If you run growth, you've had this argument in a planning meeting: someone wants to pour the next budget increment into paid, someone else says "what about affiliates," and the conversation dissolves because nobody can compare the two on the same axis. They feel like different categories (and they're priced, paced, and measured differently) so the debate goes in circles.
This post puts them on one axis: cost, risk, and timing of customer acquisition. No "affiliates are free" nonsense (they're not), no "paid is dead" nonsense (it isn't). Just the math a growth team actually needs to allocate budget.
TL;DR
- Paid and affiliate have opposite risk shapes. Paid is spend-then-find-out; affiliate is pay-only-after-it-worked. That single difference drives every other trade-off.
- Affiliate CAC is structurally lower-variance, but its capacity is harder to scale on demand. Paid scales instantly with budget; affiliate scales with relationships.
- Compare on CAC payback, not just CAC. Affiliate's payback is faster because you pay post-conversion, but recurring commissions extend the tail.
- The honest answer is "both, sequenced." Paid for speed and predictability; affiliate for compounding, defensible, low-risk volume.
- Watch the blended-CAC trap: affiliate and paid cannibalize and assist each other, so naive last-touch attribution will mislead your allocation.
The core difference: when you pay
Everything starts here.
- Paid: You pay for impressions or clicks, up front, before you know if any of it converts. Your CAC is discovered after the spend. Risk lives in the gap between spend and outcome.
- Affiliate: You pay a commission, after a real paid conversion has landed. There is no spend on traffic that bounced. Risk lives in whether you can recruit enough productive partners.
So paid is a capital-at-risk channel and affiliate is a revenue-share channel. This is why finance teams treat affiliate commissions almost like COGS and treat paid like a bet.
CAC, head to head
| Dimension | Paid | Affiliate |
|---|---|---|
| When you pay | Before conversion (per click/impression) | After conversion (per sale) |
| CAC variance | High, depends on auction, creative, LP | Low, capped by your commission rate |
| Downside risk | You can spend a lot and acquire little | You pay ~nothing if nothing converts |
| Speed to ramp | Instant, raise budget today | Slow, weeks to months to build roster |
| Scalability ceiling | High, but CAC rises as you scale | Medium, bounded by partner supply |
| Compounding | None, stop paying, traffic stops | High, partner content keeps working |
| Defensibility | Low, competitors outbid you | High, relationships don't transfer |
| Effort type | Media buying / creative ops | Recruiting / relationship management |
The headline: paid buys you predictable, instant volume at variable (and rising) cost. Affiliate buys you slower, compounding, defensible volume at capped, post-paid cost. They are complements, not substitutes.
The CAC math, concretely
Take a $30/mo product, paid CAC of $90 (3-month payback before any churn).
Paid: You spend $90 to (hopefully) land a customer. If your conversion rate dips or CPCs rise, that $90 can quietly become $140, and you've already spent it. Payback is 3 months if the customer sticks.
Affiliate at 30% recurring for 12 months: you pay $9/mo in commission, only while the customer pays you. Your effective acquisition cost is spread across the lifetime and only incurred on revenue you actually collected. If the customer churns in month 2, you stop paying commission in month 2. Your CAC self-corrects to reality.
That's the subtle power: affiliate CAC can't run away from you the way paid CAC can, because it's a percentage of money already in the door. (The trade-off: recurring commissions create a long tail of cost on your best, longest-retained customers, model it, don't ignore it. We break down recurring vs one-time here.)
Where each one wins
Paid wins when you need:
- Speed, a launch, a campaign, a board metric this quarter.
- Predictable, dialable volume you can turn up or down.
- Tight targeting and rapid creative experimentation.
- A channel that works on day one without relationships.
Affiliate wins when you need:
- Low financial risk, you only pay on results.
- Compounding volume that keeps producing after you've paid.
- Defensibility, relationships competitors can't outbid overnight.
- Trust-driven conversion, a creator's endorsement beats an ad for high-consideration products.
If your honest constraint is "I need numbers up by end of quarter," that's a paid problem. If it's "I need a durable channel whose economics improve over time," that's affiliate.
The blended-CAC trap
Here's where growth teams misallocate. Affiliate and paid don't operate in clean silos:
- A prospect sees an affiliate's review, then later clicks your retargeting ad and converts. Last-touch credits paid. You conclude paid is efficient and affiliate is weak, and you defund the channel that actually started the journey.
- Conversely, brand spend lifts the conversion rate of affiliate traffic, and affiliate content feeds your SEO and direct.
Two defenses:
- Look at incrementality, not last-touch. Run a geo holdout or pause-test on each channel and watch total new customers, not attributed ones.
- Track time-to-first-conversion and assisted conversions so you can see affiliate's influence even when paid grabs the last click.
If you allocate budget purely on last-touch CAC, you will systematically over-fund paid and starve the compounding channel. This is the most common, most expensive mistake in the whole comparison.
How to sequence them
For most SaaS growth teams, the answer isn't "pick one", it's an order of operations:
- Use paid to find product-channel-message fit fast. It gives you quick signal on which audiences and hooks convert.
- Feed those learnings into affiliate. The messaging and segments paid proved out become the brief you hand partners.
- Let affiliate compound underneath paid. As partner content accrues, it lowers your blended CAC and reduces dependence on the paid auction.
- Keep paid for the spikes (launches, seasonal pushes, filling pipeline gaps) and let affiliate carry the steady, defensible base.
You end up with paid as your throttle (instant, dialable) and affiliate as your flywheel (slow, compounding, hard to copy).
The tooling footnote
One asymmetry worth naming: paid platforms charge you on spend, and you accept that. But several affiliate tools charge a percentage of every commission on top of subscription, meaning your "low-risk" channel quietly gets taxed precisely as it succeeds. When you're comparing CAC, fold that fee into the affiliate number, or pick flat-fee tooling so the comparison stays honest. (Disclosure: I build Sled, flat $19/mo, no per-commission cut, so weigh my take on that point accordingly. The fee math is here.)
The verdict
Affiliate vs paid is the wrong framing. The right one: paid is how you buy speed and certainty; affiliate is how you build a low-risk, compounding, defensible base. A growth team that runs only paid is renting all its growth. One that runs only affiliate is leaving speed on the table. The teams that win run paid as the throttle and affiliate as the flywheel, and measure both on incrementality, not last-touch.
If you want the operating playbook for standing up the affiliate side, that's the growth team's guide to affiliate programs.
Building the affiliate side of your mix on Stripe, Polar, or Paddle? Sled gives growth teams clean attribution and a hosted partner portal for a flat $19/mo, no per-commission fees.