Most builders polish endlessly to avoid the scary part, putting the thing in front of real people. Your MVP will never feel ready. Ship it anyway, before the polishing turns into procrastination with better lighting.
You can ship when
The core function works. Users can do the main thing you promised. Test it yourself ten times without an error, then have two people who aren't family test it too. If the core action works eight times out of ten, that's good enough to ship.
It's usable on mobile. Open it on your phone and see whether you can navigate without frustration, whether buttons are big enough to tap, whether text fits the screen. It doesn't need to be a perfect mobile design. It needs to work.
Users know what to do without you explaining it. Show it to someone with no explanation and see if they understand what it does in 30 seconds, and know what to click first. If it takes a five-minute walkthrough, it's not ready.
It loads in under 5 seconds. Open it in incognito mode, which clears your cache, and time it from click to usable. Under 5 seconds is good enough, under 3 is great, over 10 means fix it before you ship.
Nothing actively looks broken. No missing images, no placeholder text, no console errors a user would see, no color combinations that make text unreadable. You don't need pixel-perfect, you need "not broken."
Don't wait for
The perfect color scheme, custom fonts, smooth animations, an empty state for every scenario, advanced filtering, dark mode, social integration, an analytics dashboard, export features, keyboard shortcuts. None of that decides whether anyone wants what you built. Ask yourself honestly: am I improving the core function, or am I procrastinating with polish?
The two-week rule
If you've been building for two weeks or more without showing anyone, you're overthinking it. Set a hard deadline, ship in three days no matter what. Fear of feedback is fear of learning, and it doesn't get easier by waiting.
Before you share your URL
Test it on mobile. Confirm the core action works. Check for obvious errors. Make sure you can explain what it does in one sentence. And if you're a little scared to share it, that's not a sign to wait, that's usually the sign it's time.
Where to ship it first
Start small: five to ten people in your actual target audience, a relevant Reddit community, a Facebook group for your niche, an industry Slack or Discord. Skip Product Hunt until v2, skip paid ads on an MVP, and don't blast your whole email list or chase press before you know the thing works.
The feedback loop
Spend the first three days just collecting feedback, don't fix anything yet. On day four, look for patterns, what are multiple people actually saying. Spend days five through seven fixing the top three issues everyone mentioned. From day eight on, add one requested feature that fits your actual vision, not every idea that comes in.
The three-people test
If three people can use your MVP without help, that's a good foundation. If they come back to use it again, that's real value. If they offer to pay, you're onto something.
If three people get confused immediately, that's a clarity problem, iterate on the explanation before anything else. If they use it once and disappear, it might not solve a real problem. If they say "interesting" and never touch it again, that's politely uninterested, and it might be time to reconsider the idea rather than the design.
The takeaway
Ship when the core works and you're a little embarrassed by the rough edges. That's exactly the right time.