You can pause a subscription or push back its renewal date instead of refunding it, and that keeps most of the money

You can pause a subscription, push its next charge into the future, or hand a subscriber free days, and each of those keeps the customer on your books without giving back a cent you have already collected. A refund does the opposite. It returns the last charge, sometimes a full year of it, and the customer is usually gone anyway. So when someone says they want out or asks for their money back, the refund is rarely your only move, and it is almost always the most expensive one.

Here is the map. Google Play lets a user pause an auto-renewing subscription, and it lets you defer the next charge by anywhere from a day to a year. Apple has no pause button, but it lets you extend a renewal date twice a year, up to 90 days each time, at no cost to the customer. What follows is how each lever works on each store, what it costs you next to a refund, and the one situation where none of them helps.

The choice you actually have when a subscriber wants to leave

A cancellation and a refund are not the same event, and treating every unhappy subscriber as a refund is how studios give back money they never had to. A cancellation stops the next renewal and leaves the money already paid where it is. A refund reaches back and pulls a charge you already booked. Between those two sits a set of store tools built for exactly this moment: keep the person subscribed, or keep the relationship warm, without reversing a payment.

Pause, defer, extend, then refund

Order matters. A pause parks the subscription so the customer can step away and come back. A deferral or an extension grants free time, which turns a would-be refund request into a goodwill gesture that costs you days, not dollars. A refund is the tool of last resort, because it is the only one of the four that takes money back out of your account. Reach for it when the others do not fit, not before.

How to pause a subscription on Google Play

Google Play is the only one of the two stores with a real pause. A user opens their subscription in the Play Store, chooses to pause it, and the store handles the rest. You do not approve it, but you do get to design around it.

How long a pause can last

A user can pause for a period between one week and three months, and the choices depend on the plan's billing cycle. The pause always takes effect only after the current billing period ends, so the customer keeps what they already paid for and the pause starts clean at the next renewal.

What a pause does to access, billing, and your queries

While the subscription is paused, the user does not have access and does not pay the renewal price. Nothing leaves your account, and nothing enters it. One integration detail catches teams out: a paused subscription is not returned by the Play Billing Library's queryPurchasesAsync() method unless you set includeSuspendedSubscriptions to true. Miss that and your app can treat a paused subscriber as fully churned when they are one tap from coming back.

What happens at the end of a pause

At the end of the pause period, the subscription resumes and Google attempts to renew it. A user can also resume manually at any point during the pause, and when they do the billing date moves to the manual resume date. If the renewal attempt fails, the subscription follows the same failed-renewal path as any other, through the billing grace period and account hold, so a pause does not remove your recovery steps, it just delays them.

Apple has no pause, so you extend the renewal date instead

Apple does not give users a pause switch. What it gives developers is the ability to move a subscriber's renewal date into the future, which grants free service time and pushes the next charge back. It is a server-side action, not something the customer taps.

Twice a year, up to 90 days each time

Using the App Store Server API, you can move the renewal date for a customer's subscription twice per calendar year, each up to 90 days into the future, and the customer is not charged for the added time. You attach an extendReasonCode that records why, such as customer satisfaction or a service issue. This is the closest Apple gets to a pause: instead of parking the plan, you extend it, so the person keeps access and the charge you would have collected simply moves out.

One call can cover an outage

There is a second endpoint that extends renewal dates for all active subscribers at once. Reach for it after an outage that hit your whole base, when handing everyone a few free days is cheaper and calmer than fielding a wave of individual refund requests. It is the mass version of the same idea: pay in free time, not in reversed charges.

Deferred billing, the lever for giving free time without a refund

Google Play's counterpart to Apple's extension is deferred billing, and it is more generous on duration. You call purchases.subscriptionsv2.defer and set a new expiry later than the current one. Billing can be deferred by as little as one day and up to a year per API call, and to go further you call again before the new date arrives.

The customer keeps everything, and pays for none of it

During the deferral period, the user remains subscribed to your content with full access but is not charged for the extra time, and the renewal date updates to the new date. Google's own examples are a retention playbook in miniature: give a week free for filling out a feedback survey, or comp access after an extended outage that kept people from using your service. Every one of those is a refund you did not write.

Wire up the notification

Deferring a charge triggers a SUBSCRIPTION_DEFERRED Real-time Developer Notification, so your backend can record the new renewal date and keep entitlements in sync. Prepaid plans work the same way: the defer call extends the expiration time rather than the renewal date.

What each option costs you in money

This is the part that should decide the order you try things in. The four moves are not close on price, and the gap is widest exactly where a refund hurts most, on a long, high-value plan.

A refund is the only move that pulls money back out

A refund returns the last charge to the customer. On a monthly plan that is one month. On an annual plan it is calculated on that last charge, which is the whole year, reversed in a single event. On top of the returned amount sit the costs you already spent serving the subscription, the compute, the third-party API calls, the storage, and any payout you sent a creator, none of which come back. A pause, a deferral, or an extension touches none of that. The only thing they cost is future revenue you postpone or a handful of days you comp.

The math a chargeback makes worse

The money case gets sharper when the alternative is not even a clean refund but a bank dispute. For Google Play orders placed after August 3, 2026, a lost chargeback costs the developer the purchase price minus Play's service fee, plus the bank's chargeback fee. A pause or a few comped days that keeps a frustrated customer from calling their bank is not just cheaper than a refund, it heads off a cost that is larger still.

When a pause or a deferral is the wrong tool

None of this works once the customer has stopped talking to you and started talking to their bank. A pause, a deferral, and an extension are all things you offer a subscriber who is still in your app. They do nothing for a dispute that is already filed.

A filed dispute has its own short clock

Once a charge is disputed, your only input is a narrow window. Apple sends your server a CONSUMPTION_REQUEST and waits up to 12 hours for consumption data. Google Play starts a 24-hour clock and takes your evidence through the orders.reviewrefund API. Miss either window and the decision is made without you. So the retention levers are for the moment before a refund request, not after a bank dispute.

A refund is sometimes the right answer

Offering free time to a customer who was genuinely wronged, and who wants their money rather than more of your product, can read as a dodge. When the fair move is to return the charge, return it. The point is not to never refund. It is to stop refunding by reflex when a pause or a comped week would have kept both the money and the customer.

The retention checklist

None of this is exotic. It is a short routine for the moment a subscriber signals they want out.

• Treat the refund as the last of four moves, after pause, defer, and extend, not the first.
• On Google Play, design for the pause: set includeSuspendedSubscriptions so your app does not mistake a paused subscriber for a churned one.
• Handle the end-of-pause renewal like a new purchase, because that is where a paused subscriber is kept or lost.
• Use deferred billing on Google Play and the renewal-date extension on the App Store to comp days instead of writing refunds.
• Record the SUBSCRIPTION_DEFERRED notification and the extension so your revenue numbers reflect the moved renewal date.
• Save the refund for the customer who was genuinely wronged and wants their money back, not for every request.
• Remember the levers stop working once a bank dispute is filed, and answer the 12-hour and 24-hour windows when one is.