Abstract
HairApp lets you snap a selfie and preview a new haircut or hair color with AI. The reviews from the last six months are almost unanimous on one thing: the advertised "free trial" charges your card immediately, the free tier is one hairstyle and one color, and the AI often ignores the style you actually picked. Near-a-million installs say the idea has real pull — the execution is what's driving people away angry enough to write long, detailed reviews about it.
Background
HairApp (HairApp: AI Hairstyle Try On) is a free-to-download app from a small studio (support and terms both point to xlabs.com.tr) that lets you try on hairstyles, colors, makeup, and a beard filter using your own photo.
What it does
Selfie in, AI-rendered hairstyle/color/makeup preview out, plus a face-shape "color analysis" and a celebrity-lookalike novelty feature.
What users report
Signing up for the advertised week-long free trial triggers an immediate charge, and only a single preset style and color are viewable without paying at all.
Why it is soft
The complaints aren't about a missing feature — they're about being billed for something sold as free, on top of an AI result many say doesn't match what they asked for.
Strengths
The concept clearly has demand — over 900,000 installs is not a small number for a niche beauty utility, and that pull exists despite a wall of one-star reviews, which says people want to try a haircut on their own face before committing more than they want this specific app. The core pipeline also works end to end: everyone who opens the app gets a selfie captured and at least one style rendered, even the reviewers furious about the paywall. That's a functioning funnel a competitor doesn't have to invent from scratch, just fix what's inside it.
Market gap
The complaints cluster into a few repeated patterns, most common first.
Bait-and-switch billing
By far the most repeated complaint: users say the app is advertised as a free trial and charges their card the moment they tap it, in currencies from GBP to PHP to NGN, with refunds frequently refused.
Nothing to try before you pay
The free tier is one fixed haircut and one color — not enough to judge whether the app is worth paying for, which several reviewers say is the actual problem with the trial.
Renders that don't match the ask
A recurring quality complaint: selecting a short style returns long hair, colors render "unnatural," and face-shape detection is described as inaccurate.
Whenever we try to check or view anything, the first thing that comes up is payment. At least some features should be free so users can develop interest and trust in the app before investing money. Your ads show things like checking hairstyles or haircuts, but when we actually try, charges appear immediately. That’s not good marketing at all. You should keep a few features free in the beginning so people start using the app regularly.
Terrible Ai app. Does not give you accurate face shape. I downloaded it because they advertised as if it would show you what you're best suited for, which it doesn't. Then if that doesn't even work I literally selected short hair and it kept giving me long hair. awful app, I can't imagine anybody paying for this.
Every single thing leads you to the free trial page, seems sketchy. Even the 5 star reviews seems suspicious, like a review bot, not one of them has ever written more than 1 sentences, and they all say generic things.
- Let people try 2-3 styles free, no card required. Trust has to be earned before a subscription is asked for — one locked preset isn't a trial.
- Never charge until the trial period actually ends. State the exact charge date and amount at signup, and honor it — this single fix removes the complaint that dominates the reviews.
- Make the AI respect the exact style and color selected. Users can tell when the output ignores their choice — that's a modeling/prompting problem worth solving before adding more filters.
Build complexity
Estimated for one developer, part-time, no funding, no paid acquisition. The core render is a thin wrapper around an existing AI image-editing API rather than a from-scratch model, which is most of why this is buildable in a month.
| Workstream | Weeks |
|---|---|
| Selfie capture + onboarding | 0.5 |
| AI style/color render + prompt tuning | 1.5 |
| Preset library + before/after UI | 0.5 |
| Honest free tier + Play Billing subscription | 1.0 |
| QA + store listing | 0.5 |
| Total | 4.0 |
Expected revenue
Assumes a fair $9.99/month subscription, no deceptive trial, and organic/App Store search traffic only — no paid ad spend, which is what likely drove HairApp's own install volume, so a solo builder's funnel starts smaller.
| Month 12, monthly | ||
|---|---|---|
| Installs | 50,000 | |
| Activated (tried a style) | 30,000 | 60% |
| Retained (returns within 30d) | 4,000 | 13% |
| Paid | 240 | 6% |
| Revenue | $2,400 |
Biggest challenges
Not the billing fix — that's a solved, well-documented pattern. The harder parts are specific to this app category.
AI quality, not just AI presence
Making the render actually match the selected length/color while keeping the face recognizable is a real prompt-engineering and model-selection problem, not a checkbox — it's the exact thing reviewers say this app gets wrong.
A poisoned category
"AI hairstyle try-on" is now a crowded genre with a reputation problem from apps like this one; without paid ads, standing out in App Store search against lookalikes is the real acquisition challenge.
AI angle
AI isn't an add-on here — it is the entire product. That cuts both ways: a commodity image-editing API (the kind HairApp itself is almost certainly built on) makes shipping a v1 in weeks genuinely realistic, but it also means the only real differentiation available to a competitor is doing that same AI step better.
Without care
A generic image-edit call, like HairApp's, that ignores the user's actual selection — exactly what reviewers say makes the current app feel broken.
With care
Deliberate prompting and a same-photo verification pass (does the output actually match the requested length/color before showing it) turns the single most complained-about part of this category into the new app's selling point, for a few cents more per render.
Conclusion
Build it. Real demand (900K+ installs on a badly-run competitor), a cheap and fast build on top of commodity AI APIs, and a market gap so well-documented it reads like a spec: don't fake the free trial, and make the render match what was asked for.
4 weeks
Time to v1
$9.99/mo
Fair subscription, real free tier
$2,400
Revenue at month 12
No paid UA
Organic-only funnel assumed