The Tip Jar That Became a $90-a-Year Subscription: CardPointers, Run by One Person

Emmanuel Crouvisier built CardPointers in 2019 to solve a problem he had himself: knowing which credit card to use for which purchase, and never missing the offers and credits that cards quietly hand out. It started as a web app, then became an iPhone app, a watch app and a browser extension. Today the paid tier, CardPointers+, costs $90 a year.

The moment that made it a business was almost an accident. In a case study published by RevenueCat, he describes it in one sentence: "My first implementation was a Tip Jar, and I threw in a subscription Pro tier at the last minute, which ended up changing the course of the app (and my life!)."

That same case study records the outcome plainly: within a year, the app was earning him more than his day job. No dollar figure for the app's revenue appears in any source opened for this article, so none is stated here. What is documented is the mechanics, and they are unusually instructive.

A tip jar is not a business model

Plenty of indie developers ship a useful free app and add a way for grateful users to leave a few dollars. It produces a trickle: a nice feeling and no runway.

The subscription changed the arithmetic in three ways at once. It converted one-time goodwill into recurring revenue. It created a reason to keep shipping features, because subscribers renew based on what happened this year rather than what impressed them at install. And it turned the free app into a funnel rather than the product.

The lesson is not that every app needs a subscription. It is that the pricing model was tested cheaply and late, as an add-on, rather than agonised over in advance. He shipped a tip jar and a Pro tier together and let the users decide which one mattered.

The fee decision that keeps 12 points of margin

Crouvisier sells subscriptions on the web, through Stripe, wherever he can, rather than exclusively through Apple's in-app purchase system. In the RevenueCat case study he puts the difference at roughly 3% for Stripe against 15% for Apple under the Small Business Program, and describes the switch to web payments as cutting fees by about 27%.

Builtplain has written about the 15% rate small developers have to apply for rather than receive automatically, and about what the platform charge buys. For an app with an established audience and a lot of returning users, the calculation is stark: twelve percentage points of margin on every renewal, permanently.

There are trade-offs he also names. Apple's checkout converts better for a cold install, since the payment method is already on file. Web payments require the user to leave the app, which loses some of them. What web payments give back, in his words, is control: the ability to offer a discount, extend a trial, fix a billing problem, or run one subscription across iOS and Android from a single ledger.

How he found users without ads

Two channels did the work, and neither cost money up front.

The first was Reddit. An early version posted into the points and miles community, by his account, drew hundreds of upvotes and users. This is the pattern from almost every product in this series: the first users come from a place where the enthusiasts already argue with each other, which is where first customers reliably come from.

The second was a platform gift. When Apple launched watchOS 6, CardPointers was among the first fully independent watch apps and was featured in the App Store's Today tab. Being early to a new platform surface is a real strategy, and it is also the sort of luck you can only prepare for.

The channel he now names as most effective is a revenue share with the people who already publish in the niche: the creators and forum moderators whose audiences are made entirely of credit card optimisers. Affiliate income reached about $1,000 a month, he says. It is the same mechanism Builtplain examined in setting up an affiliate programme on Stripe: you pay only for sales that happen, and the reviewers who would have ignored you now have a reason to write.

The economics of staying small

Two of his statements together describe why this works. His costs are "literally hundreds of dollars per month". And competitors in the same space, he says, need up to 30 people.

A rewards-optimisation app is mostly a database of card benefits, offers and rules, maintained meticulously. Larger companies in the category carry sales teams, partnership managers and content departments. A single developer with automation and a subscription price does not.

His advice on where the time goes is worth quoting for anyone who prefers building to selling: "Success comes from Closing Xcode." He also credits going full time with tripling monthly revenue within two months, achieved by focusing on marketing and messaging rather than features.

What to take from this

1. Build for a problem you have in a niche with obsessive hobbyists. They evangelise, and they read forums.
2. Ship the paid tier alongside the free thing, even as an afterthought. You cannot learn what people will pay for by thinking about it.
3. Do the fee arithmetic before you scale. Twelve points of margin compounds across every renewal for years.
4. Post the rough version where the enthusiasts are. Not a launch, a conversation.
5. Pay the niche's publishers a share. It costs nothing until it works and it buys coverage you cannot request.
6. Keep fixed costs in the hundreds. Low costs are what make a solo operator's revenue into an actual income.
7. Spend deliberate time not coding. The revenue tripled when the attention moved.

The honest caveats

Three. First, nobody has published what this app earns, so the outcome here is one person's statement that it beat his salary inside a year. That is a real fact about his life and not a benchmark anyone can plan against.

Second, the business sits on top of both Apple's platform and the credit card industry's offer structures. Either can change without notice, which is the reason for a dependency audit in a business like this. The web payment option is, among other things, a hedge against exactly that.

Third, the watchOS feature was a genuine break. He was ready for it, having built one of the few independent watch apps at launch, but readiness and a feature slot are not the same thing.

The repeatable core is smaller and sturdier: a specific irritation, a free tool that solves it, a paid tier added early, fees kept low, and the niche's own publishers paid to talk about it.