ROAS and MER get used interchangeably in most day-to-day conversations about ad performance, and that's a problem, because they answer genuinely different questions. Mixing them up doesn't just cause confusion, it leads directly to bad scaling decisions, since the two numbers can point in opposite directions in the exact same month.
The core difference
Platform ROAS is channel-specific and attribution-model-dependent. It's Meta, or Google, telling you what its own tracking believes it caused, using whatever attribution window and model that platform has configured. MER, Marketing Efficiency Ratio, is attribution-agnostic: total store revenue in a period, divided by total marketing spend in that same period, across every channel combined, Meta, Google, influencer, affiliate, everything.
MER doesn't care which platform gets credit for a sale. It simply asks: given everything we spent on marketing this month, how much total revenue came in? That single question sidesteps the entire attribution debate that platform ROAS is constantly wrestling with.
Why the two numbers diverge
Platform ROAS is vulnerable to over-attribution. Multiple platforms can each independently claim credit for the same sale under their own attribution windows and models, a customer who saw a Meta ad, then a Google Shopping ad, then converted, might get counted as a conversion by both platforms simultaneously. Sum the individual platform ROAS numbers together, and you can end up with a picture that overstates what actually happened in aggregate, since the same revenue got double or triple counted across channels.
MER can't be inflated this way, because it's calculated from actual total store revenue, a single, real number, not a sum of platform-reported conversions that may overlap. If your combined platform ROAS looks strong, but MER for the same period is noticeably weaker, that gap is a direct signal of attribution overlap or unaccounted spend, not a mysterious discrepancy.
When to trust which number
Use platform ROAS for day-to-day, in-platform decisions: pausing an underperforming ad set, comparing two creative variants against each other, adjusting bids within a campaign. It's directionally useful for these tactical calls even though it's imperfect in absolute terms, because you're comparing like-for-like within the same platform and attribution model.
Use MER for the big decisions: is total marketing spend actually profitable at the business level, should overall budget increase or decrease, is blended efficiency trending up or down month over month. MER is the number that should reconcile reasonably well with what your finance team sees on the P&L. Platform ROAS is the number your media buyers use to optimize execution day to day, it's a tactical tool, not a strategic one.
A worked example showing the divergence
Say a brand runs Meta, Google, and an influencer/affiliate program simultaneously in the same month. Meta reports 4x ROAS. Google reports 3.5x ROAS. A naive blended average across the two platforms might come out to roughly 3.7x, and if you eyeball that number without context, you'd assume the business is running an efficient, healthy marketing operation.
But if you calculate MER properly, total revenue for the month divided by total marketing spend across every channel including agency fees, tools, and influencer payouts, and it comes out to 2.2x, there's a real gap that needs explaining. That gap usually comes from one or more of: significant attribution overlap between Meta and Google claiming credit for the same customers, spend on channels (influencer, affiliate, PR) that isn't reflected in either platform's individual ROAS but is absolutely part of total marketing cost, or a chunk of platform-reported "conversions" that were never genuinely incremental in the first place.
Building a proper MER calculation
Total revenue: use your actual net store revenue for the period, ideally after returns, from Shopify or your order management system, not a platform-reported figure.
Total marketing spend: this needs to include everything, not just Meta and Google ad spend. Add agency or freelancer fees, marketing tool subscriptions (email platforms, SMS tools, analytics software), influencer and affiliate payouts, and any other cost that exists specifically to drive revenue. Leaving out agency fees or tool costs from the denominator is one of the most common ways MER gets calculated too optimistically.
Divide total revenue by total marketing spend for the same period. That's your MER. Track it monthly, and watch the trend line over time more than any single month's absolute number, since seasonality and one-off campaigns can swing a single month significantly.
Why blended platform ROAS alone misleads scaling decisions
If a team is deciding whether to increase overall marketing budget based solely on platform ROAS looking healthy, they're missing the possibility that a meaningful share of that "efficiency" is actually attribution overlap rather than real incremental revenue. Scaling budget further under that assumption often means paying more to capture the same customers multiple times across channels, rather than genuinely expanding the customer base, and MER is what eventually reveals this, usually after the fact, when overall efficiency doesn't improve the way platform numbers suggested it should.
What this looks like across brand types
A brand running a lean setup, one or two paid channels, minimal external spend beyond platform ad cost, will typically see platform ROAS and MER track relatively close to each other, since there's less room for attribution overlap and unaccounted spend to create a gap.
A brand running a fuller stack, paid social, paid search, affiliate, influencer, retention tools, and an external agency, will almost always see a meaningfully larger gap between blended platform ROAS and true MER, simply because there are more opportunities for double-attribution and more real costs sitting outside the platform dashboards entirely.
Common mistakes
Calculating MER using only ad spend, excluding agency fees and tool costs. This inflates MER artificially and defeats the entire purpose of using it as an attribution-agnostic sanity check.
Comparing MER across months without adjusting for major promotional periods. A big sale event will temporarily distort MER in ways that don't reflect the underlying trend, so compare like periods where possible.
Treating a strong platform ROAS as sufficient justification to increase overall budget, without checking MER first. This is the single most expensive mistake this framework is meant to prevent.
Not tracking MER consistently enough to see the trend. A one-time MER calculation is a snapshot. The real value comes from watching whether it's improving or declining over consecutive months.
FAQ
What's a healthy MER? There's no universal number, it depends heavily on category, margin structure, and how much of revenue comes from repeat customers versus new acquisition. What matters more than any absolute benchmark is the trend, is MER improving or declining over time as spend scales.
Should I report MER or platform ROAS to my board? MER, since it reflects total marketing efficiency at the business level and reconciles more sensibly with the P&L than any single platform's reported number.
Can MER and platform ROAS both be accurate at the same time despite looking different? Yes. This is the core point of this whole framework, they're both "accurate" for what they measure, they're just measuring different things, and the gap between them is informative, not a sign either one is wrong.
How do I explain a large gap between blended ROAS and MER to a client or stakeholder? Walk them through attribution overlap and unaccounted spend directly, using the worked example structure above, most people find this intuitive once they see the actual mechanics rather than being told to simply "trust MER instead."
The takeaway
Don't report platform ROAS to leadership as if it's the complete picture of marketing efficiency. Track MER as your top-line efficiency metric for strategic decisions, and use platform ROAS underneath it as a tactical, execution-level tool for your media buying team. Treating both correctly, for what each is actually good at, prevents exactly the kind of scaling decision that looks smart on a platform dashboard and turns out to be expensive at the P&L level.