What RevenueCat became for getting paid, Retain is building for getting kept. Retention and payments in one layer, for any business whose customers come back or do not.
Retention is what we sell. Payment is what makes it complete, because a user who does not come back never pays again. Security is the condition of being trusted with either, so it is built into the structure rather than sold as a feature.
Retain is retention and payments infrastructure for businesses. It measures what users do after signup, identifies the ones drifting from their own usual rhythm, and re-engages them with plays grounded in habit research. Once the behaviour holds, it goes quiet. Checkout, in early access on Stripe rails, ties revenue to the same layer, so you can see not just who stayed but what staying was worth.
The part we care most about is proof. A random slice of every business's users receives nothing, permanently. Whatever difference appears between that group and the rest is the product working, measured. No other retention tool ships this, and it is the reason our dashboard can be trusted when it says something worked, and believed when it says something didn't.
The tools around this problem fall into three groups. Push services send what you tell them. Analytics report what already happened. Enterprise engagement platforms do both, priced for marketing departments with staff to run them. Retain occupies the space between: it decides who needs reaching and when, proves the outcome, and is priced so that a one-person business can run it from day one.
The model is deliberate. RevenueCat became infrastructure by owning one half of the money question: whether users pay. Nobody owns the other half, whether users stay, and staying comes first, because a user who does not come back never pays again. We are building both ends of that layer: the engine underneath, the surfaces users see, and the checkout that closes the loop.
Data. We collect events, never content. No text your customers write, no titles, nothing readable. Users exist as random per-app IDs, and no identity links a person across apps, because no such key exists in the schema. Erasure is one call.
Payments. Checkout rides Stripe rails. We never hold funds and never see card details; the compliance burden sits with the processor built for it.
Behaviour. Frequency caps are enforced in the engine, the taper is mandatory, and because of the holdout, we detect when nudging harms rather than helps, and the dashboard says so. A tool that can prove its own failure is the opposite of a dark-pattern engine.
We publish the SDK source readable at retains.netlify.app/retain-sdk.js. Claims you can verify beat certifications we don't have yet.
Retain is built in Melbourne by a founder who has shipped 76 browser games and works professionally in behaviour change. Games are the most retention-optimised software ever made; their entire craft is what makes a person come back tomorrow. That craft, held to published evidence and measured against a control group, is the product.
Questions, pilots, or scepticism: Jordan.l@hotmail.com
We only earn when your business does. That is the whole incentive structure, and everything below follows from it.
Your data stays yours. We are your processor, never a broker. No user content ever leaves your product, only that events occurred, and nothing links a person across apps, because no shared identity exists in the schema.
We charge on measured results. A random share of your customers receives nothing, so every figure on your dashboard is a difference we can stand behind. Where there is no difference, there is nothing to bill.
The dashboard tells you the truth. If your curve says the product is not working, it says so plainly. An instrument that only reports good news is not an instrument.