Most teams discover marketing efficiency ratio at the same moment: the month the platform-reported numbers stop adding up.
Meta reports a 4× return. Google reports 5×. The affiliate programme reports its own conversions. Add them together, multiply by spend, and the total comfortably exceeds what actually landed in the bank. Nobody is lying. Each platform answered a slightly different question, and the same customer answered yes to several of them.
MER is the number that does not have this problem.
What is marketing efficiency ratio?
Marketing efficiency ratio is total revenue divided by total advertising spend over the same period, measured across the whole business.
That is the entire formula. Its power comes from what it refuses to do: it does not attempt to assign credit. It asks only whether the total money going out as advertising is producing enough total money coming in.
Why does per-platform ROAS overstate performance?
Because every platform measures with its own attribution window and its own definition of influence, and a buyer who saw three ads is counted by all three.
Consider a customer who sees a TikTok video, searches the brand a week later, clicks a Google ad, and buys. TikTok may claim the conversion under a view-through window. Google will certainly claim it under last click. Your accounting system records one order.
This is not a bug that better tooling fixes. Each platform is honestly reporting "someone exposed to our advertising converted." The overlap is real. The double-counting is structural.
The practical damage is in what it does to budget. Optimising each platform to its own reported return quietly shifts money toward whichever platform claims credit most aggressively — typically the one closest to the purchase, which is also the one most likely to have reached people who were already going to buy.
| Per-platform ROAS | MER | |
|---|---|---|
| Source of revenue figure | The platform | Your accounts |
| Affected by attribution windows | Yes | No |
| Double-counts overlapping channels | Yes | No |
| Can compare across platforms | Not reliably | Not applicable — it is one number |
| Useful for in-platform optimisation | Yes | No |
| Useful as a business target | No | Yes |
When should you use MER instead?
Four situations, and the first is the common one.
More than one channel is running. The moment two paid channels overlap, their reported returns stop summing to anything meaningful. MER is immune because it never disaggregates.
A large share of conversions are modelled. Where consent is declined or tracking is restricted, platforms estimate the conversions they could not observe. Those estimates are reasonable, and they are estimates. MER sidesteps the question entirely by never using platform-reported conversions.
Brand and performance run together. Brand activity that lifts conversion rates across every channel shows up in MER as improved total efficiency. In per-channel ROAS, it mostly shows up as the performance channels looking cleverer than they are.
You need a number for the board. MER is the one marketing metric that finance recognises without translation, because it is built from figures finance already owns.
How is MER actually calculated?
Three decisions have to be made before the number means anything, and they must stay fixed once made.
Which revenue counts?
Net revenue after returns and cancellations, in almost all cases. Gross revenue flatters categories with high return rates, sometimes dramatically. If you sell subscriptions, decide whether the period's revenue includes renewals from customers acquired earlier — it usually should, because the alternative understates the compounding effect of past spend.
Which spend counts?
Media cost at minimum. The more useful version includes agency fees, creative production and the tooling that only exists to support advertising. The narrower version is easier to calculate; the broader one is the one that tells you whether the function is worth its total cost.
Whichever you pick, write it down. The most common failure with MER is a definition that quietly drifts, so that an apparent improvement is actually a change in what got counted.
Which period?
Long enough that spend and its revenue land inside the same window. For impulse-purchase ecommerce that can be a week. For a considered purchase with a six-week research cycle, weekly MER is close to meaningless, and monthly or rolling-quarter figures are the honest unit.
What MER cannot tell you
It is a thermometer, not a diagnosis. It tells you the patient has a fever and nothing about why.
MER will not tell you which channel to cut. It will not tell you whether a declining ratio is caused by worse media buying, a weaker offer, seasonal demand, a competitor bidding harder, or a checkout change that shipped last Tuesday. It moves for all of those reasons and reports none of them.
This is why MER belongs at the top of the reporting hierarchy with diagnostic detail underneath, rather than replacing that detail. The ratio tells you whether to investigate. Channel data tells you where.
What is a healthy MER?
Whatever clears your contribution margin after every other cost. That is not a dodge — it is the only honest answer, and published benchmarks actively mislead here.
Work it backwards instead. If your contribution margin after cost of goods, fulfilment and payment processing is 40% of revenue, then every unit of revenue contributes 0.4 units. A MER of 3.0 means each unit of ad spend produces three units of revenue, contributing 1.2 units against 1.0 of spend — a small margin before any fixed cost. The same MER in a software business with 85% margin is a different world entirely.
The number to establish is your break-even MER: 1 ÷ contribution margin. Below it you are buying revenue at a loss. Above it, the gap is what funds everything else.
Using MER alongside CAC and payback
MER is one of three numbers that together describe acquisition health, and it is the weakest of the three on its own.
- MER answers: is the total budget producing enough total revenue?
- Blended CAC answers: what does one new customer cost, across everything?
- Payback period answers: how long until that customer has paid us back?
A business can have a healthy MER and a dangerous payback period, if revenue is arriving from existing customers while new-customer acquisition quietly gets more expensive. Reading MER alone hides that completely, because renewals and new acquisition are both just revenue in the numerator.
The pairing that catches most problems is MER plus new-customer CAC, tracked together. When MER holds steady and CAC rises, growth is being funded by the existing base — which is fine until it is not.