An offer pays $100. It converts at 1%, so a click is worth $1.00. Your traffic costs $1.20 a click.
Run 1,000 of those clicks and you get ten conversions, $1,000 in payouts and a $1,200 bill. The $100 payout didn’t save you.
The question isn’t how much the offer pays. It’s how much one click is worth once the funnel has done its work, and how much of that you can hand to the traffic source before the campaign goes red.
The math behind break-even CPC
EPC is what a click earns on average. For a CPA offer:
EPC = payout × conversion rate
Break-even CPC is the most you can pay per click and still finish at zero. In the simplest model it equals EPC.
Use the Affiliate EPC Calculator to check your own numbers. Enter revenue, clicks and total traffic cost to see how your EPC compares with what you paid per click.
Take a $100 payout at a 2% conversion rate. EPC is $2.00. At a $0.80 CPC you keep $1.20 per click. At $1.20 you keep $0.80. At $2.00 you keep nothing, and at $2.40 you lose $0.40 on every click, which is $400 per 1,000.

Figure 1. From payout to break-even CPC. $100 × 2% = $2.00 per click, the theoretical ceiling.
That $2.00 is a ceiling, and a simplified one. It assumes every conversion gets paid, the conversion rate was measured on the same clicks you’re buying, attribution is clean and nothing else costs money. The sections below take those assumptions away one at a time.
EPC vs CPC: the comparison that matters
EPC is what the click earns. CPC is what it costs. The gap, before other costs, is your gross margin per click.
The payout itself is only the starting point. Our analysis of pay-per-conversion AI SaaS affiliate programs looks at how CPA, recurring commissions, conversion events and actual program terms change the economics behind that payout.
Compare two setups. One earns $2 per click on $1 traffic. The other earns $5 per click on $4.80 traffic. The second has the bigger EPC and keeps $0.20 a click instead of $1.00. A 5% dip in its conversions drops EPC to $4.75 and the campaign goes negative. The first one barely notices.
Both numbers also need the same units and the same window. Apogee’s write-up of a merchant program reports a “$159 EPC” on 4,466 clicks, but a footnote defines it as earnings per 100 clicks, which is $1.59 a click [1]. Set that against a per-click CPC without converting and you’re off by a factor of 100. The report is from 2017 and written from the merchant side, so take it as a lesson in units, not a benchmark.
High payout doesn’t mean high EPC
Payout tells you what a conversion pays. It says nothing about how often you’ll get one.
| Offer | Payout | Conversion rate | EPC |
|---|---|---|---|
| A | $20 | 5% | $1.00 |
| B | $100 | 1% | $1.00 |
| C | $200 | 0.25% | $0.50 |
| D | $50 | 1.5% | $0.75 |
| E | $120 | 2.5% | $3.00 |
| F | $300 | 0.3% | $0.90 |
Table 1. Illustrative calculations (payout × conversion rate), not campaign data.
Three offers paying $20, $100 and $200 can earn $1.00, $1.00 and $0.50 a click. The $300 offer at 0.3% earns $0.90, only a little more than a $50 offer at 1.5%, and far less than $120 at 2.5%.
Jane Ejim, who heads ClickDealer’s e-commerce unit, Customer Direct Group, makes the same point from the network side: a $40 payout converting at 1% loses to $15 at 8%. “It’s simple math,” she said in a July 2026 CPV Lab interview [2]. That’s $0.40 a click against $1.20.
Why break-even CPC isn’t your target CPC
If break-even is $2.00 and you buy at $1.95, you’ve left five cents, 2.5% of EPC, for everything that can go wrong: tracking gaps, rejected conversions, a landing page that needs fixing, a creative that dies after a week. Any one of them erases the margin.
How much room you need is specific to the campaign. It depends on how stable your conversion rate is, how much data you’ve collected, what your fixed and variable costs look like and how fast you can cut when numbers move. A fixed rule like “always buy at 70% of EPC” skips all of that.

Figure 2. CPC zones around a $2.00 EPC. Illustrative: your own target depends on your costs.
The number that quietly breaks your EPC: approval rate
A raw conversion isn’t revenue. An approved one is.
Say 100 clicks produce 2 conversions at a 2% rate. At a $100 payout that’s $200 on the dashboard. If 75% are approved, you’re paid for 1.5 conversions: $150, or $1.50 per click. The $2.00 EPC you were tracking is a $1.50 EPC in the bank, and your break-even CPC just dropped by a quarter. Buy at $1.80 and a campaign that looked like +$0.20 a click is losing $0.30.
Approval rules differ by program. Digistore24’s help center says that when an order is returned, the affiliate commission goes back to the vendor, and it calculates cancellation rate on revenue, not on the number of orders [3]. Its blog warns that some digital products carry refund windows of 30 to 180 days, so reversals can land weeks or months after the campaign that produced the sale [4]. At the other end, Apogee’s July 2017 report shows no voided orders for one merchant program [1]. Approval is a property of the program, not a constant.
What the offer page doesn’t tell you
Two offers advertised at “$100 CPA” are different offers if one pays on signup and the other on a funded account. The conversion event drives both your conversion rate and how much gets reversed. Aragon Premium’s network data shows it: shallow events like insurance quote requests converted in the low teens at about $0.50 EPC, while event-based credit-building offers converted under 1% at $1 to $2 [5].
The rest of the fine print matters too: caps, geo and source restrictions, and delays. A PropellerAds-published case study of a media-buying agency notes that early conversions often arrived late, so the team held off cutting too soon [6].
Then there’s attribution. Aragon found that aggregate conversion rates in financial services and education had been inflated into the mid-40s by conversions attributed to clicks tracked elsewhere. At partner level, financial services ran roughly 1.5% to 13% [5]. And ROIads’ Pemex case study lists 524 registrations at a $10 payout and $5,820 in revenue, but 524 × $10 is $5,240. The page doesn’t explain the $580 gap [7]. Reconcile your tracker against the network before you trust any number, published cases included.
Real case: better conversion beat cheaper clicks
DOCUMENTED CASE. In a Mobidea Academy write-up, a publisher who goes by Garywell ran Uber’s iOS install offer on popunder traffic, mostly in the US, at $1.50 to $2.30 per install [8]. His first test, run without whitelists, returned −60% ROI on Adcash and 0% on Clickadu. He then whitelisted the zones that worked, tested a first-ride discount message and reached 80% ROI, then moved to city-specific creatives and targeting, which pushed some cities to 500%.
Here’s the arithmetic. ROI of −60% means $0.40 back per $1.00 spent. ROI of +80% means $1.80. That’s 4.5 times more revenue per dollar. Had the conversion rate stayed flat, clicks would have had to get 4.5 times cheaper to produce it. They didn’t. The message got better and the weak traffic got cut.
Treat this as a pattern, not a benchmark. It’s self-reported, the numbers date from 2017 (the page was updated in 2021), and the zone cuts and creative changes overlap, so the case can’t tell you how much each contributed. The headline claims a 172% conversion-rate increase without publishing before-and-after figures, so don’t lean on that number.
Sometimes the answer to an expensive campaign isn’t cheaper traffic. It’s making the traffic you already have convert.
EPC is an output of the whole system
Todd Stearn, CEO of The Aragon Company, published 90-day network data on September 21, 2026. Insurance EPC sat around $0.50 and financial services reached the low-to-mid teens. The more useful finding is the spread inside a vertical. On the same offers over the same window, financial-services partners earned roughly $3.50 to $23 per click, and insurance partners spanned about 15 times [5].
One insurance partner converting above 30% earned less per click than one converting below 8%. Another logged no tracked conversions and still posted the vertical’s highest EPC, because it was paid on downstream events. Stearn’s advice: “Use vertical benchmarks to orient.” Decide with your own EPC set against your own traffic cost [5]. This is one network’s data from 17 June to 14 September 2026, not an industry benchmark.

Figure 3. EPC is an output of offer, traffic, funnel and conversion event, not a property of the offer. Data: Aragon Premium [5].
The same logic applies on the page. Adam Pasha of Excity Media told WeCanTrack that a high-traffic campaign had weak EPC because affiliate links sat too early in the content. His team moved them later and made the surrounding copy more relevant, and EPC “nearly doubled” [9]. That’s a content publisher, not a paid-traffic buyer, and the interview gives no baseline numbers. The point carries over anyway: change the funnel and EPC moves, and your break-even CPC moves with it.
The same offer on three traffic sources
Same $100 payout, three sources. This is an illustrative scenario, not AffStudio data.
| Source | CPC | Conversion rate | EPC | Margin per click |
|---|---|---|---|---|
| A | $0.70 | 1.2% | $1.20 | +$0.50 |
| B | $1.20 | 2.5% | $2.50 | +$1.30 |
| C | $1.60 | 3.0% | $3.00 | +$1.40 |
Table 2. Illustrative scenario. Margin shown before approval rate and other costs.

Figure 4. Same $100 payout, three traffic sources. Illustrative scenario.
Source A has the cheapest click and the thinnest margin. Source C has the priciest click and the best margin. Ranking sources by CPC alone sends the budget to the wrong place.
Real conversion rates vary by source in the same way. Ejim says her unit sees nano creators converting at roughly 7% against about 3% for macro influencers [10]. Kyle Dana, Director of Digital Strategy and Growth at Digistore24, said in a September 29, 2026 CPV Lab interview that US traffic costs and competition have climbed, which is why US vendors are moving winning offers into Germany, France and Italy, where clicks cost less and audiences are less saturated [11]. Break-even is a per-source, per-geo number.
What other costs do to the calculation
EPC minus CPC isn’t net profit. Tracker fees, prelanders, domains, hosting, creative production, account and infrastructure costs, payment fees and any network fees all come out of the same clicks. Spread over those clicks, they lower the CPC you can afford.
Run the numbers from the approval example. $100 × 2% is $2.00. At 75% approval it’s $1.50. Take $1.00 for traffic and $0.15 per click for everything else, and you keep $0.35. Your real break-even CPC is $1.35, not $2.00.

Figure 5. From raw EPC to margin per click. Illustrative numbers.
When to kill the campaign
Break-even economics tells you what the campaign has to achieve. Statistical confidence tells you whether you’ve seen enough to say it isn’t achieving it. Those are separate questions, and mixing them up is how people kill winners early or feed losers for weeks.
Say the campaign breaks even at a 2% conversion rate. After 100 clicks with no conversions, you can’t conclude it’s below that: if the true rate were 2%, you’d see zero about 13% of the time. After 300 clicks and still nothing, that would happen about 0.2% of the time. Same economics, very different evidence. Our piece on how much to spend before killing an affiliate campaign covers the testing side in detail.
Ejim suggests giving a test 48 to 72 hours and says five clicks isn’t enough to know anything [10]. Time is a stand-in for volume, though. A slow source can take weeks to give you what a fast one gives you in a day.

Figure 6. Decision flow: test, measure, validate, then scale, optimize or stop.
A practical pre-scale checklist
Before you raise spend, know:
- Payout.
- The exact conversion event that triggers it.
- Conversion rate, measured on the clicks you’re actually buying.
- Approval rate, including reversals that land late.
- Effective EPC: payout × conversion rate × approval rate.
- Actual CPC, not your bid.
- Other variable costs per click.
- Break-even CPC: effective EPC minus other costs per click.
- Target CPC: break-even minus the room this campaign needs.
- Tracking consistency between your tracker and the network. If they disagree, start with 12 reasons conversions go missing.
FAQ
What is break-even CPC in affiliate marketing?
The most you can pay per click without losing money. In the simple model it’s payout × conversion rate. In practice it’s effective EPC minus other per-click costs.
How do you calculate break-even CPC?
Multiply payout by conversion rate for theoretical EPC. Multiply again by approval rate for effective EPC, then subtract per-click costs. $100 × 2% × 75% − $0.15 = $1.35.
Is EPC the same as CPC?
No. EPC is what a click earns you. CPC is what you pay for it. Check the unit before comparing: one published report quotes EPC per 100 clicks [1].
What is a good EPC?
There’s no universal number. One network’s 90-day data ran from about $0.50 to the low teens across verticals, and up to 15 times apart inside one vertical [5]. EPC only means something against your traffic cost, conversion event, approval rate and other costs.
Can a high-payout offer still lose money?
Yes. A $200 payout at a 0.25% conversion rate earns $0.50 a click. At a $0.70 CPC, every click loses $0.20.
How much should I pay for an affiliate click?
Less than your break-even CPC, by a margin the campaign justifies. Break-even comes from payout × conversion rate, adjusted for approval and costs. The target leaves room for volatility.





Leave a Reply