You find a hairstyle you love. You save it. You show it to your barber. You get the haircut.
Then you look in the mirror and think: why does it look so different on me?
That's the problem HairOver is built for.
What HairOver is
HairOver is an AI hairstyle app. You upload a photo, it analyses your face shape and hair, recommends cuts that are likely to suit you, and lets you try a haircut or hair color on your own photo before you commit.
The whole product answers one question: would I actually choose this haircut if I could see it on me first?
Why this problem
A saved hairstyle photo is really a photo of someone else: a different face shape, hairline, hair density and texture. The cut can be perfect and still look wrong on you, because nearly everything that decides how a cut looks is different.
Face-shape guides help a little, but I don't think face shape is a rule. It's a starting point. Texture, density, length and how you style your hair in the morning all matter too. What people want is simpler: show me, on my face.
Why a consumer product, not another client project
Most of my career has been building software for other people. With client work the question is: does it work?
With HairOver the question is: would anyone actually care?
That's a much harder question, and it's the one I wanted to learn how to answer. When you build for a client, someone has already decided the thing is worth building. With a consumer app, nobody has decided that, and you only find out after you ship.
Where HairOver actually is today
As of September 24, 2026, this is what exists:
- An Android app on Google Play. It started life as Stylora, and you can still see that name in the first version's screenshots.
- iOS is planned but not live. The App Store link on the website is still a placeholder.
- A marketing site at hairover.app, built from scratch over about a month (how I built it).
- A pricing model I've pulled apart and rebuilt on paper (the unit economics).
- A written 30-day distribution plan for X (the plan).
What's not in that list: downloads, revenue, retention. I'm not skipping them because they look bad or good. They'll get their own posts, with the real numbers, once I can publish them properly. I'd rather have a gap in this series than a made-up number in it.
The $1M MRR goal
The long-term goal I'm writing down is $1M in monthly recurring revenue.
I want to be precise about what that is: a target, not a forecast. HairOver hasn't reached it and isn't close to it. I'm writing it down because a goal that big forces useful questions early:
- What does a generated image cost, and can a subscription carry that cost at scale?
- Where does distribution come from if nobody knows the app exists?
- Which part of the product makes someone come back?
A small goal lets you avoid those questions for a long time. A big one doesn't.
Why build in public
Three reasons, in order of honesty:
- Accountability. It's harder to quietly drop an experiment when you said in public that you were running it.
- Distribution. Nobody knows HairOver exists. The founder story is one of the few channels I have that costs nothing.
- A record. Most startup stories get written after the result is known, and then they get rewritten to fit it. I want the version written before I know how it ends.
That last one is the reason this series exists. Some posts will end with "this worked." Some will end with "this didn't." Plenty will end with "I don't know yet."
What I'm testing next
The next thing I need to learn isn't technical. It's this: when a stranger uploads a photo, do they reach a generated look they care about? And after that: do they come back, and will they pay?
I don't know yet. That's what the rest of this series is for.
Key takeaways
- HairOver shows you a haircut on your own face before you get it.
- The problem is that saved hairstyle photos are photos of someone else's face and hair.
- Today: Android on Google Play, iOS planned, a marketing site live, pricing and distribution plans written down.
- $1M MRR is a target, not an achievement. Real metrics get published as they come in.