5 months in, rewrote the whole app once, finally got my first paying customer
been lurking here a few days and figured id finally write something. reading other peoples first sale posts is basically the only thing that kept me going when nothing was happening.
so. about 5 months ago i started building a desktop app for streamers. the problem its solving is niche. its called stream sniping. some random watches a streamer live, queues into their game, hunts them down using whats on stream. annoying for the streamer, ruins matches, etc. the normal fix is just adding a delay. but OBS makes you restart the entire stream every time you wanna change the delay value, which is dumb if you wanna adjust it on the fly during a session. so i built a thing that sits between OBS and the platforms, buffers the video, lets you change the delay live without dropping. also pushes to multiple platforms at once with the same delay applied across all of them.
easy to explain. way harder to actually build.
the first three or four months were me just losing my mind over resource usage. the app has to hold a full quality video stream in memory and forward it back out, no dropped frames, no re encoding, on a PC thats already running a game, OBS, discord, probably chrome. early builds worked. they were also heavy as hell. eventually i had something good enough and shipped it on electron. handed out free keys to a few streamers i knew. then april 17 someone i dont know paid for a yearly plan. screenshotted the email. probably looked at it 20 times that day.
then i did something kinda stupid. instead of trying to get the next sale, i ripped the whole app apart and rewrote it in rust.
reason was simple. one of the testers told me flat out that the memory usage was annoying. enough that he didnt wanna leave it running. and the entire point of this thing is that you leave it on the whole time you stream. so i ditched electron, rebuilt the UI on tauri, which is rust based and uses the browser engine your OS already ships with instead of bundling a whole copy of chrome. then rewrote the buffering engine in native rust too. installer went from like 120mb down to under 15mb. memory dropped to a tiny fraction of what it was. starts in basically a second.
right call. felt completely insane while i was doing it though. spending weeks rebuilding something that already worked and had a paying customer is not a thing you brag about on twitter. next time ill set a deadline. i let this one drag way past where it should have ended.
current state. 22 users. most are free keys i handed out. a few didnt continue past the trial which kinda hurt. and to be real, this whole thing is brand new to me, so i didnt have any kind of data or tracking set up to actually understand why people were dropping off in the trial. just had to guess. that part is finally in place now and since i hooked it up i havent had anyone churn or leave, but obviously thats a small sample. paying numbers are tiny. MRR is tiny. nothing impressive.
worth saying. i havent really been marketing this thing at all up until very recently. its basically just been word of mouth from a couple friends who stream to decent crowds, somewhere in the 100 to 300 concurrent range, and then their viewers and streamer buddies trickling in from there. the app finally feels like something im not embarrassed about, so im starting to actually put it in front of people. partly why im posting here.
still. its the first thing ive ever made that someone paid me real money for. five months of work, a full rewrite in the middle, a lot of late nights where i was pretty sure i was wasting my life. the money is whatever. the thing that hits different is knowing it exists for someone else now. some guy uses it. its on his PC. wild feeling.
stuff i wish id done sooner:
shipped the ugly version. i polished way too long before letting anyone use it. the feedback loop from real users moved the product more in like two weeks than i had managed alone in two months.
talked to streamers before i wrote any code. i had opinions about what a delay tool should be that turned out to be sort of wrong. wouldve saved me a lot of rebuilds.
set a deadline on the rewrite. the rewrite was right. the timeline was not.
actually marketed it. like, at all. shouldve been reaching out to streamers on kick, youtube, twitter, discord, anywhere they actually hang out, from way earlier. building in silence and hoping people find you is not a strategy. its just hoping.
set up tracking from day one. flying blind on why trial users drop off for months is not great when youre trying to figure out what to fix.
if youre still grinding on something nobody is paying for yet, idk what to tell you other than it does eventually click. no schedule, no guarantee. the work isnt wasted while you wait though. the free users teach you who its for. the first paid one teaches you the thing actually exists.
ill be in the comments. ask whatever. electron to tauri, how the buffer relay works, pricing, trying to sell to streamers as someone who doesnt stream, all of it.