Habits Garden: $600/Month, 8,500 Users — the Record of Marc Lou’s Habit App That Never Broke Out

Reading the Numbers From the Side That Didn’t Grow

Case studies skew toward success stories. Stories about hitting $10,000/month get told. Stories about stalling at $600/month don’t. Habits Garden is a rare record where the developer himself published the numbers on the side that isn’t usually talked about.

The maker is Marc Lou, a self-taught engineer based in Bali who has shipped numerous products in short timeframes — ShipFast, Indie Page, Make Landing, ZenVoice, DataFast, and more. Habits Garden is one of them: an app that visualizes habit streaks on a 365-cell grid modeled after GitHub’s contribution graph, with functionality narrowed down to essentially one feature.

The Published Numbers

Starter Story’s listing also cites an initial cost of $19.4K, but that doesn’t square with what he describes himself (stacking existing services, minimal ongoing operating costs). That figure is treated here as a listing-page estimate, likely a later updated value that can’t be independently verified, so only the higher-confidence numbers are kept in the table above.

How It Was Built

The frontend is Next.js and Tailwind CSS with DaisyUI components. The backend runs on Heroku and MongoDB. The mobile app wraps the same web implementation into native iOS/Android apps via Capacitor JS. Image delivery/CDN is on AWS, transactional email via Mailgun, payments via Stripe, and analytics via Plausible.

The design philosophy is thoroughly minimalist: narrow to one core feature, add no unnecessary complexity. No test code, ship to market and confirm with real reactions. That an app used by 8,500 people only took “about a month of actual work” to build is the result of this discipline.

The One Moment the Numbers Moved

There’s exactly one moment where Habits Garden’s numbers clearly shifted: killing the free plan in September 2022, right before launch. By his own observation, this raised the perceived value users felt, and improved the paid conversion rate. It wasn’t a feature addition or a price cut. A change that reduced options is what worked.

The other inflection point was the October 2022 launch. It got featured on Hacker News, driving 12,000 visits in a few hours. But this one didn’t change the trajectory. Set against the cumulative 8,500 registered users, those 12,000 visits that day appear not to have converted into lasting balance. The turning point was on the pricing side, not the exposure side.

What Worked, and the Structure That Didn’t

The reason killing the free plan worked comes down to removing a point of comparison. As long as a free version exists, the paid version competes not against other companies but against “free itself.” In a category like habit trackers where feature differentiation is hard to build, this internal competitor is the strongest one. Remove the free fallback, and the decision simplifies to “use it or don’t.” That monthly revenue is small relative to a base of 8,500 users is consistent with many of them landing on “don’t use it” under this simplified choice.

Multi-channel distribution via Capacitor also worked in its favor. A single codebase spans three distribution channels (web, App Store, and Google Play) and 30% of sales come through the app stores. This is passive inflow via ASO (app store optimization) rather than active acquisition. Capturing this ratio through solo development is not a small achievement.

The payment funnel design is also solid for this scale. That the web version accounts for 70% of sales means a large share of revenue avoids app store commission fees. The 83% gross margin reflects both low baseline costs and this choice of payment channel. Choosing to center the business on the web rather than the app store alone for an indie-developed app appears deliberate.

Meanwhile, the product category itself is what structurally didn’t work. Marc Lou states this clearly himself: a habit tracker is a “vitamin”, not a “painkiller”. Nobody needs it, so retention doesn’t structurally rise. His conclusion: “If you’re aiming for $10,000/month with zero employees, avoid vitamins.”

There are also sobering numbers about the payoff of building an audience. His Twitter following took 8 months to reach 1,000, then another 8 months to reach 20,000. Building in public brought early feedback and early users, but it wasn’t a fast-acting acquisition engine.

Limits and Risks

Habits Garden is currently effectively on hold. He hasn’t decided whether to invest seriously in growth or leave it running at low cost. $600/month at 83% gross margin isn’t a bad outcome in the sense that it won’t run at a loss even if left alone. But it doesn’t justify growth investment either.

If he does restart, his intention is to try partnerships with productivity influencers, the hypothesis being that vitamin-type apps get adopted through third-party endorsement. This is untested.

There’s one more weighty line in his own summary: “Whatever hypothesis I come up with is usually wrong.” So instead of trusting his own intuition, he runs on a system, post 3 times a day, ship one product a month. It’s notable that he’s explicitly given up on trusting his own ability to judge whether a single product will succeed.

How Far Can You Copy This

Three things are copyable: the build approach of narrowing to one core feature and shipping in about a month of actual work, treating the presence or absence of a free plan as something to experiment with. And the technique of wrapping a web implementation to triple the distribution channels. None of these require capital or an audience.

What’s harder to copy is the underlying decision context. Marc Lou was simultaneously running Indie Page ($1,200/month) and Make Landing ($2,600/month) during this period, which is why he could choose “put it on hold” when Habits Garden stalled at $600. Someone betting their livelihood on a single product doesn’t have that option. His follower base is also a function of 8 months × 2, not an asset available right at the start.

And what’s hardest to reproduce may be the willingness to keep publishing numbers that didn’t work out, without retracting them. The $600/month figure has itself become a benchmark for other developers.

Related Reading

• Photo AI — also indie-developed, but a product that landed on the painkiller side
• Carrd — an example of combining a no-new-features design with long-term audience building