I keep saying the goal for HairOver is $1M in monthly recurring revenue.
A goal like that is easy to say and easy to hide behind. So this post does the unglamorous part: it works out what $1M MRR would actually require, using HairOver's own numbers.
What happened
When I rebuilt HairOver's pricing, I ended up with three planned subscription prices:
- Weekly: $4.99
- Monthly: $14.99
- Yearly: $59.99 (about $5.00 a month)
And one hard cost: about NPR 15 per generated image, roughly $0.10.
Those four numbers are enough to see what $1M MRR would mean.
How many subscribers is $1M MRR?
Divide $1,000,000 by what one subscriber pays per month:
| If everyone was on… | Per month | Subscribers needed |
|---|---|---|
| Yearly ($59.99) | ~$5.00 | ~200,000 |
| Monthly ($14.99) | $14.99 | ~67,000 |
| Weekly ($4.99) | ~$21.62 | ~46,000 |
In reality it would be a mix. But the yearly row is the one I pay attention to, because the pricing plan is designed to push people toward yearly. If most people choose yearly, $1M MRR means something like 200,000 paying subscribers.
That number makes the goal concrete. Not easier, but concrete.
The milestones in between
Using the same ~$5/month yearly price, here's what each milestone means:
| MRR | Paying subscribers (at ~$5/mo) |
|---|---|
| $100 | 20 |
| $1K | 200 |
| $10K | 2,000 |
| $100K | 20,000 |
| $1M | 200,000 |
The first row is the one that matters right now. Twenty people paying every month would be the first real proof that the product is worth money to strangers. Every row after that is a harder version of the same question.
What installs would that take?
This depends on a number I don't have yet: what share of people who install actually pay.
So here's an illustration, not a measurement. If 2% of installs became paying subscribers, 200,000 subscribers would need about 10 million installs. At 5%, about 4 million.
The point isn't the exact number. It's that at $1M MRR, the question stops being "can I get users?" and becomes "can I get millions of them at a cost that makes sense?" That's a distribution problem much more than a product problem.
The cost side at scale
This is where HairOver's economics get tight.
The yearly plan includes up to 25 generations a month. At NPR 15 per image, a subscriber who uses every credit costs about $2.50 a month to serve.
After the store takes its cut, a yearly subscriber brings in roughly $4.25 a month. So a heavy user leaves about $1.75 of gross profit. That's why the yearly plan can't be generous with credits at today's cost per image.
Scale that up. At 200,000 subscribers, if everyone used every credit, generation alone could cost up to about $500,000 a month. Real usage will be lower than 100%. How much lower is something I'll only know from real data.
The biggest lever I have here isn't price, it's cost. If the cost per image dropped from NPR 15 to around NPR 6, the same plans could include roughly twice as many generations at the same margin. Cheaper previews, caching repeat results, and saving the expensive model for final images are the ideas I'm looking at. None of them are shipped yet.
One more thing to check at this scale: app store commission terms. The reduced rates I used are tied to programmes and thresholds, so at $1M MRR the fee assumptions need rechecking against the current Google and Apple terms.
What didn't work
Nothing in this post has been tested. That's the honest status. It's a map, not a report.
What I learned
Working through the math changed how I think about the goal in three ways:
- $1M MRR is mostly a distribution problem. A few hundred thousand paying subscribers means millions of installs.
- Cost per image decides what the product can be. At NPR 15, every plan has to be careful with credits.
- The first milestone is the only one that matters right now. Twenty paying subscribers comes before 200,000.
What I'm testing next
- Getting to the first $100 MRR and publishing how it happened
- Measuring real install → paid conversion, so the install estimates above can use real numbers
- Measuring real credit usage, so the cost estimates stop assuming 100%
- Reducing cost per image
When a milestone is reached, it gets its own post. Until then, this post is the plan.
Key takeaways
- At ~$5/month (yearly plan), $1M MRR is about 200,000 paying subscribers.
- The milestones: 20 → 200 → 2,000 → 20,000 → 200,000 subscribers.
- At a few percent install-to-paid conversion, that's millions of installs.
- At NPR 15/image, heavy users are expensive, so reducing cost per image is the biggest lever.
- This is math, not a prediction. The general version is in The Math of $1M MRR.