There is no fixed number. Divide your monthly program costs by the margin one partner-sourced customer brings, and you get the customers you need each month. Then divide by what an active partner delivers. Because most signed-up partners never sell, plan on recruiting several times the number of active partners you need.
A founder asks how many partners she needs to make this worthwhile. The honest answer is a short piece of arithmetic that nobody wants to do before launch, because it has an uncomfortable middle step: most of the people who sign up will never send you anything.
The break-even math, step by step
Start with what the program costs you every month, whether or not anyone sells. That is fixed cost: software, the hours someone spends reviewing commissions and answering partners, and any content or assets you maintain. Commissions are not in this number, because they only exist when a sale does.
Then find what one customer brought in by a partner is worth to you after the commission. Take the customer's margin over the period you care about and subtract what you pay the partner.
Here is a round example. It is hypothetical, so swap in your own numbers.
- Fixed cost per month: $2,000 (software plus about 20 hours of someone's time).
- Average customer pays $100 a month and stays 12 months, so $1,200 in revenue.
- Gross margin is 80%, so $960 of margin per customer.
- You pay partners 20% of payments, or $240 per customer.
- Net contribution per partner-sourced customer: $720.
To cover $2,000 a month you need about three customers a month ($2,000 divided by $720 is 2.8). Anything above that is the program making money. Anything below, you are paying to run it.
Notice what the formula did not need. It never asked how many partners you have.
So where do partner numbers come in?
They come in when you ask where three customers a month will come from. Say a typical active partner sends one customer every two months. You would need six active partners to deliver three a month. That sounds small, and for a program with a clear fit it is.
Now the uncomfortable part. Of the people who join a program, only some ever send a single click, and fewer ever make a sale. If one in five of the partners you approve turns out to be active, then six active partners means about 30 approved ones. If it is closer to one in ten, 60. You do not know your ratio until you run the program, which is the reason to launch with a small group and read the results before you promise a number to anyone.
The distribution is also lopsided. A common pattern is that a few partners account for most of the revenue, so the real question is whether you can find and keep those few. A hundred inactive partners add support load without adding revenue, and one strong partner who sells ten customers a month changes the picture alone. This is why a smaller program of well-chosen partners often beats a larger one that took anyone with a link. It also explains why recruiting matters more than raw volume; the techniques are covered in how to recruit partners who actually sell.
Why does the break-even point move?
Four things, and you control some of them.
The commission rate is the obvious one. In the example above, raising the rate from 20% to 30% moves the contribution from $720 to $600 per customer ($960 minus $360) and the break-even from 2.8 to about 3.3 customers a month. That is still manageable, but stack it with lower margin or a shorter customer lifetime and it stops being so. See how to set commission rates for the tradeoffs.
Fixed costs are the second. If the program takes 20 hours a month of someone's time today and 40 once you have a hundred partners, your cost has doubled. Software that saves review time is worth its price when your time is the larger cost, and not worth it when you have four partners and a free afternoon.
Refunds and churn are the third. A customer who leaves after two months does not deliver the $720; they deliver much less. If a program's customers churn faster than your average customer, the numbers above are too hopeful, and a longer holding step before approval, or a lower rate, may be what saves it.
The fourth is the platform fee, if you use software that takes a percentage of commissions paid. In Affiliate Tabs that fee is 3% on Launch, 2% on Growth and 1% on Enterprise, added to the plan price. On $240 of commission that is a few dollars per customer, which is real money at volume but rarely the deciding factor. Count it with the other costs.
How do you tell, early, whether it is working?
Skip vanity numbers. A signed-up partner count says little. Three measures are more useful in the first quarter.
- Share of approved partners who produced a click, and the share who produced a sale. These two ratios are your active rate.
- Cost per acquired customer: all program costs plus commissions, divided by customers it produced. Compare it with what other channels cost you.
- Time from approval to first sale. If it is long, your onboarding may be the problem, which the guide to partner program metrics looks at further.
Give a pilot long enough to cover at least one full sales cycle. Reading a program after three weeks, when your customers typically take eight to decide, will make it look worse than it is.
When is a program not worth running?
If your net contribution per customer is small, or the customers leave quickly, there may be no number of partners that pays for the fixed cost. A product with thin margin, or one that people use once, can be a poor candidate. In that case a simple referral reward for existing customers may do more at lower cost than a full program. Be willing to conclude that, and do the arithmetic before the launch, not after.
If the numbers do work, the practical plan is modest. Compute your break-even customers per month. Guess an active rate on the cautious side. Recruit the number of partners that implies, and no more, until you have measured your own ratio. The basic structure of the program itself is covered in what a partner program is and which type fits.
Common questions
- What percentage of partners are usually active?
- It varies widely by program and industry, and there is no reliable universal figure. In most programs a small share of partners produce most of the results and many produce nothing. Measure your own pilot group rather than relying on someone else's average.
- How long until a partner program breaks even?
- It depends on your sales cycle, how fast you recruit and what you spent on setup. Programs with short sales cycles can show results in the first few months; programs for high-value, long-cycle sales often take two or three quarters. Track cost per acquired customer from the first month.
- What costs should I count in a partner program?
- Count commissions, software fees, any platform fee on commissions, staff time for review and partner support, and any bonuses or assets you give partners. Leaving out staff time is the most common mistake.
- Is a few strong partners better than many small ones?
- Often yes. A handful of partners with an audience that fits your buyer can outsell a long list of inactive sign-ups. But dependence on one partner is a risk, so keep recruiting while you nurture the strong ones.