My store page converts at 69%. It gets 1.6 visitors a day.

Twelve days after production launch. Here's what Play Console just told me.

Store listing visitors: 1.61 per day
Store listing acquisitions: 1.11 per day
Conversion rate: 69%

Total installs: 37
Active devices: 25.3
DAU: 5.82
MAU: 15.5
DAU/MAU: 37.5%

Play Store conversion averages somewhere around 20-30% depending on who you ask. Mine is 69%. Nearly everyone who lands on the page installs the app.

And 1.6 people land on it per day.

The funnel isn't broken at the bottom. There's just nothing pouring in at the top.

The retention number that isn't what it looks like

DAU/MAU of 37.5% is high for a utility app. People who install it open it repeatedly.

Except MAU is 15. Three people changing behavior swings that number by 20 points. It's directionally interesting and statistically meaningless. I'm noting it, not celebrating it.

I lost 12 users and it was the right thing to happen

To clear Google's closed testing requirement I recruited 33 testers through a Korean community where people mutually install each other's apps. It worked. I got to production.

On July 4th, right after the closed track ended, active devices dropped from 35 to 20 in a single day. They uninstalled. Their obligation was over.

Here's the part I didn't expect: ad impressions kept climbing. 83, then 87, then 97 over the same window. Fill rate held at 100%.

Because those 33 people were never opening the app. They installed it, satisfied the requirement, and left it dormant. Removing them changed nothing except my dashboard.

The graph recovered to 25 within a few days. New users filled the hole.

What looked like collapse was actually the data getting honest for the first time.

The one country I never marketed to

Installs by country, most recent week:

South Korea: 21
Brazil: 3
Nigeria: 2

Korea is my own audience — Korean community posts, Korean blog. Brazil is where I live. I write in Portuguese, I post to Brazilian dev communities.

Nigeria I have done absolutely nothing for. No posts, no translations, no outreach. Those two people found the app through Play Store search.

Nigeria has one of the highest remittance corridors in the world and some of the worst off-ramp availability for stablecoins. The problem I built for is more acute there than in either country I've been marketing to.

Two people. But they came for the right reason.

Where I actually am

Web, last 7 days:

Sessions: 169
Referral: 50 — solo.xin 22, Indie Hackers 15, rest 13
Direct: 104, unattributed, my own fault
Organic search: 7

app_download_click: 16
official_quote_click: 13
provider_click: 13

App:

Installs: 37
Active: 25
Revenue: $0.07

Seven cents. Ninety-seven impressions.

Conversions on my referral events: I genuinely don't know yet. I haven't checked the exchange partner dashboards. That's the number that decides whether this is a business or a portfolio piece, and I've been avoiding it.

What I'm learning

Optimizing the wrong end. I spent days on the app itself — layout, logo system, off-ramp channel data. The store page already converts at 69%. Nobody's seeing it. I was polishing the part that worked.

Distribution effort doesn't map to distribution result. I posted across roughly 15 communities in 6 languages. Two of them produced almost all of the attributable traffic: solo.xin and Indie Hackers. The other thirteen produced approximately nothing I can measure.

Vanity metrics feel like progress. Installs went up. Then 12 people left and impressions went up anyway. I'd been watching the wrong graph.

The market you didn't pick may be the right one. Nigeria found me. I didn't find Nigeria.

Next

Fix the top of the funnel, not the bottom.

ASO. The store page works. It needs traffic.

Look at the corridors where the problem is worst, not the ones where I happen to live and speak the language.

Actually check the referral dashboards. Either it works or it doesn't and I should know.

Twelve days in. I almost quit yesterday after watching a motivational video about someone who shipped 40 apps and made a fortune. His argument was that if the market doesn't pay immediately, there's no market.

My model doesn't work that way. A user clicks through to a provider, signs up, passes KYC, executes a transfer, and then — weeks later — I get a cut. There is no scenario where day twelve tells me anything.

Three months. Then I look.

As always, happy to share raw numbers if useful.