GoldenGecko
PlaybooksGuidesFeaturesPricing
  1. Problems
  2. /Referral
  3. /People are gaming your referral program
Referral

People are gaming your referral program

Your referral numbers look great on paper, but something's off. Self-referrals, fake accounts, and reward abuse are eating your budget. You're paying for growth that doesn't exist. This is more common than you'd think — eBay's early referral program was so heavily gamed they had to shut it down entirely. Uber dealt with drivers creating fake rider accounts to collect referral bonuses in new markets. Any program that rewards signups without verifying real engagement will attract fraud, and it scales faster than legitimate referrals. If 20%+ of your referred 'users' never activate, you probably have a fraud problem.

TL;DR

"People are gaming your referral program" is a common referral problem. Key signs include spike in referrals from a small number of accounts — top 5% of referrers driving 50%+ of volume and referred users have suspiciously similar email patterns (same domain, sequential numbers, or disposable email providers). Start by trying: Move reward trigger from signup to first meaningful action (purchase, project created, 7-day retention) — this single change eliminates 80%+ of casual fraud.

Overview

If you're dealing with “people are gaming your referral program”, you're not alone. This is one of the most common referral challenges that solo founders and indie hackers face. Below you'll find the warning signs to watch for, root causes to investigate, and quick wins you can try today.

Signs you have this problem

  • •Spike in referrals from a small number of accounts — top 5% of referrers driving 50%+ of volume
  • •Referred users have suspiciously similar email patterns (same domain, sequential numbers, or disposable email providers)
  • •Referred users have activation rates under 10% compared to 40%+ for organic users
  • •Referral costs are climbing 20%+ month over month but revenue from referred users is flat
  • •Same devices, IPs, or browser fingerprints appearing across referrer and referee accounts
  • •Sudden referral volume spikes from specific geographies or during off-hours

Why this happens

  • •No verification that the referred user is a real, unique person — email-only signup is trivially gameable
  • •Reward is given at signup instead of after meaningful engagement — gamers never need to use the product
  • •No device fingerprinting or IP analysis to catch self-referrals — basic fraud that's easy to detect but often ignored
  • •Reward value is too high relative to the effort of creating a fake account — if the reward is worth more than 5 minutes of effort, it'll be gamed
  • •No cap on referral rewards per user or per time period — unlimited upside attracts professional gamers

Quick wins to try

1

Move reward trigger from signup to first meaningful action (purchase, project created, 7-day retention) — this single change eliminates 80%+ of casual fraud

2

Add basic fraud checks: same IP within 24 hours, disposable email domain blocklist, device fingerprint matching

3

Cap referral rewards at 10 per month or 50 total per user — legitimate referrers rarely exceed this, gamers always do

4

Require phone verification or a minimum activity threshold (3+ sessions, 7+ days active) before releasing the reward

When to prioritize this

When referred user activation rate is 50%+ lower than organic user activation, or when referral costs per activated user exceed your paid acquisition CAC. If more than 20% of referred accounts never log in a second time, investigate fraud before spending another dollar on referral rewards.

Related problems

Nobody uses your referral program

You built a referral program expecting viral growth, but users aren't sharing. The referral page gets visits but zero invites go out. This is one of the most common growth frustrations.

People get invited but never sign up

Your users are sharing referral links, but the people on the other end aren't converting. Invites go out and get ignored. Dropbox famously achieved a 60% invite acceptance rate by offering extra storage to both sides — most programs hover below 10%. The problem isn't your users' willingness to share. It's how the invite lands, what it says, and whether it gives the recipient a reason to care. If your referral landing page reads like a generic marketing page instead of a personal recommendation, you're burning every share your users send.

Users shared once and never referred again

Your referral program had a decent launch, but sharing dropped off a cliff. Users who referred once aren't doing it again. This is the most common referral failure mode — Uber saw it too, which is why they moved from flat bonuses to tiered ride credits that kept drivers engaged. The program feels stale because there's no ongoing reason to participate. One-time rewards create one-time behavior. If 90% of your referrers have exactly one referral, your program has a retention problem, not a reach problem.

Ready to solve “people are gaming your referral program”?

Golden Gecko matches you with the right experiments based on your specific situation. Sign up free to get step-by-step guidance and track your results.

  • AI matches playbooks to your goals
  • Step-by-step experiment guidance
  • AI interprets your results

Related problems

Nobody uses your referral program

You built a referral program expecting viral growth, but users aren't sharing. The referral page gets visits but zero invites go out. This is one of the most common growth frustrations.

People get invited but never sign up

Your users are sharing referral links, but the people on the other end aren't converting. Invites go out and get ignored. Dropbox famously achieved a 60% invite acceptance rate by offering extra storage to both sides — most programs hover below 10%. The problem isn't your users' willingness to share. It's how the invite lands, what it says, and whether it gives the recipient a reason to care. If your referral landing page reads like a generic marketing page instead of a personal recommendation, you're burning every share your users send.

Users shared once and never referred again

Your referral program had a decent launch, but sharing dropped off a cliff. Users who referred once aren't doing it again. This is the most common referral failure mode — Uber saw it too, which is why they moved from flat bonuses to tiered ride credits that kept drivers engaged. The program feels stale because there's no ongoing reason to participate. One-time rewards create one-time behavior. If 90% of your referrers have exactly one referral, your program has a retention problem, not a reach problem.

Explore more

All problemsAll playbooksReferral playbooks

Get personalized help

Get matched with the right experiments for your situation.

GoldenGecko

Always know what to test next. Proven playbooks, matched to your goals.

Product

  • Features
  • Pricing
  • Playbooks

Resources

  • Guides
  • Common problems
  • Glossary
  • Comparisons
  • Documentation

Geckoverse

  • Silver Gecko — SEO
  • Local Gecko — local SEO

Company

  • About
  • Privacy
  • Terms

© 2026 GoldenGecko. All rights reserved.