Cash or credits: which referral reward gets abused less?
Cash or credits for your referral program? What a farmed reward really costs, what you can get back, and which basic fixes break before payout.
You picked cash because you figured nobody shares a link for store credit. This morning your co-founder wants a screenshot of the referral leaderboard for this week's newsletter. The account in first place referred 200 people overnight. You click the first name on its list. Verified email, signed up at 11:41 p.m., finished the one onboarding step that counts as a qualified referral, and hasn't been back since. You click the second. Different name, signed up at 11:43 p.m., same onboarding step, same silence. By the ninth you've stopped reading the names.
At 7:05 a.m. your co-founder posted the signup chart in the team channel with three exclamation marks. A normal day brings you 15 signups, so 200 looked like more than 13 days of growth in one night. It was one referrer and one payout account, adding a new person about every 100 seconds until 5:10 a.m.
You already know what the payouts page will say, and you open it anyway. The nightly run went out at 6:00 a.m., a little over two hours ago. Status: completed.
This is a scenario, so check it against your own numbers. Say you pay $30 in cash for each qualified referral:
- 200 rewards × $30 = $6,000, your entire budget for the quarter
- 2% payout fee on $6,000 = $120
- 200 welcome credit packs × $1.20 = $240
- Total: $6,360
Thirteen weeks of budget lasted five and a half hours. If your company spends $16,000 a month, $6,360 is about 12 days of runway. You were counting on those 200 referrals turning into customers at $29 a month, which would have been $5,800 in new monthly revenue. The real number is $0, and the account that holds your $6,000 has no reason to answer your email.
At 9:30 a.m. a customer writes in: "My reward still says pending. I told my whole team about you." Eleven real customers have referred 17 colleagues since you launched on Thursday, and those colleagues finish onboarding at human speed, so none of them qualified before the money ran out. That's 17 × $30 = $510 you owe from a budget that no longer exists. You can find it somewhere else, or you can tell the 11 people who like you most that the program they just recommended is closed. The newsletter with the leaderboard is scheduled for 10:00, and your co-founder wants to know whether you're pausing the program, refilling it or quietly ending it one week after announcing it.
By that afternoon the question on the whiteboard is referral reward cash vs credits. Credits feel safer, since nobody can walk away with them. That's partly true, and the part that isn't can still cost you thousands.
Why referral rewards get farmed so fast
A referral reward pays for a new customer, and in most programs the only proof of a new customer is a new account. The person asking for the money is also the person producing the proof. In the scenario above, $6,000 for about 5.5 hours of work is roughly $1,090 an hour, which is a very good night for someone with a few inboxes.
Swapping cash for credits does help. It also hides the problem, because a credit has two prices: the one on the label and the one you pay when someone spends it.
Referral reward cash vs credits: the real cost
Take the same $30 reward in four forms. Cash costs $30.60 with the 2% fee. Credits cost whatever it takes to serve them: $18 where serving costs 60% of list price, $3 where it costs 10%. Credit that only comes off a paid invoice costs nothing on an account that never pays.
| Reward | Cost if farmed | When value leaves | Can you get it back? | Appeal to referrers |
|---|---|---|---|---|
| Instant cash | $30.60 | Next payout run | Rarely | Highest |
| Cash after a hold | $0 until the hold ends | After the hold | Yes, until it's paid | High |
| Spendable credits | $3 to $18 | As soon as they're used | Only the unused part | Medium |
| Invoice credit | $0 unless they pay you | Never leaves as cash | Yes, until it's applied | Lower |
What you actually lose depends on how much you can take back:
Run the 200 rewards through it. Cash is 200 × $30.60 × 1 = $6,120. AI credits with 40% still unspent at 8:20 a.m. are 200 × $18 × 0.6 = $2,160. Invoice credit is $0.
Real customers are different. $30 of credit given to someone who would have paid you anyway replaces $30 of revenue, so it costs close to the full $30.
What you can get back the morning after
Once cash says completed, your options are a polite email and whatever your payout service allows. Credits have two states. Unspent credits come back with one database update. Spent credits are gone, because you've already paid for the compute, and in an AI product the gap between those states can be minutes. See how to take back free credits after an abuse wave.
So which gets abused less? Invoice credit, by a wide margin, because a farmed account has to become a paying customer before the reward is worth a cent. Spendable credits come next, depending on your serving cost. Instant cash draws the most effort, since it's the only reward worth the same outside your product. If credits can move between accounts or be pooled in a workspace, treat them as cash.
Invoice credit still needs a cap. One paying account with 200 referrals holds 200 × $30 = $6,000 of credit, which at $29 a month is about 207 months, or more than 17 years free.
The basic fixes, and how each one breaks
Start with the design of the reward, because it's cheap and it limits the worst night.
Pay on the first paid invoice, and keep the reward under it. At $30 on a $29 invoice, a farmed account still clears $1. At $20 it loses $9.
Hold every reward past your refund window. A pending reward can be withdrawn, so write your terms to match.
Set a daily ceiling for the whole program. At $200 a day, the worst night costs $200 and a $6,000 budget lasts at least 30 days.
Cap rewards per referrer, and cap how fast reward credits can be spent, so there's still something left to take back.
Those limit what you can lose. Telling a real friend from a second account is harder, and most teams try these next.
| Basic fix | What it catches | What it misses | Who it wrongly blocks | What it costs to keep working |
|---|---|---|---|---|
| Email verification | Typos and dead addresses | Any second real inbox | Friends whose code lands in spam | Deliverability work |
| Unique emails | The same address twice | Every different address | Almost nobody | Very little |
| Disposable-domain blocks | Known throwaway services | New domains and ordinary mailboxes | People using privacy forwarding | A list someone updates forever |
| CAPTCHA | Simple scripts | People claiming by hand | People using screen readers | Fees and lost signups |
| Phone or card requirement | Casual second accounts | Anyone with another number or card | Friends who wanted to try it first | Text charges and unfinished referrals |
| One referral per IP address | Repeats on one connection | Any other network | Colleagues in the same office | Exceptions and tickets |
| VPN blocks | Known VPN ranges | Everything off your list | Staff on company VPNs | Another list to maintain |
| Cap per referrer | One account overrunning | A second referrer with its own cap | Your best advocate | Handling exceptions by hand |
| Smaller rewards | Some value per claim | More accounts make up the difference | Every honest referrer | Fewer referrals |
| Manual review before payout | What a person can spot | Whatever looks ordinary | Whoever looks odd to a tired reviewer | 200 claims × 2 minutes = 6 hours 40 minutes |
Email is where the limits are easiest to see. Gmail ignores dots, so jane.doe@gmail.com and janedoe@gmail.com reach one inbox, and mail to janedoe+anything@gmail.com lands there too. That's how email works, and it's how aliases create duplicate accounts.
Colleagues are the group referral programs can least afford to lose. People refer the person at the next desk, on the same office connection. One rule tuned too tightly turns away the exact referrals you were paying for. Meanwhile the domain list goes stale, a CAPTCHA does nothing about someone claiming by hand, and the review queue grows on the nights nobody's awake to read it.
Evaluate the claim before the reward moves
Every row in that table looks at one narrow thing about one signup, and a referral farm only needs each fresh account to look fine on its own. AbuseZero starts from the question those rows can't answer: does this claim look like a repeat? With multi-layer AI risk analysis, AbuseZero examines 140,000+ signals, detects linked claims and flags abuse that single checks miss, and those signals are read together, never one at a time. Where a unique-email rule stops, AbuseZero checks mailbox aliases, disposable domains and mail routing. Where a one-per-IP limit stops, it sees beyond an IP address with VPN, hosting, network reputation, Tor and proxy intelligence. A reused card isn't a clean start under a new account either, because AbuseZero detects repeated payment fingerprints across users, even when new accounts and emails hide the connection, and it recognizes repeat devices across new accounts, even when email addresses keep changing. There's also a proprietary algorithm, with AI that checks the claim again. Creating a fresh account isn't enough to erase previous activity, and fresh accounts are what a referral farm runs on.
Cash or credits, it's one API call before the reward moves and a clear allow or deny, under rules you control. Set up the reward as a benefit, and Observe mode shows what would be denied before anyone is blocked. You don't have to build your own abuse prevention system. Keep your holds and ceilings beside it, and start with the quickstart.
Frequently asked questions
Is cash or credit better for a referral program?
Credit is usually safer, and invoice credit is the safest. Cash appeals most to real referrers, so if you use it, pay after a hold and only on a real payment.
Can I take back a referral reward after it's paid?
Cash, rarely. Unspent credits, yes, with one update to the balance, provided your terms allow it. Spent credits are gone.
Is referring someone at the same company abuse?
No. It's often the best referral you'll get. Write down what you mean to give each person before you turn anyone away. AbuseZero's evaluation applies the limits you set, and they're yours to change.
Pay for referrals that are real
Start with Builder
Pay as you go. Protect your first reward today.
See pricing
Builder, Pro, Scale and Enterprise.
Talk to us about Enterprise
High volumes and custom terms.