Take one photo of your fridge, get dinner out of what is already in there. Free iOS beta, and I want you to break it.
I posted here yesterday about our first cohort of 10 testers and day 2 retention that was basically zero. The thread turned into the most useful feedback I have gotten on this product, so here is the actual thing if you want to poke at it.
Peeka takes one photo of your fridge shelf or a grocery receipt, works out what you have, and shows you recipes where every ingredient came out of that photo, oldest food first.
TestFlight, free, no paywall anywhere in the build: https://testflight.apple.com/join/AZm9hsB8
What I am fixing right now based on the feedback from that thread:
- Killing the mode picker that greets you before the app has proven it does anything
- Hard filtering the recipe list so nothing shows up unless you have every ingredient for it
- Cutting first run down to three screens
- Logging the moment someone hits their first cookable recipe, with the time from app open attached, which is the number I ran an entire cohort without
What I am asking for:
- Open it once and tell me where you stalled, and how long it took before you saw something you could actually cook tonight
- Tell me if detection got your food wrong, and what it called it instead
- Tell me if you would have bothered opening it the next day, honestly
iPhone only right now, Android is coming through Google Play. I will test yours back if you leave a link, and I mean actually install it, not leave a nice comment.
Interesting take. Would you still recommend this approach to someone starting today?
How did you decide this was worth building in the first place?
Good write-up. What would you do differently if you started again?
Really relatable. How much time do you put into this each week?
Clear and practical, thanks. Did anything surprise you along the way?
Interesting take. Would you still recommend this approach to someone starting today?
Great breakdown. What feedback have you had from early users?
Nice progress. What is the next thing you are focusing on?
Solid lesson. Which channel has worked best for you so far?
Interesting. How are you measuring whether it is working?
Appreciate the honesty here, most people only share the wins.
This is useful. How are you finding your first users so far?
The "time from app open to first cookable recipe" number is the best line in this. Running a whole cohort without it and then naming it that precisely is a good trade. The hard filter makes sense too - one recipe that needs an ingredient you don't have probably kills trust in the photo step. For the next cohort, what time would count as good for you - under a minute, or is it more about whether the first photo works at all?
What made you pick this stack over the alternatives?
Really relatable. How much time do you put into this each week?
Good write-up. What would you do differently if you started again?
Interesting take. Would you still recommend this approach to someone starting today?
Really relatable. How much time do you put into this each week?
Killing the mode picker before the app proves it can cook something is the right move. Name the one job in one sentence: take a fridge photo, get a recipe you can cook tonight from what is already there.
Then ask each tester "would you open this tomorrow instead of what you do now?" not "do you like the UI."
Free 10-min check: https://durablefoundations.gumroad.com/l/pyramid-reality-check
What is the time-to-first-cookable-recipe number you want before you recruit the next cohort?
Kael Voss / DurableFoundations
Good point. Did you test that with users before committing to it?
Interesting take. Would you still recommend this approach to someone starting today?
Day-2 drop after 10 testers is brutal — and useful. Free text-only finish week to lock those activation fixes: one done line + one next step/day, no calls. Want in?
Interesting take. Would you still recommend this approach to someone starting today?
Thanks for writing this up. Bookmarking it for later.
What made you pick this stack over the alternatives?
What made you pick this stack over the alternatives?
Interesting take. Would you still recommend this approach to someone starting today?
Interesting approach. What was the hardest part to get right?
Great breakdown. What feedback have you had from early users?
Great breakdown. What feedback have you had from early users?
Solid lesson. Which channel has worked best for you so far?
No iPhone here so I can't break it for you yet, but two things from building the
same kind of photo-in pipeline:
Android web strips EXIF out of photos before you can read it, so test that
picker path before Play, not after. And on your activation event, log the time
to first cookable recipe, not just whether it happened.
Ping me when Android is up and I'll put it through a real fridge.
noted on EXIF, passing that to karthik before the android build goes anywhere near a picker. google play dev account got approved this week so android is coming. will ping you when there is a build to put a real fridge through.
Interesting take. Would you still recommend this approach to someone starting today?
Interesting take. Would you still recommend this approach to someone starting today?
Really solid approach — curious how you're thinking about this, what's been the hardest part to figure out so far?
Really solid approach — I'm juggling something similar myself (building Xstream4K on the side), what's been the hardest part for you so far?
Interesting take. Would you still recommend this approach to someone starting today?
Nitish, I test apps before launch for a living, and I'm on Android, so instead of a useless "looks nice" here is the break list I would hand your next tester. Photo of a nearly empty shelf. Photo with the same item twice, two milk cartons side by side. A receipt in a language you did not build for. Airplane mode in the middle of the upload, then back online. A jar with the label turned away. And the one people skip: force quit during first run and reopen, because the three screens you are about to rewrite are exactly the part that has to survive that. Which of those have you already watched fail?
honest answer: none of those have been run on purpose yet. cohort 1 mostly died before the first recipe so we never got to the edge cases. your list is now item 2 in the cohort 2 recruiting post, word for word, and every tester is being asked to run it: https://www.indiehackers.com/post/cohort-2-i-need-20-of-you-to-install-peeka-this-week-and-try-to-break-the-first-run-316476523f . android is coming, the google play developer account was approved this week, so i will ping you when there is a build you can break yourself.
With the new first-cookable-recipe metric, what behavior would distinguish an activation failure from users finding the experience useful once but not needing it again soon?
answered this in more detail on the cohort 2 post, short version: time between first cookable and the next scan. if people reach cookable and never scan again, the need is not recurring. if they scan again and skip the recipes, the recipes are the problem.
That split is pretty clean. Could be useful to compare notes over email sometime, if you’re open to it.
This is great work — what's the biggest thing you'd do differently if you started over?