Cost Per Sale vs Cost Per Click: What Brands Should Pay For, and What Creators Should Charge

Cost per sale and cost per click are two honest ways to pay for the exact same visitor, and picking between them is one of my favorite conversations in this industry, because it’s really one question wearing a dozen costumes: who carries the conversion risk? Under cost per sale (CPS, also called pay per sale), the brand pays a commission only when a purchase completes. Under cost per click (CPC), the brand pays for the click itself, sale or no sale. Every other term in the negotiation, the fee, the listicle position, the contract length, flows downstream from that one question.

Quick note on where I sit: I’ve spent about 10 years founding SaaS companies, and affiliate programs kept turning out to be a big part of why they worked. I built native affiliate programs from scratch for my own products, then started Rekomi to make the software version with everything I’d ever wanted in one. I like cost per sale for a simple reason: you only pay for what you actually get, and it’s genuinely good for creators too, because recurring commissions become real monthly recurring revenue for them. But here’s the part I enjoy most: established blogs that charge flat placement fees or CPC are running a smart business, and brands that buy those placements are sometimes getting the cheapest acquisition available anywhere. Both sides negotiate better when they can see the other’s math, so let’s put all of it on the table.

And this article really is the math: plain definitions, a break-even formula you can apply in one line, real click-through data for pricing listicle positions, and a three-way revenue comparison on the same 10,000 visitors. Every number is either cited at its source or recomputable from assumptions I state out loud, because that’s the deal I want with you: see my math, check my math.

TL;DR:

Cost per sale: the brand pays a commission only on completed purchases, so its acquisition cost is capped by construction at commission rate × order value.

Cost per click: the brand pays for traffic whether or not it converts, taking on the conversion risk plus click-fraud exposure.

The break-even rule: a CPC placement beats a CPS commission only when CPC < commission % × AOV × conversion rate.

For creators: flat or CPC placement fees are a smart hedge (predictable revenue that doesn’t depend on the brand’s checkout), and when the sides disagree, a hybrid of a reduced flat fee plus a CPS kicker is usually the deal both should want.

Cost per sale vs cost per click: the two models, plainly

What is cost per sale?

Cost per sale is the amount a brand pays for each completed purchase: total spend divided by the number of sales it produced. If a campaign costs $5,000 and generates 200 sales, the cost per sale is $5,000 ÷ 200 = $25. That’s the cost per sale formula in its general form, and it’s how you calculate cost per sale for any channel, paid or organic.

Affiliate marketing runs a stricter, and honestly more elegant, version. The brand doesn’t spend first and divide later; it promises a commission per sale up front, so the cost per sale is fixed before a single click happens: CPS = commission rate × average order value. A 20% commission on a $100 order is a $20 cost per sale, written into the agreement. And if the customer refunds, a properly built program reverses the commission, so the realized cost of a junk sale rounds to zero.

Two properties of CPS drive everything else in this article, and if you’re the brand, they’re both wonderful:

• Payment happens after revenue exists. The brand never pays for traffic that didn’t buy.
• The conversion risk sits with the partner. An affiliate who sends 10,000 clicks that never convert earns nothing for them.

What is cost per click, and who uses it?

Cost per click is the price paid for a single click, calculated as total spend divided by total clicks: $300 spent for 400 clicks is a $0.75 CPC. It’s the model search and display advertising run on, and it’s also how some publishers price direct placement deals, where the brand pays for every reader who clicks through regardless of what happens on the landing page.

Under CPC the risk positions flip exactly. The publisher gets paid on delivery of the click; the brand owns whatever happens next, including nothing. And if you arrived here searching “what is a good cost per click,” here’s the honest answer: a good CPC is one that produces a cheaper sale than your CPS alternative, and the break-even formula below tells you exactly where that line sits.

Where do CPA and CPL sit between CPC and CPS?

CPA and CPL are the middle rungs of a ladder that transfers conversion risk from the brand to the partner one step at a time. The CPA vs CPC question is really a question about which rung you’re standing on.

The naming overlap trips people up, so here’s the clean version. CPA vs CPC is simple: CPA pays on a conversion, CPC pays on a click. CPA vs CPS is fuzzier because “acquisition” can mean any event the brand defines (a lead, an install, a demo booked), while cost per sale means a completed purchase specifically; every CPS deal is a CPA deal, but not the reverse. Cost per lead and cost per conversion are useful internal metrics, but when money changes hands between a brand and a partner, the contract should name the exact event that triggers payment.

The brand’s view: why cost per sale is the structural winner, and when CPC still makes sense

What does paying cost per sale buy a brand?

A capped acquisition cost, agreed in advance, funded entirely out of realized revenue. I’ll say it plainly: I find that close to beautiful. At a 20% commission, your CAC is 20% of order value on every order, by contract; spend can never run ahead of sales because spend is a function of sales. Pay per sale also scales without a budget conversation: more commission owed means more revenue arrived, which is the one cost overrun nobody escalates.

Refund reversals complete the picture. A program that reverses commissions on refunds and chargebacks pays for kept revenue, not gross revenue, so a customer who buys and refunds costs you no commission at all.

The three risks CPC hands a brand

CPC is a perfectly honest model; it just hands the brand three risks at once, and they compound:

• No outcome guarantee. A CPC budget can burn to zero with no sales attached. The publisher fulfilled the contract; you got clicks. Nothing in the model requires that anyone buy.
• Traffic quality you can’t fully audit. Fraud Blocker’s analysis of 96 million clicks across 85,000 Google Ads accounts found an average invalid click rate of 11.5% (Fraud Blocker). The same report estimates roughly 22% of global digital ad spend is lost to fraud, about $84B in 2023 and projected to reach $133B by 2026. Those are one vendor’s figures, not an industry consensus, but even at half the rate the point stands: some fraction of the clicks you buy were never a human with intent.
• Incentive drift. A partner paid per click optimizes for clicks. Curiosity-bait anchor text, prominent buttons, link-shaped furniture everywhere. None of it is dishonest, but none of it selects for buyers either.

When does a CPC or flat placement genuinely make sense for a brand?

Four cases, and they’re common enough that I’d never tell a brand to refuse placements on principle:

• Guaranteed distribution on a page that already ranks. If a listicle owns the SERP your buyers search, a placement buys you tomorrow what SEO would take 18 months to maybe deliver.
• Competitive defense. Slots on a ranking page are scarce. If you pass, your competitor takes it, and now the page that owns your keyword recommends them.
• Awareness. Sometimes the impression is the point, and paying per view or per click prices it honestly.
• Speed. A placement can be live next week. Recruiting affiliates who produce equivalent volume takes months.

CPC vs CPS: when does the placement beat the commission?

A CPC placement is the better buy when: CPC < commission % × AOV × conversion rate. Here’s where the inequality comes from: a placement’s effective CAC is the click price divided by your conversion rate, and a CPS deal’s CAC is the commission rate times average order value. Write both models as a cost per acquired customer and the comparison collapses to that one line.

Run it both ways: the same formula produces opposite answers on realistic inputs.

Example A, CPS wins. A brand pays $2.00 per click on a placement, converts visitors at 2%, sells a $100 AOV product, and its affiliate program pays 20%. Placement CAC: $2.00 ÷ 0.02 = $100 per sale. CPS CAC: 20% × $100 = $20 per sale. The commission is 5x cheaper. The break-even click price here is 0.20 × $100 × 0.02 = $0.40, so at anything above $0.40 per click, this brand is overpaying versus its own affiliate program.

Example B, the placement wins. A brand negotiates $0.60 per click on a high-intent “best X” listicle whose readers convert at 8%, on a $250 AOV product, with the same 20% commission. Placement CAC: $0.60 ÷ 0.08 = $7.50 per sale. CPS CAC: 20% × $250 = $50 per sale. The placement is more than 6x cheaper, and the break-even CPC is 0.20 × $250 × 0.08 = $4.00, so $0.60 is a genuine bargain!

Same formula, opposite verdicts, and I love how cleanly it falls out of two inputs. The swing variables are conversion rate and AOV, and notice who holds that information: the brand knows both numbers and the creator usually knows neither. Keep that asymmetry in mind; it explains most of what creators do next.

The creator’s view: why charging per click or per slot is smart business

Why do established blogs charge flat fees or CPC instead of commissions?

Because a commission is a variable claim on a machine the creator can’t inspect. Under a CPS deal, the blog’s revenue depends on the brand’s checkout conversion rate, its attribution setup, the cookie window, and its refund rate. All four are invisible from the outside, and all four can quietly change after the deal is signed.

A flat monthly fee or a CPC rate removes every one of those dependencies, and I respect the logic:

• Predictable revenue. Affiliate earnings per click swing with seasonality, landing page changes, and pricing tests the creator never hears about. A flat fee pays the same in February as in November.
• No attribution trust required. If the brand’s tracking undercounts, a CPS creator eats the loss silently and can rarely prove it. A CPC deal is settled on the creator’s own click logs.
• Refund immunity. CPS commissions reverse when customers refund. A placement fee is earned when the audience is delivered, which is the part the creator actually controls.

Brands should recognize this logic instantly; it’s why they pay salespeople a base salary instead of pure commission. Asking a publisher to take 100% outcome-based pay is asking them to underwrite your funnel sight unseen. Some will, at a price; the established ones with ranking pages usually won’t, and they’re not wrong.

The same trust math explains why CPC affiliate marketing, where the program pays per click instead of per sale, survives only where the publisher’s traffic is auditable. Nobody keeps paying for clicks they can’t verify came from humans.

How should a creator price listicle positions?

By click share, because attention inside a ranked list concentrates brutally at the top. The best public proxy we have is organic click-through rate by SERP position, and FirstPageSage publishes the most-cited figures (last updated May 28, 2025): position 1 earns a 39.8% CTR, position 2 earns 18.7%, position 3 earns 10.2%, position 4 earns 7.2%, position 5 earns 5.1%, and position 10 earns 1.6% (FirstPageSage). One caveat to carry with you: those figures describe SERPs without map packs or AI overviews, and a listicle isn’t literally a SERP. Treat the curve as a proxy for how attention decays down a ranked list, not a guarantee.

The pricing consequence is blunt, and I mean this as a compliment to the bloggers who’ve already figured it out. Position 1 draws roughly 7.8x the clicks of position 5 (39.8 ÷ 5.1), so if the #5 slot in your “best X” roundup is worth $500/mo, the #1 slot is worth about $3,900/mo on click value alone. A rate card built off the same multiples:

If I ran a content site, here’s what I’d want to know: rate cards in the wild are usually far flatter than this curve; 2x for #1 over #5 is common. If that’s you, your top slot is underpriced by multiples, and the sponsored listicle placement market will happily keep buying it from you at a discount. Reprice it; the math is on your side.

What should a creator check before accepting CPS instead?

Five things, and a brand that answers them readily is a brand worth taking upside on:

• The brand’s actual conversion rate. Ask for it. A serious program will share landing-page-to-paid numbers or offer a paid test month so you can measure your own.
• The attribution window. A 7-day cookie on a considered purchase quietly donates your conversions to “direct”; 30 to 90 days is normal for SaaS. And remember Safari caps script-set cookies at 7 days regardless of what the program promises, so ask how the cookie is set, too, and whether conversions are confirmed by billing events rather than a pixel; my teardown of tracking architecture explains exactly what to look for.
• Recurring vs one-time. A 20% recurring commission on a subscription is a different asset class from a 20% one-time bounty, and I mean that in the best way. The payout mechanics matter as much as the headline rate.
• Refund policy and reversal mechanics. Reversals are fair, but you want them itemized, not netted invisibly out of a lump sum.
• Tracking you can trust. Pixel-based tracking dies to ad blockers and misses renewals; billing-verified tracking reads the sale from the payment processor itself. That gap is the design choice behind Rekomi’s tracking, which sits on billing events instead of pixels.

Then convert the offer into an expected earnings per click and compare it to your placement rate: expected EPC = conversion rate × commission % × AOV. If the terms in that sentence are new, my EPC walkthrough covers the formula, the reporting conventions, and the recurring-revenue adjustment that makes subscription offers look better than their first month suggests.

The AdSense comparison: the same 10,000 visitors, three ways

I’ve been waiting to get to this one, so let me set the assumptions carefully. One content page, 10,000 buyer-intent visitors a month, three ways to monetize it. All three sets of numbers below are illustrative constructions from cited inputs, not study results.

Option 1: display ads. Premium ad management networks are the realistic ceiling here. One publisher’s first-party comparison reports a Mediavine travel site running $22 to $44 RPM across the year (about $33 to $35 on average) and $47 to $57 on Raptive in Q4 2025, with a second, smaller niche site earning about $9 on Mediavine Journey and $12 to $15 after moving to Raptive (This Week in Blogging). Raptive accepts sites at 25,000 monthly pageviews and advertises a 15% RPM lift guarantee for larger sites (100k+ pageviews) (Raptive).

AdSense alone typically pays materially less than these managed networks, so call a typical content site $15 to $35 RPM outside Q4 (RPM meaning revenue per 1,000 pageviews, per Google’s definition). On 10,000 pageviews: $150 to $350.

Option 2: a direct CPC placement. A brand pays $0.75 per click on its link in the page. If 5% of readers click, that’s 500 clicks: $375, regardless of whether anyone buys.

Option 3: CPS affiliate marketing. Same 500 clicks go to a brand that converts 6% of them at a $150 AOV with a 20% commission: 30 sales × $30 = $900.

Now read the crossover, because that’s where the real decision gets made. On low-intent traffic, display wins: there’s nothing to convert, so getting paid per impression beats getting paid per outcome that won’t happen. On buyer-intent pages, CPS wins by multiples, but only if the offer converts.

Cut that 6% conversion rate to 1.5% and the CPS revenue drops to 7.5 sales × $30 = $225, below the $375 placement fee. That sensitivity is exactly the variance creators are pricing against when they ask for flat fees, and it’s why the CPC placement sits between the other two as the hedge: more than display on intent traffic, no exposure to the brand’s funnel. Every model in this table earns its spot; they’re just pricing different risks.

Everything in that CPS estimate assumes each sale gets recorded honestly, and that assumption is about tracking architecture, not good intentions. That’s why billing-verified attribution matters: commissions computed from the payment processor’s own events mean refunds reverse automatically and subscription renewals keep accruing to the referring partner, so both sides are settling on real revenue instead of a pixel’s guess. It’s the difference between a partner who trusts your $900 column and one who demands the $375 flat fee, and it’s the problem Rekomi’s pricing is built around: a flat monthly plan plus a small percentage that only accrues as payouts actually flow.

Should a brand pay a blog that charges for placement?

Sometimes yes, and happily, the deciding math is a single division. The result surprises most brands: a blog charging thousands per month can still be their cheapest channel.

What is the effective CAC of a paid placement?

Effective CAC = placement fee ÷ expected conversions. Worked example: a $2,000/mo slot on a page that sends you 800 clicks a month, at your measured 5% conversion rate, produces 40 sales. Effective CAC: $2,000 ÷ 40 = $50. Now compare that against your other channels: your CPS commission CAC (20% × $150 AOV = $30 in the running example) and your actual paid search CAC from your own ad account, which for plenty of competitive B2B categories runs well past $50. The placement loses to your affiliate program per sale, but if affiliate volume is capped and search is more expensive, $50 can still be the best marginal dollar you spend, with the awareness value riding along free. I’ve watched brands walk away from exactly this deal because the sticker price felt high; the division says otherwise.

When are positions 1 to 3 worth the premium, and when is position 5 the value buy?

Pay the premium for positions 1 to 3 when the page already owns the SERP your buyers search, because the CTR concentration from the rate card section means most readers never engage below the top of the list. Buying #1 on the page that ranks #1 is buying the single highest-intent click stream in your category, and there’s exactly one of it.

Position 5 is the value buy when rate cards are flatter than the click curve, which they usually are. If #1 costs 4x the #5 price but delivers 7.8x the clicks, #1 is the bargain; if #1 costs 10x, position 5 wins on $/conversion. Run the effective-CAC division on each slot offered and buy the cheapest sale, not the highest position.

Which negotiation levers make both sides better off?

• The hybrid deal. This is the one I get excited recommending, because it’s the win-win more often than either pure model. Instead of $2,000 flat, structure $800 flat plus a 10% commission. At the same 40 sales × $150 AOV, the creator earns $800 + $600 = $1,400 with genuine upside if the brand converts well; the brand’s CAC becomes $35 with downside capped at $800 even if conversions crater. The creator keeps a floor, the brand caps its risk, and both now want the same thing: conversions!
• A pilot month at a reduced rate, with both sides reviewing click and conversion data before committing to a year.
• A position-upgrade option, locking today’s price for a future move up the list.
• Reporting rights. Clicks visible to the brand, conversions visible to the creator. Deals thrive when both sides can see; they fall apart when one side is blind.

What compliance duties do paid placements carry?

Two, and both are easy to get right. First, the FTC requires disclosure of material connections between an endorser and a brand, which explicitly includes payment, free or discounted products, and personal, family, or employment relationships (FTC, Disclosures 101). A paid listicle slot is a material connection, period, and the disclosure duty applies whether the payment was a flat fee, a CPC rate, or a commission.

Second, Google’s guidance is that links which are “advertisements or paid placements (commonly called paid links)” must carry the sponsored value, with nofollow also acceptable (Google Search Central). Here’s my one hard stance for the brand side: a brand paying extra for followed links is buying a link-spam penalty risk with a receipt attached, not SEO value. And a creator selling followed placements is wagering the ranking page their entire placement business stands on. Mark the links rel=”sponsored” and compete on the click stream, which is what you were actually buying. Everyone sleeps better.

The decision framework

For brands:

• Run cost per sale as the backbone. Pay for outcomes by default; it’s the only model where CAC is capped by contract. If that program doesn’t exist yet, my program-launch guide for SaaS is the long version, and it’s a fun build.
• Buy placements selectively, where the page already owns the SERP you want and the break-even math (CPC < commission % × AOV × conversion rate) clears with room to spare.
• Never sign CPC with partners whose traffic you can’t audit. Remember the 11.5% invalid-click figure; on an unauditable partner you have no idea if your number is better or worse.

For creators:

• Price top positions like the real estate they are. The click curve says #1 is worth about 8x #5; don’t sell it for 2x.
• Take CPS upside where the brand demonstrably converts and the tracking is billing-verified rather than pixel-based. That combination is where commissions out-earn any flat fee, especially recurring ones.
• Push hybrids when uncertain. A floor plus a kicker converts a trust problem into a shared incentive, which is a genuinely nice trick.

Five questions that come up in every CPS vs CPC negotiation

Is cost per sale the same as cost per acquisition?

Not quite. Cost per acquisition (CPA) measures the cost of any defined conversion event, which might be a lead, an install, or a trial start; cost per sale measures the cost of a completed purchase specifically. Every CPS is a CPA, but a $40 CPA on free-trial signups is very different money from a $40 CPS on paid orders. (One naming note: “Shopify cost per sale” usually refers to Shopify’s transaction and processing fees, a platform cost, not this marketing metric.)

What is the difference between cost per click and cost per acquisition?

Cost per click prices the click; cost per acquisition prices the outcome. They connect through conversion rate: CPA = CPC ÷ conversion rate. A $1 CPC at a 2% conversion rate is a $50 CPA, which is why a cheap click isn’t necessarily a cheap customer.

What is a good cost per click?

A good CPC is one below your CPS-equivalent cost, which equals commission % × AOV × conversion rate. For a brand paying 20% on a $150 AOV converting at 5%, any CPC under $1.50 acquires customers cheaper than its own affiliate program; above that, the commission is the better deal. There’s no universal benchmark that beats your own math, and your own math takes two minutes.

Should affiliates be paid per click or per sale?

Per sale, in almost all cases, which is why CPC affiliate marketing is rare and pay per sale affiliate programs are the industry default. Paying affiliates per click invites click fraud and rewards traffic volume over traffic quality; commissions align the affiliate’s income with the brand’s revenue, and aligned incentives are the whole magic of this channel. The exception is an established publisher with an audited, ranking page, where a click fee is really a placement fee under another name.

How much should a listicle placement cost?

Anchor it to click share by position. Using published CTR-by-position data as the proxy, position 1 draws roughly 7.8x the clicks of position 5, so a fair rate card scales the same way: if #5 rents for $500/mo, #1 justifies around $3,900/mo. From the brand side, sanity-check any quote with fee ÷ expected conversions and compare the result to your other channels’ CAC.

If you land on pay-for-outcomes

Before your next partnership conversation, spend ten minutes building your own break-even card: your commission % × AOV (that’s your CPS ceiling), your measured conversion rate, and the resulting maximum CPC you should ever agree to. It’s a genuinely enjoyable ten minutes, because you walk out knowing exactly which deals are good for you. Brands that show up with those three numbers negotiate placements well; creators that show up with their click data and a rate card priced off the CTR curve negotiate even better, and the best conversations happen when both sides do.

And if the outcome side of your program is the part you want rock-solid, where sales get tracked, refunds get reversed, and commissions actually match billing, Rekomi is that layer, and 14 days of trial is long enough to watch real commissions reconcile against real invoices.