with the skill
**I wouldn’t launch an uncapped $50 reward for a free signup.** You’d be paying for an easily manufactured event, not necessarily a valuable customer.
This is a classic **Cobra Effect** risk: incentives produce more of what you reward—signups—even when what you actually want is retained, valuable users.
## How I’d set it up
### 1. Define a qualified referral—not just a signup
Use a structure like:
> “Earn $50 when a new customer joins through your link, becomes a paying customer, and remains eligible through the refund period.”
Require the referred customer to be genuinely new and prohibit self-referrals, duplicate accounts, and coordinated abuse.
If free users create meaningful value—for example, through collaboration or marketplace activity—you could reward **verified activation plus retention** instead. But first establish that those behaviors predict enough business value to justify $50.
**For a free-plan product, my default would be:**
- Offer a modest, cost-controlled product benefit when the friend reaches a meaningful activation milestone.
- Pay the $50 only after a qualifying paid conversion and an appropriate waiting period.
Show the rest (44 more lines)
Make all conditions visible before someone shares a link.
### 2. Check whether $50 is affordable
Evaluate the **total cost per incremental valuable customer**, not the cost per signup.
Include:
- Cash rewards and any recipient incentive.
- Payment fees, fraud losses, support, and program administration.
- Whether referred customers would have signed up anyway.
- Contribution margin and how long it takes to recover acquisition costs.
For example, if you pay $50 per free signup and only 5% become paying customers, **the reward cost alone is $1,000 per paying customer**—before fraud or operating costs.
### 3. Start with a controlled pilot
- Invite a limited group of existing, engaged users.
- Set an overall budget and clear per-referrer limits.
- Use unique referral links and a published attribution window.
- Show rewards as **pending → qualified → paid**, with clear timing.
- Review suspicious activity before payout; automate as you learn.
- Keep a comparable unexposed group where feasible to measure incremental lift.
Don’t scale based on referral volume alone.
## What could go wrong?
| Risk | Why it happens | Mitigation |
|---|---|---|
| **Fake accounts and self-referrals** | A free account costs less to create than the $50 payout | Eligibility checks, velocity limits, risk-based verification, delayed payouts |
| **Real but low-value users** | People recruit friends who want the incentive, not the product | Reward activation, retention, or payment—not registration |
| **Spam and damaged trust** | Referrers chase cash through unsolicited messages | Clear sharing rules, disclosure of the incentive, complaint monitoring |
| **Paying for organic growth** | Existing prospects use referral links before joining | Measure incrementality, not just attributed signups |
| **Refund or payment abuse** | People qualify, collect rewards, then reverse payments | Wait through relevant refund periods and review suspicious transactions |
| **Unexpected cash exposure** | Rewards become payable before customers generate margin | Budget caps, payout forecasting, cohort monitoring |
| **Legitimate users being blocked** | Families and workplaces share devices or networks | Use multiple risk signals and offer an appeal process |
Also review applicable tax, privacy, promotion, and payment requirements. Don’t upload friends’ contacts or contact them without appropriate permission.
## What I’d measure
Track **incremental retained customers, contribution margin, payback period, fraud rate, and referral complaints**. Compare referred cohorts with other acquisition channels.
**Bottom line:** use the free plan to make trying the product easy—not to make collecting $50 easy. Reward the earliest *verifiable behavior that reliably creates value*, and scale only once the economics hold up.