← Back to Blog

How to Run a Mobile Referral Program (Android, iOS & Web)

A complete guide to designing, building, and tracking mobile referral programs across Android, iOS, and web — covering deferred deep linking, double-sided rewards, fraud prevention, K-factor measurement, and real examples from Dropbox, Lyft, Rappi, and Robinhood.

$ ls contents/
  1. How mobile referral programs actually work
  2. Designing your referral program
  3. Real examples worth learning from
  4. Promoting referrals inside your app
  5. Using influencers in your referral program
  6. The technical foundation: cross-platform attribution
  7. Fraud prevention
  8. Measuring what matters
  9. Common mistakes that kill referral programs
  10. Pre-launch checklist

Word of mouth has always been the most trusted channel in marketing. A friend's recommendation cuts through ad fatigue in a way no banner ever will. Mobile referral programs formalize that trust — turning satisfied users into your most effective growth engine, rewarding them for doing what they'd do anyway: telling people about something they love.

But "formalize" is the operative word. A referral program without proper tracking infrastructure is just noise. You can't reward accurately, you can't measure what's working, and users who brought real friends — but never saw a reward — will quietly stop sharing.

This guide covers the full picture: how referral programs work on mobile, how to design them well, how to track them reliably across Android, iOS, and web, and what separates programs that drive real growth from those that quietly fade out.

How Mobile Referral Programs Actually Work

The mechanics are simple at their core:

1
Existing user gets a unique referral link or code

Tied to their profile, shareable anywhere they naturally communicate — WhatsApp, SMS, Instagram DM, email.

2
A friend clicks the link and installs the app

They complete a qualifying action — the trigger event you define — that confirms genuine engagement.

3
Both parties get rewarded

The referrer for sharing, the new user for joining. Programs that reward both feel generous — the new user receives a gift, not a sales pitch.

The "both parties" part matters more than most people think. Programs that only reward the referrer feel transactional. Programs that reward both feel generous — and this design pattern accounts for the overwhelming majority of successful referral programs.

The Technical Challenge Most Teams Underestimate

The flow sounds simple. The implementation isn't.

When a user clicks a referral link on mobile, they get redirected through the App Store or Google Play before your app opens. That redirect strips URL parameters. By the time your app launches, you've lost all referral context — you no longer know who referred whom.

User clicks referral link
  → Redirected to App Store / Play Store ← attribution breaks here
    → App installs and opens
      → Referral context is gone

On web, similar problems arise from Safari's Intelligent Tracking Prevention, cross-browser sessions, and ad blockers. On iOS, Apple's App Tracking Transparency framework adds another layer of complexity.

Deferred deep linking is the solution. It saves the referral context on a server before the app store redirect, then retrieves it after install using device signals — matching the new install back to the original click.

User clicks referral link
  → LinkTrace logs: device fingerprint + referral code
    → App Store redirect → Install
      → App opens, SDK matches fingerprint
        → Referral context restored: "Priya referred this user"
          → Both users rewarded correctly

This is the infrastructure problem that LinkTrace solves — privacy-first referral and install attribution that works across Android, iOS, and web, without requiring users to consent to tracking. You integrate it in two API calls and the hard part is handled.

Skip the attribution plumbing

LinkTrace handles deferred deep linking across Android, iOS, and web. Two API calls and you're live.

$ start-free --linktrace

Designing Your Referral Program

Before a single line of code, lock in these three decisions.

1. What triggers the reward?

The trigger event is the most consequential design choice in your referral program. Pick it too early and you'll attract fraud. Pick it too late and participation suffers.

Trigger Fraud Risk Best For
App install Very high Not recommended
Account creation Medium Only for very low-value apps
First purchase / transaction Low E-commerce, fintech, delivery
First meaningful action Very low Most consumer apps
Subscription start Very low SaaS, subscription apps

The general rule: pick a conversion event far enough down the funnel that it requires genuine engagement, but not so far that it feels impossible to reach.

2. What's the reward?

Match the reward to your product. The best referral rewards enhance the core product experience rather than sitting outside it.

Cash / wallet credits work best when monetary value is central to the product — fintech apps, delivery platforms, payment apps. The value proposition requires no explanation.

App credits (ride credits, order discounts, in-app currency) keep value in your ecosystem. Users have to return to redeem them, which drives retention on top of acquisition.

Premium feature unlock or extended trial — near-zero marginal cost for you, high perceived value for users. Excellent for SaaS and subscription apps. On iOS, this also sidesteps App Store restrictions on promotional pricing.

Product-aligned benefits — Dropbox's extra storage is the canonical example. The reward is the product itself, creating a viral loop: more referrals → more storage → more reasons to invite more people.

3. Who gets rewarded — and how much?

Always use double-sided rewards. When both the referrer and the new user receive something, the new user feels welcomed rather than recruited. Your cost per referred user goes up slightly, but conversion rate and user quality go up significantly more.

On reward sizing: start conservative and test. If you're spending ₹200 to acquire a user through paid channels, a referral program that costs ₹80 (₹50 to the referrer, ₹30 to the new user) represents a meaningful saving — and referred users typically have higher LTV than paid users to begin with.

Real Examples Worth Learning From

Case Study
Dropbox — Product as the Reward

Dropbox's referral program is the most studied case in mobile growth, and for good reason. In 15 months, they grew from 100,000 to 4 million users, with 35% of daily signups coming from referrals.

What made it work: the reward was storage — the exact thing users came for. There was no cash to track, no complicated redemption. Refer a friend, both get 500MB. The reward was built into the product, creating a loop where users who wanted more storage had a natural incentive to keep inviting.

Key lesson: When the reward enhances the core product, it creates a genuine growth loop rather than a one-time promotion.

Case Study
Lyft — Double-Sided from Day One

Lyft's referral program gave ride credits to both the existing rider and the new user. This wasn't just generous — it was strategically smart. The new user received a free first ride, dramatically lowering the barrier to trying the product. The referrer felt like they were doing their friend a favour, not spamming them.

Key lesson: The new user's reward is often more important than the referrer's. If the friend has no reason to click through, no amount of referrer incentive will move the needle.

Case Study
Rappi — Referrals at Scale with Influencers

Colombian super-app Rappi took referral marketing to its logical extreme. Rather than treating referrals as purely peer-to-peer, they recognised that influencers and affiliates are just referrers operating at scale — the mechanics are identical, only the audience size differs.

Influencers shared unique codes across social channels. New users installed the app, logged in via social account, and the coupon applied automatically to their first purchase. The result: over 50 million total app installs, with an average of 50% of new users in each new city launch coming from influencer and referral campaigns.

Key lesson: Referral infrastructure that works for individual users can scale to influencer programs with minimal additional engineering. Build the foundation right and influencer campaigns become an extension of the same system.

Case Study
Robinhood — Referrals Before Launch

Robinhood built a waitlist of nearly one million users before their app launched, almost entirely through referrals. The mechanic: sign up, see your position in the queue, move up by referring friends. This turned a constraint (limited beta access) into a growth driver. FOMO and the desire for early access did the marketing.

Post-launch, they shifted to gamified referrals offering random free stock — variable rewards that added excitement a fixed discount could never match.

Key lesson: Referral programs don't have to wait for launch. Pre-launch referrals can build a ready audience before you've shipped a single feature.

Promoting Referrals Inside Your App

The best referral programs don't feel like marketing campaigns — they feel like a natural part of using the product. Here's how to get placement and timing right.

Where to Surface It

Profile / Settings screen — always accessible for users who go looking. This is table stakes.

Post-action prompts — shown after a moment of delight: first successful order, first milestone reached, first savings goal hit, first workout completed. This is where participation rates are highest. Users who've just experienced value are emotionally primed to share it.

Onboarding nudge (carefully) — only appropriate if there's an immediate reward for the new user that makes the referral make sense at signup. Don't prompt cold.

Push notifications — a well-timed push can re-engage dormant referrers: "You've earned ₹200 in referral credits. Share with one more friend to unlock a bonus."

What the Referral Screen Should Show

Four things, immediately legible:

01
What you get — "Earn ₹100 wallet credits"
02
What your friend gets — "They get ₹50 off their first order"
03
Your unique link — easy to copy, easy to share
04
Your progress — referrals pending, completed, rewards earned

Progress visibility is not optional. Users who can't see the status of their referrals will stop trusting the program and stop sharing. Transparency is what keeps them engaged.

Share Channels That Actually Convert

Channel Conversion Notes
WhatsApp / iMessage Highest Personal context makes people click. Prioritise these.
SMS High High open rates. Works for users off heavy messaging apps.
Telegram High Increasingly important in many markets.
Email Medium Lower conversion but works for higher-consideration products.
Social media posts Low Great for brand awareness, not referral completion.

Pre-populate the share message. Remove every tap you can. A two-tap share flow outperforms a five-tap flow every time.

Using Influencers in Your Referral Program

Influencer marketing and referral marketing are the same mechanic at different scales. An influencer with 200,000 followers is just a very high-reach referrer with a unique code.

This means your referral infrastructure — unique codes, attribution, reward fulfillment — can power influencer campaigns with minimal extra work. Give influencers unique codes distinct from regular user links so you can track them separately, set different reward structures if needed, and measure their conversions exactly as you would any other referral source.

The Rappi model: They didn't build a separate influencer platform. They extended their referral program to include influencers and affiliates as a new tier of the same system. Referral conversion depends on trust more than reach — a personal recommendation from a creator someone follows daily carries far more weight than a display ad, even at a fraction of the audience size.

The Technical Foundation: Cross-Platform Attribution

Regardless of whether your users are on Android, iOS, or web, you need three things from your attribution infrastructure:

1
Unique link generation per user

Every user gets a trackable link tied to their account. Short links (e.g. ltr.in/abc123) work better for sharing than long parameter-laden URLs.

2
Deferred deep linking

Captures referral context before the app store redirect and restores it after install. Without this, you lose attribution for every user who doesn't have your app installed.

3
Cross-channel attribution

The same referral link must work whether clicked from WhatsApp on Android, Safari on iPhone, or a desktop browser. The user should always land in the right place and carry the right referral context.

LinkTrace handles all three across Android, iOS, and web. It's built as a privacy-first alternative to Branch and Firebase Dynamic Links — the same deferred deep linking and attribution infrastructure, designed to integrate in minutes rather than days. If you're building or rebuilding your referral program, this is the layer to get right before anything else.

Fraud Prevention

Any program with financial incentives will attract abuse. These are the patterns to watch for and how to catch them:

Fraud Pattern How to Catch It
Self-referral Check if referrer and referee share the same email, phone, device ID, or payment method.
Multi-account farming Device fingerprinting (which LinkTrace captures) catches most of this.
Click injection (Android) Check the time delta between click and install — under 2 seconds is suspicious.
Velocity spikes Set a daily cap per referrer and flag outliers for review before paying out.
IP cycling Rate-limit completions per IP and check against known VPN/proxy ranges.

Right posture: Build basic checks from day one, don't wait until fraud costs become visible. The threshold for legitimate users is rarely affected by reasonable fraud controls.

Measuring What Matters

K-Factor — The North Star Metric

K = (average invites sent per user) × (conversion rate of those invites)

A K-Factor above 1.0 means each user brings in more than one new user on average — true viral growth. Below 1.0, you still get organic uplift on top of paid channels; you're just not fully self-sustaining.

K-Factor What It Means
> 1.0 Viral — each cohort grows the next
0.4 – 1.0 Strong organic contribution
0.2 – 0.4 Meaningful supplement to paid
< 0.2 Needs significant optimization

For most consumer apps, 0.2–0.4 is a realistic and valuable target. Dropbox at peak was around 0.7 — don't measure yourself against that.

The Full Metrics Scorecard

Metric Formula Healthy Benchmark
Participation rate Sharers ÷ Eligible users 10–25%
Share rate Completed shares ÷ Referral screen visitors > 40%
Invite CTR Clicks ÷ Invites sent 20–35%
Referral conversion rate Signups ÷ Clicks 10–25%
Cost per referred user Total reward cost ÷ Referred users < Paid CAC
Referred user LTV Track cohort separately > Average LTV

Track referred users as a separate cohort in your analytics. In most apps, referred users retain better and generate higher LTV than users from paid channels — and quantifying this gives you the business case to invest more in the program.

Common Mistakes That Kill Referral Programs

Rewarding on install. Install fraud is widespread. Fake installs from device farms or click injection will drain your budget before you notice. Always require a meaningful action — first purchase, first transaction, first subscription.

One-sided rewards. If only the referrer benefits, new users feel like they're being recruited. Double-sided rewards convert dramatically better.

No visibility into referral status. Users who share and hear nothing will assume the program is broken. Show them exactly which referrals are pending, which have completed, and when they'll receive their reward.

Prompting before users have experienced value. A new user who just installed your app hasn't decided whether they like the product yet. Wait for a moment of genuine delight before asking them to share.

Building attribution yourself. Deferred deep linking across app store redirects is genuinely difficult — device fragmentation on Android, ATT restrictions on iOS, Safari ITP on web. Tools like LinkTrace exist precisely so you don't spend a sprint building and debugging attribution infrastructure.

Pre-Launch Checklist

$ checklist --pre-launch
  • Reward trigger event defined — not install, a real conversion
  • Double-sided reward structure confirmed
  • Unique referral link generated per user
  • Attribution SDK integrated (Android, iOS, web) — LinkTrace recommended
  • Deferred deep linking tested end-to-end on each platform
  • Backend referral and reward tables created
  • Fraud checks active: device dedup, self-referral detection, velocity limits
  • In-app referral screen built with progress visibility
  • Share sheet pre-populated for WhatsApp, SMS, copy link
  • Analytics tracking: invites sent, clicks, installs, conversions, rewards issued
  • Full user journey tested: share → click → install → trigger event → reward

A mobile referral program, done right, is one of the most efficient growth levers available to any app — on Android, iOS, or web.

Referred users arrive with trust already built. They convert at higher rates, retain longer, and refer others more readily than users from any paid channel.

The design principles are straightforward: double-sided rewards, a meaningful trigger event, transparent progress tracking, and share flows optimised for the channels people actually use (WhatsApp and direct messaging, overwhelmingly).

The technical foundation — accurate cross-platform attribution through app store redirects — is where most implementations quietly fail. Getting this right is the difference between a referral program you can trust and one that erodes user confidence every time a reward goes missing.

Start tracking referrals the right way

Privacy-first referral and install attribution for Android, iOS, and web. Two API calls and you're live.

$ start-free --linktrace