How to Build a Referral Program for a Mobile App
Referral programs aimed at subscription apps have not yet matured, and the public record of well-executed examples is thin. That scarcity pushes most teams toward mechanics proven in other consumer industries, then reworked to fit in-app purchase rules. The design problem is simple to state and difficult to solve: hand both the person sharing and the person arriving something genuinely valuable, while keeping the program hard to exploit and inexpensive next to a subscriber's lifetime value. This guide walks through the business case, funnel math, the reward routes available on iOS and Android, and a build checklist that keeps App Review and fraud risk in view.
Does a referral program earn its place in your acquisition mix?
Alex Bauer's framing is that referral schemes have served many industries well for a long time: they are easy to grasp, they work from an audience you already have, and they spare you the cost of buying ad campaigns, because your users do the inviting. Andrew Chen, a general partner at Andreessen Horowitz, has put a successful referral program at 20-30% of an acquisition mix. Treat that as an outside estimate rather than a promise for your particular app, but it explains the interest the channel attracts. Referral is no magic fix, though. Shipping one will not produce explosive growth on its own, and subscription developers still have to operate inside Apple's rules on promotions and its in-app purchase API. Publication date: pending owner confirmation.
What other industries already prove
Because referral programs for subscription apps are still maturing, looking at what app referral programs in general can do is a sensible starting point for developers. With paid advertising under pressure, apps search for other routes to grow acquisition and engage the users they already have.
Model the referral funnel before you build it
Suppose you have 10,000 active users and want to maximize new subscriptions. Put the call-to-action where users will actually see it, not buried in settings. As Alex Bauer notes, the failure mode is not weaving the referral program into the product experience. Be bold with the CTA, but keep it from reading like an ad.
The strongest programs — Dropbox, Uber, and Robinhood among them — are woven into the product in so many places and so tightly that a user cannot miss the option, he says. When incentives line up with product use, users feel more engaged.
Weigh the implementation cost of a referral program against product work and other marketing. Consider its lifecycle too: Dropbox's peak of 35% of users gained through referral (a historical third-party figure) faded as word-of-mouth spread and the program went viral.
Referral-sourced users are among the highest-quality users you can get, Alex says. How leaky your funnel is after acquisition depends on the channel, and referrals are one of the best ways to bring in users who actually stick around.
Design rules that keep a program honest
Incentivize both sides and choose the incentives carefully. A financial reward is a natural starting point, but you can also hand users meaningful value for free rather than making them pay — comparable to Dropbox's free storage. The reward has to be integral to the product and hard to game.
A referral program must beat your customer acquisition costs (CAC) and be established early in the user journey. Gamifying the process engages and respects your current users without alienating them.
Attracting the right users — and protecting against fraud and abuse — means firing completion events at the right moment. Alex advises picking a conversion event far enough down the acquisition funnel that repeating it is too much pain for a user. Incentivizing the initial purchase also discourages fraudulent users from trying to replicate or automate the program for their own benefit.
It is, in Alex's words, the ultimate cost-per-action campaign. Developers need to confirm they are gaining value equal to the giveaway incentive and establish how long that takes.
Reward mechanics: referee and referrer
A referral scheme has to give something to the newcomer and something to the person who sent the invitation, and those two payouts need not match. Since no settled convention governs these mechanics, teams usually have to define their own rules from scratch. Rewarding the newcomer is the simpler problem, and there are four ways to approach it. One is a dedicated SKU carrying an introductory offer that appears nowhere else in the app: for instance, the paywall might advertise a free week, while a second SKU quietly grants a free month. Introductory offers come in several shapes, including no-cost access, pay-as-you-go pricing, and recurring discounts or lengthened trial windows; the limitation to keep in mind is that they reach new subscribers only. A second option is Stripe. Reviewers scrutinize anything that appears to divert users from in-app purchase, yet Apple has no authority over messages sent outside the app and the store, which is exactly where a friend-to-friend email referral lives. Stripe then makes it possible to build a highly adaptable billing relationship with each user, though if Stripe is not already part of your stack, weigh whether the referral program alone justifies the tax and international compliance burden. A third option is a promotional entitlement issued through subscription backend infrastructure that tracks each user's status: once the entitlement lapses, the account reads as unsubscribed and the standard paywall appears. The trade-off is that nothing renews automatically, which leaves a business choice for the app owner — either route users into an introductory-offer trial that converts unless cancelled, or hand out a free month and let the paywall return afterward. Offer codes are the fourth possibility, workable if you can get them functioning; the newcomer receives a simple link and uses it as they wish, though one of the earlier approaches may prove less trouble. Rewarding the referrer is where things get complicated, and several of the same avenues reappear with different plumbing. Stripe is often the cleanest answer for paying referrers because its API is approachable and pleasant to integrate with; Alex calls it among the best APIs he has encountered. Being programmatic, it suits people who already pay you, and either an in-app claim button or an automatic offer banner serves equally well. Because email sits outside the App Store, Stripe makes both charging and rewarding straightforward, and it can absorb leftover time from Google Play or Apple's App Store when a subscriber cancels on either platform. The drawbacks deserve equal attention: Apple quietly absorbs a lot of grace-period and failed-payment handling that developers would otherwise own, and moving to Stripe can shift that work back onto you. Promotional offers in the App Store are capped at 10 active offers at any one time, require the recipient to be a current or former customer, and force a confirmation dialogue plus Apple's payment flow for every reward; implementation is laborious from initiation through eligibility checks, so this is not a realistic first version. Apple's server-to-server subscription extension API sidesteps some promo-code headaches by applying extensions automatically, but it permits only two uses per year with a ceiling of 90 days per period, so a continuous 180 days cannot be assembled; it does allow accrual, which makes it a stronger candidate for an early build. A promotional entitlement API is likewise simple to apply on the referrer side, and again it will not renew on its own — yet that can be a feature if your strategy leans on free users as a marketing force, since it pays them without tying them to a payment flow. Offer codes can serve referrers too, within limits: one code per active offer, no more than ten active offers, and 150,000 codes annually, and a single-use offer makes for a weak incentive.
Build checklist
A subscription app's referral program deserves a short pre-launch review against the points below. Ship a minimum viable version rather than waiting for the polished end state; the goal is to get the app into people's hands, so begin with reward types that are quick to implement. Place the call-to-action where it can be seen — never tucked into settings — while keeping its tone informational rather than promotional. Whatever flow you construct, confirm that App Review can actually exercise it, and when separate SKUs exist that are invisible inside the app, spell them out in the review notes. Confine the program and every reward it pays to your own app; programs that resemble offer walls and pay people for installing a different app are off limits. Deferred deep linking should not be confused with tracking: Apple defines tracking as matching an install back to an advertisement, whereas a friend's link that leads someone to install an app in order to claim an in-app offer does not fall under that gate. Build a shareable link that carries the referrer's identity, resolve it through a deferred deep link, and record the association in subscriber attributes and analytics events. Give the newcomer a reason to feel personally welcomed, so that opening the app makes them want to redeem what was offered. Requiring referrers to claim their reward by hand is a useful brake on abuse and fraud, and it tends to make people more invested in the referrals they bring. None of this replaces a broader marketing plan: referrals may well supply a meaningful slice of acquisition, but the overall mix still has to hold together for the app to succeed. The material here draws on a webinar conversation between David Barnard and Branch's Alex Bauer.
| Route | Side | Key constraint | Renewal behavior |
|---|---|---|---|
| Separate SKU with introductory offer | Referee | Applies only to new subscribers | Can auto-convert or expire to paywall |
| Stripe | Referee and referrer | External billing; taxes and global compliance | Flexible, programmatic |
| Promotional entitlement (third-party backend infrastructure) | Referee and referrer | No auto-renewal; owner must confirm provider | Expires to paywall |
| Offer codes | Referee and referrer | 1 code per active offer, up to 10 active offers, 150,000 codes/year | Single-use codes weak as incentive |
| Promotional offers | Referrer | Max 10 active simultaneously; current or prior customer; confirmation dialogue | Confirmation flow required |
| Subscription extension API | Referrer | Max 2 uses/year, 90 days each; 180 consecutive days not possible | Accrual allowed |
Limits reflect platform rules at the time of writing; verify current App Store and Google Play documentation before launch.
Frequently asked questions
Can a referral program alone grow a subscription app?
No. Referral is one lever among several. It can represent a meaningful share of acquisition, but it needs to sit alongside product work and other marketing, and Apple's promotion restrictions and in-app purchase API shape what you can offer.
What is the difference between referee and referrer rewards?
The referee is the new user arriving through the link; the referrer is the existing user who shared it. They do not have to receive the same reward. Referee rewards are generally easier to implement; referrer rewards often need billing or entitlement infrastructure.
Is deferred deep linking the same as tracking?
No. Apple treats tracking as matching a user's install to an ad. A friend-to-friend link that leads to an in-app offer after install is not gated the same way.
How do I stop people from gaming the program?
Pick a conversion event far enough down the funnel that repeating it is painful, require manual reward claiming, and avoid offer-wall style programs that reward downloading another app.
What should I build first?
Start with an MVP reward route rather than the ideal end state. Separate SKUs with introductory offers and promotional entitlements are the most MVP-friendly starting points; promotional offers and Stripe integrations add more work.