MER (marketing efficiency ratio)

The formula, and the two variants people mean by it

Blended MER is total revenue divided by total advertising spend. Every order counts, whoever placed it and whatever brought them; every dollar of media counts, whichever platform it went to. Both numbers come from systems that cannot flatter themselves — the order table and the ad invoices — which is the entire reason the metric is trusted.

Contribution MER replaces revenue with contribution margin: revenue less cost of goods, fulfilment, payment processing and returns, before media. It is the more honest version, because it breaks even at 1.0 by construction rather than at some margin-dependent number nobody has computed. A brand that reports contribution MER never has to argue about whether a given ratio is good; below 1.0 the advertising is not paying for itself.

A third cut, usually written aMER or acquisition MER, divides new-customer revenue by total advertising spend. It answers a different question: how much of the return is new business rather than repeat orders the advertising did not cause. Take an example month — $500,000 of revenue, $100,000 of media, a 45% contribution margin before media, and $300,000 of that revenue from first-time customers.

Why the metric exists: channel ROAS does not add up

Continue the same month. Meta reports a 4.0 ROAS on $60,000 of spend, which is $240,000 of claimed revenue. Google reports 6.5 on $40,000, which is $260,000. The email platform claims $150,000 of flow-attributed revenue. Those three numbers sum to $650,000 against $500,000 that actually reached the bank — $150,000 of revenue claimed by somebody that does not exist.

Nothing there is a bug. Each platform is answering "of the conversions I can see, which ones did I touch inside my own window", and a shopper who clicks a Meta ad, later clicks a branded Google ad and finally converts from an abandoned-cart email has genuinely been touched three times. The error is only in the addition, which nobody performs deliberately and every channel-level dashboard performs implicitly.

MER refuses to participate. It has one numerator and one denominator, both auditable, and it is the same number whether your attribution window is one day or twenty-eight. That immunity is bought by giving up all diagnostic power: a falling MER tells you the business got less efficient and offers no opinion about which channel did it.

What counts as a good MER

There is no universal number, and any source quoting one has quietly assumed a margin structure. Break-even blended MER is the inverse of your contribution margin before media: at a 45% margin it is 1 / 0.45, or 2.22. At a 25% margin it is 4.0. Two brands reporting the same 3.0 MER can be one comfortably profitable and one losing money on every order, and the difference is entirely in the cost structure.

Read the table below as a floor rather than a target, for one reason: blended MER includes revenue from returning customers who would have ordered without any advertising at all. A subscription brand with a large active base can clear its break-even MER on repeat revenue alone while every acquisition dollar loses money. This is why aMER is worth computing alongside — it strips the base out.

Read MER against a spend change, not in isolation

A single month's MER is a level. The decision in front of you is almost always about a change — spend more, spend less, keep it flat — and levels do not answer questions about changes. What does is the marginal MER: the extra revenue divided by the extra spend.

Take the example month at $100,000 of media and $500,000 of revenue, a blended MER of 5.0. The next month spend rises to $150,000 and revenue to $600,000, a blended MER of 4.0. The headline still looks strong. But the increment bought $100,000 of revenue for $50,000, a marginal MER of 2.0 — below the 2.22 break-even at a 45% margin, so the extra $50,000 lost roughly ten cents on the dollar while the reported ratio stayed comfortably above every benchmark in the category.

That is the whole practical use of MER: not the number, but the slope. Scale until marginal MER approaches break-even, then stop, and understand that the blended figure will keep looking healthy for a long time after the marginal one has stopped being.

The caveat is that a month-over-month comparison attributes everything to the spend change. Seasonality, a price rise, a promotion, a new product and an organic spike all land in the same delta. The fix is to make the spend change deliberate, hold everything else still, and give it long enough to clear the purchase cycle — at which point you are running a crude incrementality test, which is the correct destination.

Reconciling Google Ads ROAS with your blended number

The common version of this question is a brand looking at a 6.5x in the Google Ads interface and a blended MER that will not support it, trying to work out which one is lying. Neither is. They measure different things, and the gap decomposes into a small number of known causes.

• Reporting dates differ. Google credits conversion value back to the date of the ad interaction; your P&L records it on the order date. Over a single week that offset alone can be most of the discrepancy — compare over a period long enough for it to wash out.
• Branded search is usually the largest single cause. A shopper searching your brand name was already coming. Split branded and non-branded into separate campaigns and separate lines before comparing anything to blended revenue.
• View-through and engaged-view conversions are inside the reported figure for Demand Gen, PMax and YouTube. They are not clicks, and they are the least likely conversions to be incremental.
• Google deduplicates against Google. It has no knowledge of the Meta click three days earlier, so an order claimed at full value here may be claimed at full value there too.
• Performance Max reports one blended figure across Shopping, Search, YouTube and Display. A strong Shopping return and a weak Display one arrive as a single average with no way to separate them from inside the report.
• Reconcile at the total, not per channel: sum every platform's reported conversion value and compare it to actual revenue. The excess is the double count, and its size is the measure of how much you should trust channel ROAS as a steering input.