Ad platforms report the revenue they can connect to their ads. Your store reports the revenue it actually collected. The two numbers measure different things, so they won't match, and the gap can widen as you add channels, retargeting, and email.
That gap is where budget decisions go wrong. A campaign gets more budget because its ROAS looks strong, and total profit doesn't move. This guide defines both metrics, shows when each deserves your trust, and gives you a routine for reading them together.
Define MER, ROAS, and margin precisely
Agree on definitions before anyone compares numbers. A quiet change to what counts as revenue or spend can move either metric more than a real change in performance.
MER = Total revenue ÷ Total marketing spend
Uses revenue from your store or finance system. It ignores attribution entirely.
ROAS = Attributed revenue ÷ Ad spend
Measured by platform, campaign, or ad. The revenue is whatever that platform's attribution model credits to its ads.
aMER = Revenue from first-time customers ÷ Total marketing spend
Also called acquisition MER. Some teams divide by acquisition spend only. Either works if you stay consistent.
Contribution margin = Net revenue − COGS − Shipping and fulfillment − Payment fees − Returns cost − Marketing spend
COGS is the cost of goods sold. The result is what's left to cover fixed costs and profit. Calculate it before marketing too, because break-even targets start there.
Two choices shape every number above. For revenue, use net revenue after discounts and refunds, and leave out taxes. For spend, decide whether to include only media or also agency fees, creative production, affiliate commissions, and tools. There's no universal right answer. The wrong answer is changing the definition between reports.
Why platform ROAS over-counts
Each platform measures its own ads in isolation, using its own attribution rules. None of them sees the full path to purchase. Add their reported revenue together, and the total can exceed what your store took in.
- Attribution overlap. A shopper clicks a Meta ad, later clicks a Google Ads result, then clicks through a Klaviyo email and buys. All three can report the order as theirs.
- View-through credit. Some attribution settings credit a purchase to an ad that was shown but never clicked, even if the ad had no influence.
- Modeled conversions. When consent choices or browser privacy features block tracking, platforms estimate the conversions they can't observe. Estimates are useful, but they aren't orders.
- Existing demand. Brand search, retargeting, and ads shown to past customers can claim sales that would have happened anyway.
- Gross values. The purchase value sent at checkout doesn't reflect later refunds or returns unless you adjust it.
- Shopify net revenue: $50,000
- Revenue reported by Meta: $31,000
- Revenue reported by Google Ads: $27,000
- Revenue attributed by Klaviyo: $12,000
- Sum of claims: $70,000, which is $20,000 more than the store collected
Invented figures. Some store revenue also came from unpaid channels that no tool claimed, so the real overlap is larger than the difference.
When to trust ROAS and when to trust MER
ROAS is for decisions inside a channel
ROAS works when you compare things measured the same way. Two campaigns in one platform share its attribution model, windows, and blind spots, so the bias is similar on both sides of the comparison.
- Moving budget between prospecting campaigns in the same platform
- Comparing creative, offers, and audiences in a structured test
- Setting targets for automated bidding, which optimizes to the platform's own numbers
- Finding products or campaigns that have stopped converting
Compare like with like. Brand search and retargeting reach people close to buying, so their ROAS can look stronger even when they add little new revenue. Judge them as separate groups with separate targets. Our paid media page shows how this shapes campaign structure.
MER is for decisions about the total budget
MER doesn't care which platform claims a sale. It asks a simpler question: for everything spent on marketing, how much revenue came in? That makes it the right number for decisions no single platform can see.
- Setting total monthly or quarterly marketing spend
- Checking whether growth stays profitable as spend rises
- Planning budgets with finance, using numbers that match the bank account
- Testing platform claims when they drift away from store revenue
MER has limits too. It can't say which channel caused a change. It includes revenue from returning customers, organic search, and word of mouth. And spend today can drive orders weeks later, so one week can mislead. For channel-level answers, analytics and attribution methods such as holdout tests, where a comparable group sees no ads, help fill the gap.
Set targets from contribution margin
A target ROAS or MER only means something next to your margins. A ratio that keeps one store profitable can lose money for another with higher product costs or more returns. Build your targets from your own numbers.
Break-even MER = Net revenue ÷ Contribution before marketing
Contribution before marketing is net revenue minus COGS, shipping, payment fees, and returns cost. At break-even, marketing uses up all of it.
Target MER = Net revenue ÷ (Contribution before marketing − Required contribution)
Required contribution is what the business needs after marketing to cover fixed costs and profit. Recalculate it at each planned revenue level.
- Planned net revenue for the month: $400,000
- COGS, shipping, payment fees, and returns cost: $220,000
- Contribution before marketing: $180,000
- Break-even MER: $400,000 ÷ $180,000 = 2.22
- Required contribution for fixed costs and profit: $60,000
- Maximum marketing spend: $180,000 − $60,000 = $120,000
- Target MER: $400,000 ÷ $120,000 = 3.33
Invented figures that show the arithmetic only. Use your own margin data.
Break-even ROAS = Revenue ÷ Contribution before ad spend
Calculate both figures for the products a campaign actually sells, not for the whole store.
Break-even ROAS assumes every attributed sale is real and caused by the ad. Platform numbers don't meet that standard, so a target set at break-even can lose money in practice. Set in-platform targets above break-even, then adjust them as MER and holdout results show how much claimed revenue holds up.
Separate new and returning customer revenue
Returning customers can buy with little or no paid push, and their orders raise MER whether or not acquisition is working. A store can hold a steady blended MER while each new customer costs more, and the problem only surfaces when the returning base stops growing.
New-customer CAC = Total marketing spend ÷ New customers acquired
Use the same spend definition as your MER. Compare the result with the contribution a new customer brings over time.
- Quarter A: $300,000 revenue on $100,000 spend, so MER is 3.0
- Quarter A new-customer revenue: $150,000, so new-customer MER is 1.5
- Quarter B: $330,000 revenue on $110,000 spend, so MER is still 3.0
- Quarter B new-customer revenue: $132,000, so new-customer MER falls to 1.2
Invented figures. Blended MER held because returning customers covered a decline in acquisition efficiency.
Split orders by first-time and returning customers using Shopify order data, and track new-customer MER and new-customer CAC next to blended MER. Google Ads offers a new-customer acquisition setting for some campaign types. It uses customer lists and its own detection, so check its counts against your store data.
Read both numbers together
Neither number is enough alone. Use this table when they disagree, because that's when the decision matters most.
| Situation | What to trust | Decision |
|---|---|---|
| Platform ROAS rising, MER flat or falling | MER | Hold the budget. Check overlap, brand and retargeting share, and returning-customer claims before scaling. |
| Platform ROAS falling, MER steady | MER, after a tracking check | Confirm conversion tracking still works. Don't cut spend on platform ROAS alone. |
| Both rising | Both, with a new-customer check | Scale in steps. Watch new-customer MER and contribution margin each week. |
| Both falling | MER for the size of the problem, ROAS for the cause | Rule out site, stock, pricing, and seasonality first. Then look for weak campaigns. |
| Blended MER steady, new-customer MER falling | New-customer MER | Don't add budget on blended results. Fix acquisition offers, targeting, or creative first. |
| Choosing between campaigns in one platform | ROAS, like for like | Move budget toward stronger campaigns. Judge brand and retargeting separately. |
| Setting next quarter's total budget | MER and contribution margin | Set spend at the level where the last dollars still clear target MER. |
Pitfalls that distort both numbers
- Promotions. Discounts can lift revenue and MER while cutting margin per order. Judge promotional periods on contribution margin, not on ratios.
- Seasonality. Peak periods can raise conversion rates across every channel. Compare with the same period last year, or with a baseline you built, not with the prior week.
- Returns. Products with frequent returns can show strong ROAS on gross order value and lose money after refunds. Use net revenue, and check return rates before scaling a product.
- Brand search. Brand campaigns catch people already looking for you, so their ROAS says little about what they add. Test with a controlled holdout, and account for competitors bidding on your name.
- Reporting dates. Platforms can report conversions against the date of the ad interaction, while your store uses the order date. Recent days keep changing, so don't judge them yet.
- Definition changes. Adding a cost to spend or a revenue source to revenue changes MER overnight. Note every change in the report.
How to use this every week
A fixed routine keeps both numbers honest. Run it on the same day each week, with the same date ranges and definitions. Read weekly numbers as a trend, and make total budget decisions on longer windows.
- Pull net revenue, total marketing spend, and new-customer revenue from store and finance data.
- Calculate MER, new-customer MER, and new-customer CAC, and compare each with its target.
- Pull platform-reported revenue and ROAS, and compare the platform total with store revenue.
- Where MER and ROAS disagree, use the decision table to choose which number to trust.
- Adjust spend within channels using ROAS, and adjust the total using MER.
- Log what changed and why, so next week's review can check the result.
Build your own baseline before setting alert thresholds. A run of comparable weeks shows how much MER and the platform gap move on their own, so you can tell noise from a real shift. For how these metrics feed wider growth planning, see e-commerce marketing.
