how I got 25K+ users for my math puzzle app — with zero paid ads

Arin Agrawal — Sun, 12 Jul 2026 21:08:08 +0000

how I got 25K+ users for my math puzzle app — with zero paid ads

I didn’t know what ASO was. I just wanted to see if the idea worked.

I published CalcQuest in my first semester of college.

I didn’t know about ASO optimisation, store algorithms, or organic acquisition funnels. I didn’t have a launch strategy. I had a puzzle mechanic I hadn’t seen anywhere on the Play Store, and I wanted to know if it actually worked — so I asked my friends to download it and tell me.

That was my launch on 3 Sept 2024.

the first thing that actually moved the needle: ratings, not downloads

Somewhere in the middle of building, I started paying attention to what the top math puzzle games were doing — not just the games themselves, but how they marketed. I noticed something.

Ratings compound in a way ads never will.

A 4.9★ app with 800 ratings is a different product than a 3.7★ app with 50,000 downloads. The store treats it differently. Users trust it differently.

I stopped being obsessed with download counts and started nudging users — the ones who were clearly invested, who kept coming back — to rate CalcQuest on the Play Store.

Not everyone. The ones who were already in it.

The idea was:

Ask for ratings after delight, not before it.

That shift changed how the app felt in the store. And at some point, CalcQuest started appearing in the Explore tab. In recommendations. I didn’t pay for that placement. The algorithm just started treating it like something worth surfacing.

i didn’t publish on iOS for 1.5 years — on purpose

The App Store costs money to enter. The Play Store doesn’t. As a college student with no budget, that mattered.

So I made a decision: I’d wait until the Android numbers gave me a reason. I wanted a threshold — real traction, not just hope — before I spent the money on iOS.

1.5 years in, we had it. We launched on iOS.

That delay forced something useful: I had to get really good at the Play Store before worrying about a second platform.

Everything I learned about optimisation, about what made the listing convert — I brought all of it to the App Store when we eventually got there.

i realized blogs that weren’t really blogs

I was inspired by how Duolingo does content. They don’t write ads. They write articles about language learning — and Duolingo just happens to be in them.

So I started writing about:

Mental maths
Puzzle types
How to actually get better at arithmetic

Not "download CalcQuest" — just useful stuff about the thing CalcQuest is built around. And inside those posts, naturally, CalcQuest shows up.

Blogs worked like:

How to improve mental math
Strategies for solving puzzle types
etc.

Two things happen when you do this:

Readers find you through search.
Google starts understanding what your app is actually about.

The blogs and the Play Store listing started reinforcing each other.

When I published a major CalcQuest update, I'd also share the behind-the-scenes on my main Twitter and Reddit accounts. Not just to build in public — though that's part of it — but because it helps Google index properly and understand that CalcQuest is a real, maintained product.

the number that surprised me the most

After a while I opened MixPanel and looked at where my users were actually coming from.

Not India.

Not where I expected.

The top markets were:

Brazil
USA
Indonesia
India

I didn’t localise for any of them. I didn’t run geo-targeted campaigns. Nothing.

The reason, I think, is simple:

CalcQuest is a numbers game. Numbers don’t have a language.

The puzzle mechanic works the same in Portuguese as it does in Hindi.

That’s not a strategy I planned — it’s just what happens when you build something where the core experience doesn’t depend on text.

The game spread because of what it wasn’t, as much as what it was.

40+ releases before I started thinking like a product person

Around versions 40–45, something shifted in how I thought about CalcQuest.

I stopped trying to build for "everyone who likes games."

I got specific.

People who like quick brain challenges.
Students who like numbers.
Casual gamers who are bored of hyper-casual clones.

That’s it.

That’s the whole audience.

The app itself got ruthlessly focused around one simple problem:

Make math feel addictive.

Fast onboarding
Satisfying feedback
Short game loops
Progressive difficulty
Clean UI
No ads interrupting the experience

The whole thing stays under 30 MB.

That’s not an accident — that’s a product decision.

When you know exactly who you’re building for, everything else gets easier.

What features to add.
What to remove.
What to put in the store listing.
What to write about in blogs.

Clarity isn’t just a philosophy thing. It’s actually a distribution strategy.

what I still watch

I don’t check download counts every day.

What I actually get excited about is opening MixPanel and seeing returning users.

People who came back.

People the puzzle actually held onto.

That’s the real number.

Not how many people found the app — how many people stayed.

I’m still figuring out retention.

Still shipping updates based on what the data shows me.

Still keeping the app under 30 MB.

Still writing blogs about mental maths that happen to mention a puzzle app I made.

25K users. Zero paid ads. 150 countries.

I don’t say that to flex.

I say it because when I was a first-semester college student who just wanted to test an idea with his friends, I had no idea any of this was possible.

Most of it happened slowly, then all at once.

Ping me on LinkedIn if you need startup advice, tech help, or emotional support after your server bill arrives.

— Arin

P.S. CalcQuest is free on Android and iOS.

Give it five minutes.

You’ll either love it, or you’ll have opinions.

Both are useful to me.

DEV Community: Arin Agrawal

how I got 25K+ users for my math puzzle app — with zero paid ads

how I got 25K+ users for my math puzzle app — with zero paid ads

the first thing that actually moved the needle: ratings, not downloads

i didn’t publish on iOS for 1.5 years — on purpose

i realized blogs that weren’t really blogs

the number that surprised me the most

40+ releases before I started thinking like a product person

what I still watch