Referral Program Calculator

A free tool for founders deciding whether a customer referral program pays for itself. Enter your active customers, the share you will ask, the share who refer, referrals per referrer, the share of referrals who become customers, the rewards on each side, the program's fixed cost, lifetime value and the CAC of the channel referral would replace. It returns referral CAC, the break-even reward, payback, LTV:CAC, what referral saves against the other channel and the loop factor k. It runs the engine of the referral-program-design agent skill in your browser; nothing you enter is sent anywhere. Every rate is yours: the calculator holds no benchmarks.

How it works

• Asked = customers × share asked. Referrers = asked × share who refer. Referrals = referrers × referrals each. New customers = referrals × share who become customers.
• Referral CAC = (rewards + fixed cost) ÷ new customers. The break-even reward is the CAC it replaces minus the fixed cost per new customer: the most the two rewards together can cost per referred customer.
• Payback = referral CAC ÷ monthly gross margin per customer. k = new customers ÷ customers for one period; with k below 1, each customer you win turns into 1 ÷ (1 − k) customers if referred customers refer at the same rate.
• A reward paid per referral, not per customer, costs the reward ÷ the referral-to-customer rate for each referred customer.

It flags rewards above the CAC they replace, referral CAC above the other channel or above lifetime value, rewards paid per referral, one-sided rewards, rewards larger than a month of customer spend, asking every customer, a missing fixed cost and inputs you mark as guesses. Tick what applies and it points to the rules to check: the FTC Endorsement Guides, the FTC consumer reviews rule and Google's policy, the Anti-Kickback Statute and HIPAA marketing authorization. Pointers, not legal advice.

k here is what the program adds in one period. To model a full referral, content or paid loop stage by stage, use the Growth Loop Calculator.

Limits

• If your ask and conversion rates are guesses, as they are before launch unless you have run a pilot, mark them, run the first period as the test, and replace them with what you measured.
• Referred customers are not assumed to be worth more. The often-quoted 16% comes from one German bank; enter a separate lifetime value only if you have measured your own referred customers.
• Not legal advice. Check what the pointers mean for your program with counsel before launch.

Why I built it

Most referral programs I see start with the reward: what would make a customer share? I would rather start with the ceiling. If the CAC you are replacing is known, the most you can pay per referred customer is known too, and the reward design has to fit under it.

The research and the rules are in AAJ's post Referral Programs for Startups. To check the lifetime value and CAC you enter, use the Unit Economics Calculator.

Run this in your agent

The free AAJ agent skill Referral Program Design does this job inside a coding agent, on your own data. Install it with npx skills add sarojkjha/aaj-marketing-skills --skill referral-program-design.

Using Claude or ChatGPT instead? Add the AAJ engines as a custom connector, with no install: https://aaj-engines-mcp.sarojkjha10.workers.dev/mcp