Arbitrage ROI Calculator: Find Your Break-Even CPC for E-Commerce Offers

What arbitrage ROI actually measures

In paid-traffic arbitrage you do not own the product or the checkout. You buy clicks from an ad network, send those visitors through a content page or pre-sell page, and then pass them to an e-commerce offer that pays you either a commission or a margin. Every campaign decision reduces to one comparison: what does an average click cost, and what does an average click earn? If the second number is larger, you have a business. If it is smaller, no amount of creative testing will save the campaign unless it moves one of the inputs in the formula below.

That is why break-even CPC is the most useful single number in arbitrage. It converts a messy funnel into a bid ceiling. Once you know that a click is worth, for example, 18 cents after refunds and fees, you can instruct your bidding tool never to pay more than that, and you can judge every traffic source against the same yardstick. The calculator above computes it for you, and the sections below show every step by hand so you can audit the result.

The math, step by step

All formulas below use rates as decimals (a 3% conversion rate is 0.03). The calculator accepts percentages and converts them for you. Five quantities build on each other.

1. Offer EPC

Offer EPC is the average earnings per click that reaches the offer. Your affiliate network or tracker may report it directly. If it does not, derive it from three inputs:

Offer EPC = Conversion rate x AOV x Payout share

Payout share is the commission percentage when you promote someone else's store. If you sell the product yourself through a dropshipping or private-label store, use your gross margin percentage after product cost and shipping, because that is the money left to pay for ads.

2. Revenue per ad click

You pay for the click on your ad, not the click to the offer. When you use a pre-sell page, only a fraction of visitors continue to the merchant. That fraction is the landing-page CTR. If you send traffic straight to the offer, set it to 100%.

Gross revenue per ad click = Landing-page CTR x Offer EPC

3. Net revenue per ad click

Refunds and chargebacks reverse revenue, and fees reduce what you keep. Both scale with revenue, so they act as survival rates multiplied together:

Net revenue per click = Gross revenue per click x (1 - Refund rate) x (1 - Platform fee)

4. Break-even CPC, profit and ROI

Break-even CPC = Net revenue per click
Profit per 1,000 clicks = 1,000 x (Net revenue per click - CPC)
ROI = (Net revenue per click - CPC) / CPC
Max CPC for a target ROI = Net revenue per click / (1 + Target ROI)

Break-even CPC equals net revenue per click because profit is zero exactly when cost and revenue are equal. ROI divides by CPC, so a CPC of zero would be a division by zero. The calculator reports ROI as undefined in that case rather than showing infinity or a misleading number. It also refuses to run when landing-page CTR is zero or any percentage is out of range, and it tells you which field to fix.

5. Required conversion rate for a target ROI

Rearrange the chain to solve for the conversion rate. You need net revenue per click equal to CPC times one plus your target ROI, so:

Required CR = CPC x (1 + Target ROI) / (CTR x AOV x Payout x (1 - Refund) x (1 - Fee))

The denominator is the revenue you would earn per ad click at a 100% conversion rate. If it is zero, because AOV, payout or CTR is zero or refunds and fees consume everything, no conversion rate can reach the target, and the calculator says so. In EPC mode the same rearrangement gives a required offer EPC instead, since conversion rate is not a separate input there.

Worked examples (illustrative)

Every figure in this section is an illustrative example chosen for clean arithmetic. None is a benchmark or a recorded campaign result. Use them to check that you understand the formulas, then enter your own numbers above.

Example A: a healthy spread

Assume a CPC of $0.12, a landing-page CTR of 35%, an offer conversion rate of 3%, an AOV of $80, a payout share of 25%, an 8% refund rate and a 3% platform fee. Offer EPC is 0.03 x 80 x 0.25 = $0.60. Gross revenue per ad click is 0.35 x 0.60 = $0.21. After refunds, 0.21 x 0.92 = $0.1932. After fees, 0.1932 x 0.97 = $0.1874. That is the break-even CPC.

Paying $0.12 leaves $0.0674 per click, or $67.40 per 1,000 clicks, which is a 56.2% ROI. For a 100% ROI target you would need a conversion rate of 0.12 x 2 / (0.35 x 80 x 0.25 x 0.92 x 0.97) = 3.84%. For plain break-even you need only 1.92%, so this offer has roughly a 36% cushion on conversion rate (3.00% against 1.92%) before it turns unprofitable.

Example B: EPC mode with no fees

Assume the network reports an offer EPC of $0.45. Landing-page CTR is 25%, refunds are 5%, there is no platform fee, and CPC is $0.08. Gross revenue per click is 0.25 x 0.45 = $0.1125, and net is 0.1125 x 0.95 = $0.1069. Profit per 1,000 clicks is $26.88 and ROI is 33.6%. To hit a 50% target ROI at the same CPC you would need an EPC of 0.08 x 1.5 / (0.25 x 0.95) = $0.505, so this offer falls short and you would negotiate a higher payout or improve the pre-sell page before scaling.

Example C: a campaign that cannot work

Assume a $0.30 CPC, 20% landing-page CTR, 2% conversion rate, $50 AOV, 30% payout, 10% refunds and a 5% fee. Offer EPC is 0.02 x 50 x 0.30 = $0.30, gross revenue per click is $0.06, and net revenue per click is 0.06 x 0.90 x 0.95 = $0.0513. Profit is minus $248.70 per 1,000 clicks and ROI is minus 82.9%. Breaking even would take a conversion rate of 11.7%, almost six times the assumed 2%. When the required rate is that far above anything plausible, the fix is not optimization; it is a cheaper traffic source, a better-paying offer, or walking away.

Scenario table for Example A

The table repeats Example A at different CPCs, holding net revenue per click at $0.1874. All values are illustrative.

Notice the asymmetry. Raising CPC from $0.12 to $0.15, a 25% increase, cuts profit nearly in half, from $67.40 to $37.40 per 1,000 clicks. Small bid changes move profit a lot when the spread is narrow, which is why a bid cap derived from break-even CPC is worth setting before launch.

Where to get each input

A calculator is only as honest as its inputs. Pull each one from a source you can defend, and prefer your own tracked data to anyone's averages.

• CPC: read it from your ad platform's reporting for the same traffic source and geography you plan to scale. Blended CPC across countries hides the fact that cheap clicks often convert worst.
• Landing-page CTR: measure it in your tracker as clicks to the offer divided by landing-page visits. Do not use the ad's own CTR, which measures a different step.
• Offer EPC or conversion rate: early EPC figures are noisy. At a 3% conversion rate you need about 1,000 clicks to the offer just to expect 30 sales, and ratios built on 30 sales still swing widely. Treat early results as a range and test the low end in the scenario tables.
• AOV and payout: use the merchant's actual commission terms. Tiered payouts, cookie windows and product-category exclusions change effective payout, so check the program terms rather than the headline percentage.
• Refund and chargeback rate: use a rate from reversed commissions over a meaningful window. Reversals usually arrive weeks after the sale, so a campaign that looks profitable in week one can drift lower.
• Platform fee: add every percentage-based cut between the shopper's card and your bank account, such as network fees, tracker costs allocated per click, and payment processing if you run your own checkout.

Payment fees: a fixed amount that percentages hide

If you run your own store, card processing is a real cost. Per Stripe's public pricing page, its standard US rate for domestic cards was listed as 2.9% plus 30 cents per successful transaction when this page was written; confirm the current rate and your own contract, since negotiated and regional pricing differs. The fixed 30 cents matters more on small orders. On an $80 order the effective rate is 2.9% + 0.30 / 80 = 3.28%. On a $20 order it is 2.9% + 0.30 / 20 = 4.4%. If your AOV is low, enter a higher platform fee than the headline percentage, or the calculator will flatter the campaign. Dispute fees vary by processor and are separate again, so treat them as part of your refund and chargeback assumption.

Common mistakes that make arbitrage look better than it is

Ignoring the second click. A common error is treating the offer's EPC as if it were earned on every ad click. If only a third of your visitors continue past the pre-sell page, revenue per ad click is about a third of the offer EPC, and a bid cap built on the unadjusted figure loses money from day one.

Using revenue instead of profit. A payout share of 100% only makes sense if the offer pays you the entire order value. For your own store, margin after product cost and shipping is the correct figure; otherwise the break-even CPC will be too high.

Skipping the refund lag. Reversals can land long after the spend. Hold back a reserve in your reporting, or run the calculator with a refund rate a few points above what the first week suggests.

Optimizing for ROI percentage alone. A 200% ROI on a few dollars a day is not the same business as a 40% ROI on thousands. Use the profit per 1,000 clicks figure alongside ROI, and remember that scale usually forces you to bid higher, which compresses the spread in the first scenario table.

Treating one scenario as the forecast. Use the second scenario table. If a 20% drop in revenue per click flips the campaign to a loss, you do not have a safety margin, you have a coin flip.

Using traffic-cost data alongside this calculator

If you are still deciding which network to buy from, our native ad benchmarks page for Taboola and Outbrain lists CPC, CTR and CPA figures for e-commerce and SaaS. We have not attached primary sources to those figures, so treat them as directional starting points to test against your own tracker, not as verified industry averages. A sensible workflow is to take a CPC range from that page or a network's own estimate, run it through the calculator with your offer data, and see how much conversion rate you need before spending a cent.

Once a test budget confirms your numbers, return to the scenario tables, set your bid cap at or below the maximum CPC for your target ROI, and re-run the calculation each week as refund data matures. If you also work with creators, the influencer rate calculator on our home page prices sponsored posts using the same logic of comparing what a placement costs with what it can return.

Limits of this model

The calculator assumes constant averages. Real campaigns vary by placement, device, time of day and creative, and your blended numbers can hide profitable and unprofitable segments. It does not model taxes, currency conversion, ad-platform minimum spends, traffic quality penalties or cash-flow timing, and it treats refund and fee rates as proportional to revenue. Use the output as a decision aid for bid ceilings and go or no-go checks, not as a forecast or financial advice. Check each ad platform's and merchant's terms before you run traffic, including rules on disclosures and pre-sell pages.

Frequently asked questions