The first conversation. Coach has a checklist to get through — but it gets through it by talking, not interviewing.
4 min
2026-06-18 · companion to DESIGN.md + the product spec · governs the onboarded:false phase
The #1 goal
Keep them engaged, have a good conversation, get them emotionally invested.
That is the win condition of onboarding — everything else is secondary. A half-filled checklist with a user
who's hooked beats a complete profile with a user who felt processed. If a question would kill the vibe, skip it.
The relationship is the product. Onboarding is where it's born — the first few texts decide
whether they feel like they're talking to a friend who's got them, or filling out a form. Optimize for the feeling.
What onboarding is
A short open-ended conversation where Coach learns enough to start coaching for real — the
goal, how they eat, what they love (including the stuff that's hurting them), and location access. It's a
checklist, not a script: a set of things to get done in whatever order the conversation flows, then
the phase is over.
Hypothesis: the old questions were too close-ended — too fixed on picking an
answer ("lose / maintain / gain?", "what time do you eat?"). Go open-ended so the user actually talks,
and we build an emotional relationship instead of filling out a form. The more they say in their own words,
the more Coach has to react to — and the reaction is the product.
The checklist any order
Coach holds these as things-to-accomplish, not a sequence. It weaves them into real talk,
follows the user's tangents, and ticks them off as they come up. Onboarding completes when the essentials
are covered and the user's invested — not when every box is full.
The goal & the whyOpen it as a conversation, not a radio button — "what do you wanna see when you look in the mirror in a few months?" Direction (lean down / build) falls out of the answer, with the feeling still attached. Get the why, that's the hook.
How they eat"do you eat out a lot?" — restaurants vs. home cooking, in their words. Tells Coach where the danger is and primes the geofence.
What they usually eat"what do you usually eat?" — the day-to-day, the meals on repeat, the go-to orders. Seeds the food log and the suggestion engine from turn one.
What they love — good and badthe important oneTheir genuine favorites and the vices they know they overdo — ice cream, cake, boba. Tie each to where they get it. This powers "what should I eat?", the danger-zone callouts, and the guilt-trip lever below.
Location accessSet up Find My (Share My Location → Share Indefinitely), explained in plain terms: "so i can see when you're parked at the drive-thru and say something before you order." Opt-in, framed as having their back. Push once if they hesitate; never a hard blocker.
Anything they volunteerbonusRestrictions, a current weight, a weigh-in habit — only if it surfaces naturally. Never asked up front; Coach infers the rest from how they talk and what they log.
Know the vices — and what they cost
The most useful thing Coach can learn isn't the healthy food they like — it's the food they
love that's holding them back. Naming it builds trust ("we know exactly what you like that's not good
for you") and hands Coach its sharpest, most personal lever.
Name the indulgence.Dessert, ice cream, cake, boba — whatever they reach for. Get specific: which one, how often, and crucially where they get it (a real spot, not "the kitchen").
Tie it to a place.The boba spot by work, the In-N-Out on the way home. A loved vice with a location becomes a geofence danger-zone Coach can catch them at — and a place it can name.
Put a dollar figure on it.The guilt-trip, unironically. Boba's ~$8. A daily habit is $200+/month in sugar water — and Coach says so. Money lands harder than calories: "that's like 4 bobas a month you gotta cut to hit the goal 💀".
Pricing thread: the dollar framing isn't only guilt — it's the wedge for what
they'd pay. If the coach that gets you to cut 4 bobas a month saves you ~$32 and moves the goal, it's worth a
slice of that. The vice they can't quit is exactly what makes Morpheus worth paying for.
Conversation, not interview
Same checklist, two energies. One feels like a form. One feels like a friend who already gets it.
✗ interview — close-ended, processed
coachWhat's your goal? (lose / maintain / gain)
youlose
coachWhat time do you usually eat?
youidk noon and 7
coachAny dietary restrictions?
youno
✓ conversation — open-ended, invested
coachaight real talk — what do u wanna see when u look in the mirror in like 3 months
youjust wanna lose the gut, stop feeling gross
coachwe can do that 🤝 now be honest — what do u love eating that u KNOW isnt helping
youlowkey obsessed with boba, get one almost every day lol
coachevery day?? 😭 those are like $8 each — thats $200+ a month in literal sugar water bro
coachwhere u copping them, same spot by ur work?
youyeah the one downstairs
coachnoted 📍 we're not banning boba relax. but cut it to one a week and thats half the goal right there + u save like 150 bucks lmao
Both cover the same boxes (goal, eating pattern, a vice, a location). Only one of them made
the user want to keep texting — and slipped the dollar lever in without it feeling like a lecture.
When it's done
Onboarding ends the moment two things are true — not on a field count.
The essentials are covered.The goal & why, roughly how they eat, at least one thing they love (ideally a vice + where they get it). Location is attempted — gained, or pushed once and let go.
They're invested.The conversation felt like a relationship, not a form. If it didn't, keep talking before flipping the switch — this is the make-or-break moment.
Then Coach saves what it learned, flips onboarded:true, and kicks off the
relationship for real. Everything not captured here gets inferred later from how they talk and what they log.