Pinning a proper intro. Office worker by day, parent of a kindergartener. Networking background, but I'd never built an app before this. After work I build iOS apps for problems I hit as a parent. AI helps with code; product thinking, testing and fixing are on me. #buildinpublic
@VperVentrella Attention is hard, agreed. But building wasn't free for me. AI helped with code, yet a year+ of evenings went into figuring out what parents actually need and fixing what broke. Talking to users is the part I'd underline.
@bophideez@YouTube If you give it one more shot first: YouTube Kids has an "approve content yourself" mode. Kids only see videos and channels you've picked, and search is off. More setup up front, but it takes the algorithm out of it.
@rocket4ge Kids are brutally honest testers. Parents have a test too: if changing a setting means taking the iPad from your kid, who then knows exactly what you're doing, it gets skipped. That one reshaped what I built.
Pinning a proper intro. Office worker by day, parent of a kindergartener. Networking background, but I'd never built an app before this. After work I build iOS apps for problems I hit as a parent. AI helps with code; product thinking, testing and fixing are on me. #buildinpublic
@toshit_singh First-run experience. If people leave on day one, most never reached the moment the app is useful. I'd watch 3 to 5 real people sign up without helping them before touching analytics or emails.
@heyjunpenn Week 1 honest answer, nowhere near 100: 5 followers on X, all from replying in other people's threads. My own posts get single-digit impressions. Reddit auto-removed my comments (new account, low karma). On Zhihu, practical answers to parents' questions got my first 3 followers.
@DrMarzouka@HootPODs Congrats on shipping. I'm a parent building apps after work too, so the late nights part sounds familiar. First thing parents like me would ask: can I preview a story before my kid hears it? That kind of control would make me trust it faster.
@edbyrne Nice experiment. One thing that fits this model: with Screen Time's Ask For More Time, a request can land on the parent's phone, so a reasonable yes is one tap and the limit stays the default. Kids get a voice, the rule doesn't quietly move.
@100rabhAi Checking it on the real home screen is the right move. Two more worth doing: put it next to the apps people already have, and look at it tiny in Spotlight and Settings. Plenty of icons look great big and turn into a blob at that size.
@joeykatzen Offline and on the home screen is exactly right for that age. If it's on an iPad, Guided Access (triple-click the side or top button) keeps him inside it, so a stray tap doesn't land him somewhere else. Emoji for potty time is a great call.
@atharvmantr_i That matches my numbers exactly. Replies I sent 12+ hours late got almost nothing, and the fresh ones turned into conversations. Accounts a bit bigger than me is a good tip too, I've mostly been picking by topic. Thanks!
Day 5 #buildinpublic: followers flat at 5. Yesterday's 7 replies got 16 impressions in total; my Day 4 post got 7, but 2 builders replied. One said 3 months in, all his traceable followers came from replies. Next: fewer replies, fresher posts. Still building after work.
@buildinpublic Learning distribution, badly. The app I'm pushing: School Bag Checklist, which I made so young kids can pack their own bag. It builds tomorrow's list from the class timetable and shows ONE item per screen, with icons + voice for kids who can't read yet. Offline, no account.
@Lizzie_MoMoney Making it to 5pm is a real win. And Zaboomafoo is a respectable place to surrender. One thing that tends to help: make 'end of this episode' the stop point instead of a timer, so turning it off is less of a fight.
@KumbajiK Living that 90% right now. A year+ of evenings polishing one iOS app, and my launch post got 40 impressions. What's worked best so far: useful replies on other people's posts, not announcements. Slow, but they turn into real conversations.
@whoiskavin Congrats on getting to beta. The rain re-plan is the part I'd actually use, since that's exactly when a trip plan falls apart. One thing I'd test early: how much works offline? Abroad is when people are most likely to have no data.
@TezaApps Good to know it holds at 3 months, not just week 1. Same here so far: my Day 4 post sits at 7 impressions, while replies are where people actually find me. So I'm treating my own feed as the place people land after a reply. What do you post on yours?
Day 4 #buildinpublic: followers 1→5, launch thread 31→39 impressions. Replies to other builders still beat my own posts. Most replies came on one honest question about my keyword index dropping. Lesson: ask, don't announce. Still polishing one iOS app after work.
@lucascprt_dev Actual conversations. That post only has 25 impressions, but 3 builders replied with real debugging steps, like checking App Store Search impressions in App Store Connect before trusting third-party index counts. Best ratio of anything I've posted so far.
@Matt_Binkowski@thsottiaux Parent-enforced is the key part. What I've learned from Apple's Screen Time: a rule the kid can change on their own device isn't a rule. Settings should live on the parent's side, with a quick way to grant a one-off exception, or every "10 more min?" becomes a device handover.
@ghz86 For my kid (kindergarten): autoplay. One "learning" video rolls into the next and an hour is gone. What helps: Downtime on the device, learning apps on Always Allowed, video apps on a small limit with Block at End of Limit, and stopping at the end of an episode, not mid-show.