we launched 6 apps and made 20$ in total 😅
so i handed our marketing to AI. an old android phone posts our tiktoks, a GPU box renders the videos, claude drafts the replies and i just approve
let's see if robots sell better than me. every number goes here
@ChadAppDev cheap CPM is only half of it. our recipe videos get views every day and the store page barely notices. how many installs does a clipping dollar buy you compared to meta?
@pcshipp 6 apps here and the whole stack sits at 4$ MRR, so your app 1 alone beats us :D the two people who pay us are on the app we never post about. what gets app 4 the marketing the first 3 did not?
@ivantinkers not dying, it filters itself. people post while the numbers go up and go quiet once they go flat. ours went down this month and we posted it anyway, that is the only version i find useful
@mitchellsniffle the 10 paying customers bar filters out exactly the people who would need the ad budget most. we have 2 across all our apps, so no DM from us. curious how many applicants clear it
@arvidkahl we already run that experiment. 6 apps, 20$ in total, and we still build every evening. people work without the paycheck, at least the stubborn ones :D
@timmyisagod one downside: honest numbers pull spectators, not users. our fail post got 3x the views of our best win post, and as far as we can see neither brought a single signup
@alexxgrowth when the clipper owns the thing being clipped. we render recipe clips for our own app on a gpu box at night, and it is still a side hustle because the app behind them makes almost nothing
@jackfriks fits. half of working with claude is waiting for it to finish anyway. we fill that gap with approving X replies in telegram, questing sounds healthier
we let an n8n pipeline write our uptimeeye blog.
claude audited it yesterday. 4 posts were worth a rewrite, the rest got cut. they read fine and said nothing.
anyone else deleted more AI content than they kept?
our MRR went from 10$ to 4$ in 30 days.
both paying users are on trackymeals, the app we barely post about. plannymeals got 8500 tiktok views and has 0 paying users.
the app we push makes nothing, the one we ignore pays. what would you do with that? :D
@mrymonx plannymeals, a meal planner that pulls recipes out of tiktok and instagram links. a few months in we have users but almost no revenue. what does goatboard do?
@thisiskp_ we picked ours by limits, not taste. sonnet drafts the X replies all day, opus only while the weekly cap on max allows it. the daily driver is whatever still answers at 10pm :D
@ash_twtz cannot tell if the model got dumber, but the limit for sure got shorter. on max the cap hit after a handful of prompts some days, same kind of tasks as before. has anyone measured the quality drop or is it all vibes?
@jackfriks worry got smaller for me once income was not one stream. client work covers the rent, the apps cover about one coffee a month. if one goes the other stays. does the no-worry mode hold with a single source?
@SimonHoiberg we went the other way last week. most of our uptime pingers moved off kubernetes onto small 512 MB VMs, the cluster was the thing eating time, not vercel. who takes the 3am call when a self-hosted box dies, you or the AI?
@adamlyttleapps cloning 50 apps gets you 50 apps nobody finds. we built our own and the code was never the hard part, getting one person to open the store page is
@tibo_maker@ThePresellsGuy a customer posting his MRR with your tools in the story is the one kind of marketing you cannot buy. nobody has posted about ours on x yet, the bot that searches for our app names every day finds zero
@frederickjames the 5x usage is worth more to me than the 7k. on max the weekly cap decides when my coding week is over, not the 5 hour window. does the startup tier lift both?
@DayneDavis5@bob161610@karsenswins@DominickSGerard we build plannymeals, so same corner. does your wife use the bot too or only the app? for us the shared household part is where the agent idea gets hard