt · run it now

Here are observation notes from users trying my product. Cluster the confusions into patterns, rank them by how much they block the core action, and give me a fix list of at most 5 items. Then tell me: does this evidence say persevere, zoom in on one feature, or pivot? (Day 2 you graded the PROBLEM with REAL/WEAK/DEAD. Today you grade the PRODUCT. Same honesty.)

Watch it once, live

Exercise 1 · debrief

"I ranked the checkbox confusion first because 3 of 5 testers got stuck exactly there" is a defended ranking. "I just moved it to the top" is not. Disagreeing with Claude's order is great, as long as a quote backs the disagreement.

Ten minutes, wired for good

You've been reading confusion logs all morning. Now you give your product a memory: a simple count of what people actually do, every day it's live.

Three numbers, nothing more

1) how many people completed the ONE thing your product is for. 2) where people quit before finishing it. 3) how many came back a second time. That's it. Resist counting everything just because you can.

Copy this prompt · run it in Claude Code

Add a simple counter to my app: every time a user completes [my core action], log it with a timestamp to a small file or table. Show me the running total somewhere I can check any time, like a console log or a tiny admin page. Do not track anything else yet.

You first

Say it out loud

Exercise 2 · setup

Not a group vote, not a staff opinion: your team's read of your evidence. Public declaration is the point. Once you've said it to the room, it's real, and the room remembers who was honest.

The unpopular truth

A pivot means you built real evidence, read it honestly, and had the guts to act on it with hours on the clock. That deserves louder applause than persevere ever will.

Checkpoint

Every team declared: persevere, zoom, or pivot. Backed by a quote, not a feeling.

“Do things that don't scale. Charge for it.” Today you proved it.

Go deeper · optional

All three are on the portal too.

This afternoon

Next up: brand, deck and the launch video. Then role-split tracks, then the first-dollar hunt, real people, real yuan, real bell.