A mini program left alone for 8 years — how is it doing now?

November 22, 2026・3133 words・3 min read
Cayden
AuthorCayden

Focused on independent product development, recording bits and pieces of the development process

Welcome to follow WeChat Official Account CayDock

A mini program left alone for 8 years — how is it doing now?

Back in 2018, WeChat mini programs were still new. I wanted to try one, so I threw together a small “days tracker” called Time Odometer (时光里程表) and wrote a post about choosing mini programs. The idea was simple: low-frequency, high-need tools fit the format.

After a small push, traffic settled around 200–300 visits a day. Honestly, for a solo project that felt great — people were actually using it, not just a demo.

The books never worked out, but I couldn’t shut it down

Feeling good is one thing. The numbers didn’t add up.

Backend cost money. Ads made maybe a few mao a day — nowhere near enough, never mind the time. I should have stopped.

Then someone left a message: please keep maintaining it.

That got me. Fine, leave it running. Not like I had to ship every week.

Years went by. “Unmaintained for 8 years” isn’t perfectly true — I still fixed reviews and compliance now and then — but there was no real ops. It just kept running for almost 8 years.

A big reason: LeanCloud. Auth, storage, file upload — barely any self-hosted backend. Passion project economics on someone else’s free/cheap tier.

How the numbers slid

Roughly three stretches.

Right after launch and that small push, daily visits held around 200–300. Best looking part of the curve.

Then I stopped pushing and mostly stopped caring. Organic traffic drifted down. Cumulative users still climbed, but daily actives fell. Expired verification, search limits — harder to find, uglier chart.

As of Nov 2026, cumulative users sit around 63.8k. Yesterday: 30 visitors, 40 opens, zero new users. About 4.2k people added it to “My Mini Programs.” Down an order of magnitude from the peak, but not zero.

Active overview (2026/11/21)

Over the last two months (late Sep–Nov 21), daily visitors averaged about 33 — peaking near 50, dipping into the low twenties. A flat long-tail line, not a growth curve.

Daily visitors trend (last ~2 months)

Why touch it now

LeanCloud is shutting down.

Do nothing and the cloud data goes with it. So the move is simple: no more cloud sync, local-first.

Not because sync is hard — because the cost can’t be covered. Better cut the dependency now than die on a vendor sunset notice.

For me it’s a forced architecture change: BaaS → zero backend, smaller package as a side effect. Already in review; annual verification is in progress too.

Closing

It started as an experiment. Stayed alive because someone said don’t stop. Changing now because the cloud is leaving.

Traffic was never huge. Revenue never mattered. But for a while a few hundred people opened it every day — that stuck with me.

I can’t “maintain” it as product ops. I can at least clean up the dependencies so it doesn’t die with someone else’s shutdown email.

Passion project, cheaper version: don’t run a backend.

Indie development is hard. Especially now that AI has broken the cost of “just building it,” coding isn’t the scarce part — promotion and ops are. Leave a project alone and it won’t die overnight. It fades: daily actives slide, harder to discover, until it’s quiet numbers in a dashboard.

If you’re keeping a side project that once spiked, never paid for itself, and you still can’t delete — I’d like to hear your numbers.

Share:

Leave a Comment

Minimum 3 charactersYour email will not be published. Required fields are marked with *

No comments yet. Be the first to share your thoughts!