Almost every "how to start an affiliate program" post is written for a solo founder who installs a tool on a Sunday afternoon. If you're on a growth or marketing team, that's not your problem. You already know how to set up a tool. Your problem is different: you have a channel mix, a CAC target, a board deck, and a finite number of headcount-hours, and someone just asked whether affiliates are worth a slot in next quarter's plan.
This post answers that question, not "what is an affiliate program," but "how does a real team run one as a measurable acquisition channel that compounds."
TL;DR
- Affiliate is the only channel you pay for after the conversion, not before. That makes its CAC structurally lower-risk than paid, but it's slower to ramp and needs an owner, not a side-of-desk volunteer.
- Treat it like a channel, not a feature. Give it a CAC target, a monthly recruiting quota, and a single DRI. Programs without an owner die at ~15 affiliates.
- The metric that predicts success is not "number of affiliates." It's active affiliates (drove ≥1 conversion in 90 days) and revenue concentration. Expect 80% of revenue from your top ~10%.
- Your first 20 partners are a sales motion, sourced from existing power users and adjacent (non-competing) tools, not a marketing broadcast.
- The tooling decision is mostly about fees and attribution, not features. Percentage-of-commission pricing quietly taxes your best channel as it scales.
Why affiliate belongs in a growth team's mix (and why it usually doesn't get a slot)
Every channel a growth team owns has a risk shape. Paid is pay first, hope later, you spend, then find out if it converted. Content and SEO are pay first, wait months, hope.
Affiliate inverts it: you pay a commission only after a real, paid conversion has already landed. There is no spend on traffic that bounces. That's why CFOs like it and why it survives budget cuts that gut paid.
So why doesn't it get a slot? Three honest reasons:
- It ramps slowly. A paid campaign shows signal in a week. An affiliate program shows real signal in a quarter or two, as partners produce content that then accrues SEO and audience trust. Growth teams optimizing for quarterly OKRs under-invest in it for the same reason they under-invest in SEO.
- It looks like ops, not growth. Approving partners, answering "where's my payout" emails, and reviewing for fraud feels like support work. Teams assign it to whoever has spare time, it gets neglected, and then they conclude "affiliates don't work for us."
- The wins are lumpy and attributed elsewhere. A great affiliate drives a burst, then quiet, then another burst. In a last-touch model, half their influence gets miscredited to "direct" or "organic."
None of these are reasons to skip it. They're reasons to resource it correctly.
The mental model: affiliate is a portfolio, not a funnel
Paid acquisition is a funnel you optimize. Affiliate is a portfolio you build. You're recruiting a roster of partners, most of whom will produce roughly nothing, a handful of whom will carry the channel. This is not failure, it's the literal shape of the channel, the same power law you see in VC, content, and app stores.
Two consequences that change how you should plan:
- Recruiting is the job. Since output is concentrated in the top decile, your expected revenue scales with how many quality partners you can get to first conversion. A recruiting quota (e.g. "10 qualified partners activated per month") is a better leading indicator than any dashboard vanity metric.
- You can't predict your winners up front. The partner with 50k followers often underperforms the one with a 1,200-person niche newsletter whose audience trusts them. So you recruit broadly within fit, then double down on whoever actually produces.
The metrics that actually matter
Kill "total affiliates" as a headline number on day one. It rewards collecting dead signups. Track these instead:
| Metric | What it tells you | Healthy signal |
|---|---|---|
| Active affiliates (≥1 conversion / 90d) | Real channel size | Growing month over month |
| Activation rate (signups → first conversion) | Quality of recruiting + onboarding | 20–40% for targeted recruiting |
| Revenue concentration (top 10% share) | Portfolio health | 60–80% is normal, not a problem |
| Blended affiliate CAC | Commission paid ÷ new customers | At or below your paid CAC |
| CAC payback | Months to recover commission | Faster than paid, since you pay post-conversion |
| Time-to-first-conversion | Onboarding friction | Shorter = better assets/links |
The two that predict whether the channel compounds: activation rate (are recruited partners ever producing?) and active affiliate growth (is the productive roster expanding?). Everything else is downstream.
Who owns it: the part teams get wrong
The single biggest predictor of an affiliate program's success isn't the tool or the commission rate. It's whether one named person owns it. Programs run "by the team" plateau around 15 affiliates and quietly rot.
The owner (call it a Partnerships or Growth Marketing DRI) runs three loops:
- Recruiting loop (weekly): source and pitch a target number of new partners. This is outbound. It looks like sales because it is.
- Activation loop (per new partner): get them their link, give them assets that actually convert, and remove every excuse not to post.
- Retention loop (monthly): check in with top producers, ship them new creative, pay them fast and correctly. Your best partners get poached by competitors who treat them better.
You don't need a full headcount on day one. You need ~25–40% of one competent person's time, with a quota and a number to hit, not "manage the affiliate inbox when you get a sec."
Recruiting your first 20 partners (as a team, not a solo founder)
The advice "post your program in affiliate directories" is dead. Here's the motion that works at team scale:
- Mine your own product data. Pull your power users, high usage, NPS promoters, anyone who's tweeted about you. They already advocate; you're just adding a reason and a link. Highest activation rate of any source, by far.
- Map adjacent, non-competing tools. List products your ICP already uses alongside you. Their creators, agencies, and consultants are your highest-fit partners. Reach out person-to-person, not via a "Dear Affiliate" blast.
- Recruit the consultants and agencies in your category. They recommend tools for a living. A 20–30% recurring commission turns your product into part of their revenue model.
- Make the ask specific and personal. Lead with why them, reference their actual work, and put the commission as the close, not the hook. Budget ~30 quality touches to land 10 active partners. That's a normal sales conversion rate, and it's why this is a sales motion.
Skip influencer cold-blasts and directory submissions until you have product-channel fit proven by your first cohort.
The commission structure decision
Two questions settle most of it:
- Recurring or one-time? For subscription products, recurring commission (paying on renewals, often for 12 months or lifetime) is what attracts serious partners, because it builds them an income stream. One-time payouts attract coupon-site traffic, not advocates. (We wrote the full breakdown here.)
- What rate? 20–30% is the SaaS norm. The exact number matters less than whether it's meaningful enough that a creator will spend an afternoon making content. Below ~15% recurring, most won't bother.
Set a cookie/attribution window you can defend (30–60 days is standard) and write your program terms before you launch, not after the first dispute.
The tooling decision (the boring part that compounds)
For a growth team, feature checklists matter less than two things:
- Attribution you trust. If you can't cleanly tie a conversion to the partner who drove it, every other metric above is noise, and partners will dispute payouts. Get clean tracking across your billing provider (Stripe, Polar, Paddle, or webhooks) before you scale spend.
- Pricing that doesn't tax your best channel. Several incumbents charge a percentage of every commission on top of the subscription. That's a tax that grows precisely as the channel succeeds, the worse the incentive, the more it punishes your wins. A flat fee means the channel's economics improve as it scales, instead of degrading. (Here's what that % actually costs as you grow.)
Full disclosure: I build Sled, which is flat $19/mo with no per-commission fee and native Polar/Stripe support, so I'm not a neutral party on point 2. The principle stands regardless of which tool you pick: don't run your most CAC-efficient channel on pricing that scales against you.
The 90-day plan
A realistic ramp for a team starting from zero:
| Phase | Weeks | Focus | Target |
|---|---|---|---|
| Stand up | 1–2 | Pick tool, wire attribution, write terms, set commission | Tracking verified end-to-end |
| Seed | 3–6 | Recruit from power users + adjacent tools | 10–15 partners, first conversions |
| Activate | 7–10 | Ship creative, fix onboarding friction, pay fast | 30–40% activation rate |
| Double down | 11–13 | Identify top producers, give them more, recruit lookalikes | Active roster growing, CAC ≤ paid |
If by day 90 you have a handful of genuinely active partners and a blended CAC at or under paid, you've found the channel. Now it compounds, every piece of partner content keeps working long after you paid for it.
The honest verdict
Affiliate is not a quick-win channel and anyone selling it as one is lying to you. It's a compounding, low-financial-risk channel that rewards teams who resource it like a real channel and punishes teams who treat it like an inbox.
If you can give it one owner, a recruiting quota, a CAC target, and a quarter of patience, it becomes one of the few channels whose unit economics get better as it grows. If you can't commit that, don't half-start it, a neglected program just generates payout emails and a bad conclusion.
Running affiliates for a SaaS on Stripe, Polar, or Paddle? Sled gives you clean attribution and a hosted partner portal for a flat $19/mo, no per-commission fees. See how it compares.