Shipping in public has completely evolved. Ideas that used to take months to build and deploy can now be built and shipped in days.
While building several full-stack AI web apps recently, I noticed a massive shift in developer velocity when moving from heavy server infrastructure to client-side WASM & edge computing.
Here are 3 stack decisions that cut our dev cycles by 80%:
Hello guys. I recently published the Version 1 of our app. After publishing I feel like I should have done more with the UI. I m considering genuinely going and re designing the UI and re publishing the app. Maybe I published it earlier than I should have. Or maybe I m just overthinking now that the apps actually usable. Has anyone been in this situation? If so what way did you decide to go about it. Any advice is appreciated. I m considering dedicating 2 full days to relishing the UI.
Building a pricing tracker meant scraping pricing pages. Everyone in the category does it. I checked the terms on the first source, found they didn't allow it, and emailed them instead.
That became 204 sources reviewed one at a time. Terms, robots, whether caching a snapshot counts as redistribution. Some yes, some silence, a few no, which left holes in my data I still haven't filled. Four months gone. Two competitors shipped the same idea in that window by just scraping.
Honest case for: a product selling trustworthy evidence shouldn't be built on data taken without asking. Honest case against: not one user has ever asked me where the data came from.
When you're building for a problem the market hasn't fully named yet, how do you stay convinced you're right and not just stubborn?
Our thesis a year ago: a wave of EU regulation is coming, small companies won't handle it alone. On paper we were nearly alone in the space.
Reality was messier. Two releases, both flops. Wrong channels first, then trying to reach every persona at once, then over-promoting a brand we couldn't articulate simply. And LLMs got better at sounding confident about compliance answers, which for a stretch made us wonder if we were building the buggy whip. (They got better at sounding confident, not at being reliable. That's exactly the gap we still see.)
The space is now crowded, but the gap is still unclosed.
I am pretty sure it's not just me, but this feels weird.
We can literally vibe code an app in a few hours now. Yet I still keep Googling the same tiny utilities over and over.
Not SaaS products. Just simple tools. Things like Markdown editor, JSON visualize, color palette extractor, text formatter. None of these are complicated anymore. Most could probably be built in one evening.
Classic early-stage mistake, sharing it so you can skip it.
We assumed gamification would drive engagement from day one. Badges, progress bars, real-time feedback. We even built a full leaderboard before launch.
Then we stepped back and realized: a leaderboard is a multiplier. It amplifies engagement that already exists.
If the core loop isn't working yet, it just multiplies zero.
I saw a great discussion recently about building a personal brand before the product is even ready vs building the product first and revealing yourself later.
For those of you targeting specific, legacy enterprise problems, which route did you take? It's tempting to build in stealth to protect the core idea and architecture. On the other hand, building a brand (in the sense of building trust) seems critical for those first few partners willing to hand over their messy internal data for you to structure.
If you had to start your current project over, how would you go about it? What have been some of the most unignorable observations centering on this aspect?
