What the stores take, and what comes back
Three chapters, five minutes. By the end you will know the size of the toll on your own book, the margin that returns permanently, and the cash a store settlement queue is holding.
Get the one-page summary, written for your role.
The opportunity
Store billing does more than take a fee — it caps margin, delays cash, and stands between you and the subscriber.
Price presentment, payment method, renewal, receipt — and 15–30%. What reaches you is a payout report, 30–60 days later.
Your checkout, your renewal terms, your receipt, your payout in T+2 — and a subscriber you can reach. The app doesn’t change; the ownership underneath it does.
Apple takes 30% on year-one subscriptions and 15% on renewals; Google Play takes 15% flat. Those rates apply to every renewal, upgrade and additional subscriber — there is no scale tier where the store stops applying. Apple settles up to 60 days after the sale, Google roughly 30, and while the store holds those funds you cannot deploy them.
The third cost is the quietest. Cancellation flows, win-backs, upgrades, dunning, payment recovery and pricing tests all work better on web rails. While subscribers stay on store billing, those plays are store-mediated or need workarounds.
You don’t need to disable in-app purchase to fix any of it. Web-first acquisition starts new cohorts on your rail — the established pattern. Store-to-web migration is the missing layer: it moves the subscribers you already have.
The margin reclaim
A store book loses roughly a fifth of every renewal. On web it loses the processing rate — and the difference holds for the life of the subscriber.
Where $17 comes from. A 50/50 monthly/annual subscriber mix at $10 a month and $100 a year. Per 100 subscribers that is $11,000 billed across 650 transactions — $16.92 each. Monthly-heavy books land lower, which makes the flat per-transaction fee a bigger drag.
Source: the same model the audit runs — Stripe at the published rate for your billing country, plus Stripe Billing at 0.7%. Representative book at $10/mo and $100/yr.
That is +16.6 points of ongoing margin — $2.81 more of every $17 payment reaching your P&L, for as long as the subscription lives. The fee mix changes once; the lift compounds.
One caveat belongs on this comparison. The 22% is not pure toll — it buys dunning, tax handling, receipts, refunds and chargeback cover. Move billing to the web and that work moves with it, so the gap above is the raw rail economics, not the finished P&L: whoever runs that work, in-house or bought, costs something that is not in this figure.5
One rail is deliberate rather than cheapest: the architecture runs on Stripe Billing — the 0.7% above — instead of a proprietary billing engine. That is what keeps the subscribers portable.
The cash-flow release
The part nobody prices in: a 48-day settlement queue collapses to T+2, and the float comes home once.
Not revenue you didn’t have — revenue you already earned, sitting in the stores’ queue. The queue drains in proportion to what migrates: at the 55% base case, $270K comes home once, and for 46 days both streams land at the same time.
Source: Recurr audit model at $5M ARR. Held float is net of the store’s cut, which is why a better store deal leaves more of your money in the queue. Store settlement modelled at 48 days average; Stripe payout T+2 on the United States schedule.4
Act I sized the opportunity on a representative book. Act II shows what the market has already done, runs the model end to end, and then shows how the same arithmetic gets proven on a slice of your own base.
4 Stripe published pricing by market, July 2026.
5 These figures are gross of cost, in two ways. Running billing — dunning, tax, receipts, refunds, chargebacks — moves with the billing and costs something wherever it sits. So does the migration itself: vendor fees, internal engineering and support time, campaign cost. Neither is netted here. The audit nets both against your own numbers.