From 1,000 to 8,000 downloads and $35 to $280 MRR: 5 months of growing an AI calorie tracker with $0 ads

Hey Indie Hackers,

Back in April, I shared how Caldef, the AI calorie tracker I built to fix my own health, hit 1,000 downloads and $35 MRR in its first two weeks. Many of you left brutally honest feedback, and I took it seriously. Here's the 5-month update.

The Numbers

8,000+ downloads (up from 1,000)
$280 MRR (up from $35)
130 active subscribers, a mix of monthly and yearly (up from 14)
~1.6% of downloads are active subscribers (up from 1.4%)
$0 spent on ads

Quick recap for new readers

Not long ago, I was obese with high blood pressure. My doctor told me that if I didn't change my lifestyle, I'd be on medication for the rest of my life. I built Caldef to track what I was eating, lost 13kg in about two months, got my blood pressure back to normal, and then put the app on Google Play.

The uncomfortable insight

Downloads grew 8x. MRR grew 8x. That's not a coincidence.

Almost all my growth came from reaching more people, not from converting them better. My conversion rate barely moved from 1.4% to 1.6%. Right now, each download is worth about $0.035 in MRR.

If I want another 10x in revenue, I either need 10x the downloads or a better funnel. Fixing the funnel is the cheaper path, so that's my focus now.

The revenue math nobody talks about

130 subscribers and $280 MRR works out to about $2.15 per subscriber per month. That's lower than my ₱179 (~$3) monthly price suggests, for two reasons:

Yearly plans are discounted and spread across 12 months
Google takes its cut

I'm okay with that. Caldef is priced for the Philippines and other emerging markets, so I'll never have MyFitnessPal-level revenue per user. The trade-off is that yearly subscribers pay upfront and don't have a chance to churn every month.

What makes Caldef different

AI logging that understands real food: Type "two cups of rice, grilled tilapia, and sinigang," or snap a photo of your plate, and get the full nutritional breakdown in seconds.
Cultural intelligence: It handles Filipino and other non-Western dishes that traditional food databases ignore.
Localization: Onboarding asks for your country and suggests meals you actually recognize and cook.
Price: ~$3/month instead of ~$20/month, a price that works in emerging markets.

Growing beyond one channel

In my first post, I said relying only on Facebook groups was fragile. Since then, I've expanded into Reddit, TikTok, and my own website at caldefapp.com.

The lesson that hasn't changed: the personal story beats everything. People don't want another corporate wellness app. They want a tool built by someone who was in their shoes.

What's working vs. what's not

✅ The AI logging hook still sells the app. People get it in one try.
✅ My personal story still outperforms any other marketing angle.
✅ RevenueCat is still flawless for subscriptions.
❌ Conversion is stuck around 1.6%. I've grown the top of the funnel, but I haven't cracked what makes someone pay.
❌ Google Play review has been painful. I've hit multiple rejections over permissions and subscription terms mismatches. Read the policies twice before you submit.

The Stack

Frontend: React Native (Expo)
AI: OpenAI GPT-4o-mini
Backend/Auth: Supabase + Google Sign-In
Payments: RevenueCat

My questions for the IH community

My conversion has stayed around 1.5% from 1,000 to 8,000 downloads. For mobile subscription apps, what actually moved your install-to-paid rate: a hard paywall, a shorter trial, showing the paywall during onboarding, or something else?
For those running low-priced subscriptions in emerging markets, how did you raise revenue per user without pricing out your core audience?
How do you keep habit and health app users engaged past the first month?

Brutally honest feedback is welcome again. It helped a lot last time.

👉 App: https://play.google.com/store/apps/details?id=com.cedricdev.diettracker
👉 Website: https://caldefapp.com/

Thanks for reading. Let's build.