It's easy to feel successful looking at the wrong number.
A post gets a lot of views. The app gets a burst of downloads. The chart goes up. And the business can still be going nowhere, because none of those numbers tell you whether people got value and paid for it.
What this means
Metrics are just a way of asking "is this working?" at each step between a stranger hearing about you and that stranger paying you, again and again.
The trick is to look at them in order. Each step depends on the one before it.
The chain, simply
Here's the chain for a typical app, with what each step actually tells you:
Visitors. People who saw your page or listing. Tells you: your distribution reaches someone.
Downloads / signups. People who tried. Tells you: your promise is interesting.
Activation. People who reached the moment the product is actually useful. Tells you: the product delivers on its promise.
Engagement. How much they use it. Tells you: it's part of what they do.
Retention. People who come back after a day, a week, a month. Tells you: the value lasts.
Conversion. People who pay. Tells you: the value is worth money.
Revenue / MRR. How much comes in, and how much of it repeats every month. Tells you: whether there's a business.
CAC and LTV. What it costs to get a paying customer, compared with what they bring you over time. Tells you: whether growth makes money or burns it.
You don't need all of these on day one. But you should know which one is your weakest right now.
Why 100,000 downloads isn't a business
Downloads are the easiest number to grow and the easiest to be fooled by.
Here's a made-up example, to show the math:
- 100,000 downloads
- 20% activate → 20,000 people actually use it once
- 10% of those come back after a week → 2,000
- 5% of those pay → 100 paying customers
A hundred customers from a hundred thousand downloads. The headline number looks huge. The business is tiny. And if those downloads came from ads, you may have paid for all 100,000 to get those 100.
That's why I try not to celebrate downloads on their own. They're the start of the story, not the result.
What I learned
When I wrote HairOver's distribution plan, I gave every funnel stage one metric so I couldn't hide behind a big number:
| Stage | Metric |
|---|---|
| Attention | Impressions |
| Curiosity | Profile visits |
| Trial | Clicks / installs |
| Activation | First generation rate |
| Retention | 7-day / 30-day retention |
| Advocacy | UGC shares |
| Revenue | Paid conversion |
Real example: HairOver's key metric
For HairOver, the one number I care about most right now is first generation rate: of the people who upload a photo, how many actually see a haircut on their own face?
That's HairOver's activation moment. If people don't reach it, nothing else matters: not retention, not pricing, not marketing. More traffic would just mean more people leaving before the good part.
Retention is my other big question, because haircuts are infrequent. People don't need a hairstyle app every day. So I'm watching whether people come back at all, and what brings them back. I explain why in HairOver's growth post.
I'm not publishing HairOver's funnel numbers in this chapter. They'll go in the HairOver guide once they're reliable, and I'll compare them against the plan instead of picking the flattering ones.
What to do
- Write out your chain from first contact to repeat payment.
- Define your activation moment in one sentence: when does a new user first get real value?
- Pick one metric per stage. Just one.
- Find your weakest stage and work on that before adding more traffic.
- Review weekly for product metrics, monthly for revenue.
Mistakes to avoid
Celebrating downloads. They're the start, not the result.
Tracking everything. Twenty metrics means none of them get your attention.
Adding traffic to a leaky funnel. If activation is broken, more visitors just means more people leaving.
Changing the metric after the fact. Decide what success looks like before the numbers come in.
Ignoring retention. Without it, you're refilling a bucket with a hole in it.
My current thinking
Early on, I think only two metrics really matter: activation (does it work for people?) and retention (do they come back?). If those are healthy, conversion and revenue have something to build on. If they aren't, no amount of marketing fixes it.
For HairOver, both are still open questions. That's the honest status.
Key takeaway
Look at your metrics in order, from visitors to repeat revenue, and find the weakest link. Downloads show curiosity, activation shows the product works, retention shows lasting value, and only revenue shows a business.
Next: Learning From Failure.