This Metric May Change The Way You Report Your Meta Traffic
I was reviewing a monthly performance report with a brand I work with. The paid social numbers looked good. ROAS was up, cost per acquisition was down, the agency had put together a clean slide deck with a lot of green arrows. The founder seemed happy.
I asked what had happened to overall revenue compared to the same month last year.
It had gone up by about £30,000.
The paid social spend that month was £18,000. The agency was claiming £270,000 in attributed revenue. A ROAS of 15. That is not an uncommon ROAS quoted by agencies. There is a big difference between £270k and £30k. That’s not a small discrepancy.
Nobody in the room had done that comparison before. Not because they were lazy or because the agency was obviously wrong. It was because the two numbers lived in different places, got reported in different meetings, and nobody had been asked to put them next to each other.
That gap, between what the channel says it drove and what the business actually took in, is the most consistently under explored number in ecommerce. And the way to close it is not a more sophisticated attribution model. It's a much simpler calculation that's been around for years and that most brands aren't using.
What Marketing Efficiency Ratio?
Marketing Efficiency Ratio. Total revenue divided by total marketing spend. That's it.
If you made £1,334 last month and spent £18 on marketing, your MER is 74. For every £1 you put into marketing, £74 came back.
I want to explain why that matters by using an analogy that I think is clearer than any of the technical explanations I've encountered.
Imagine you run a lemonade stand. At the end of the month you count all the money in the tin: £1,334. Then you count everything you spent on posters and flyers to tell people about the stand: £18. Divide one by the other and you get 74. That means for every £1 you spent on posters, £74 came into the tin.
That's MER. It doesn't try to work out which poster brought which customer. It doesn't care whether someone came because of a flyer, or because their mate told them, or because they walk past every day. It just asks: all the money in, all the poster money out, what's the ratio?
Why is that useful? Because the person who makes your Facebook posters is going to tell you their posters brought in loads of customers. Maybe they did, maybe they didn't, and you can't check by asking them. But you can check the tin. If you double the Facebook poster money and the tin gets fuller by more than that, the posters are working. If you double it and the tin stays the same, the posters aren't doing much, whatever the poster person says.
And the part it catches that channel attribution can't: if the poster person says "my posters brought in £270 this month" but the tin only went up by £30, someone is counting customers who were coming anyway. MER can't be fooled that way, because the tin is the tin.
Two things worth knowing before you start using it. It only works when you compare it to something. For example, he same month last year, or a period before a spend change. This is because a big sale month will always look great on its own. And it's a whole-business measure, so it can't tell you whether the Facebook posters or the Instagram posters did the work. It only tells you whether all the posters together earned their keep.
Why fashion brands need MER
Channel-level attribution has always had problems. Last-click models credit whoever was standing at the door when the customer walked in. Multi-touch models try to distribute credit across the journey but require assumptions about weighting that nobody agrees on. View-through attribution, where a channel claims credit for anyone who saw an ad and later bought, has been the source of more inflated agency numbers than I care to count.
But there are two things happening right now that make the problem considerably worse for drop-model brands.
The first is iOS and privacy changes. The signal that paid social platforms used to use for attribution has degraded significantly over the past few years. The numbers they report are increasingly modelled rather than measured. Some of the gap between what a platform claims it drove and what the business actually took in is an artefact of this modelling. The platform fills in the gaps where tracking broke down and landing on a number that is, consistently, higher than the true figure.
The second is agentic commerce. I've written about this across several pieces in this series. AI shopping agents are increasingly doing the consideration work before a customer reaches your site. That traffic has no UTM parameter. There's no click your paid channel can claim. No impression your social platform can count. The customer arrives, buys, and every attribution model in existence records it as direct or organic.
As agentic traffic grows, the share of revenue that paid channels can see shrinks. The denominator in your ROAS calculation gets smaller. Your ROAS looks better. The agency presents a slide deck with green arrows. And nobody checks the tin. It’s what’s in the tin that matters!
MER captures all of it. Agentic traffic that converted, word-of-mouth from the last drop, organic search, email. None of it is invisible to MER because all of it ends up in the tin.
How to implement MER
The calculation is straightforward. Total revenue for the period divided by total paid marketing spend for the period. Some brands include all marketing spend including agency fees, creative production, influencer. Some include only the direct media spend. I'd say be consistent above all else. Pick your definition, document it, and use it the same way every month. The absolute number matters less than the trend.
Then the harder part, which is the comparison.
MER needs a benchmark to be meaningful. The two most useful comparisons are the same period last year and a pre-intervention baseline. If you're trying to evaluate whether a new agency is performing better than the last one, you need a period before the agency change to compare against, at the same time of year if possible. Seasonality in fashion is significant enough that comparing January to October tells you almost nothing.
For drop-model brands the calculation gets slightly more interesting because revenue is lumpy. A drop month and an off-drop month are different beasts. They shouldn't be averaged together carelessly. I'd track MER separately across drop periods and non-drop periods, because the two states have different dynamics and blending them obscures both.
The practical check and the one I'd run monthly is this: take the total revenue increase from last year. Take the total additional marketing spend. Divide revenue increase by additional spend. If that number is above one, marketing is generating incremental return. If it's below one, you're spending more to make less, regardless of what the channel reports say.
That's a rough version of MER delta and it's the number that most quickly surfaces whether the poster person's claims are connecting to the tin.
The conversation with the agency
This is the part most brands avoid because it feels confrontational. It isn't. It's just asking for accountability at the right level.
Most agencies report at the channel level because that's what their tools measure and that's the framing that makes their numbers look best. That's not dishonest. It’s the frame they were given when they were hired. But it's incomplete, and part of your job as the brand is to set the right accountability frame.
The ask is simple. Alongside whatever channel metrics the agency reports, you want to track business-level MER over time. You're not asking them to stop reporting ROAS. You're adding a second measure that sits above it and that the whole team, brand, agency, founders, looks at together.
The useful question is:
Given this month's marketing spend increase of X, how much additional revenue did we see versus the same period last year?
If the agency can't engage with that question, that tells you something.
Most good agencies are comfortable with this framing because it lets them take credit for things that channel attribution undersells such as brand-building, upper-funnel work, anything that drives demand without leaving a traceable click. MER doesn't discriminate. If the spend is working anywhere in the system, the tin gets fuller.
Where this connects to everything else
The structured data and entity trust work I write about in this series is, among other things, a bet on channels that don't leave attribution trails.
An AI shopping agent recommending your products doesn't fire a pixel. A knowledge panel surfacing your brand in a search doesn't get credited to SEO. A Perplexity recommendation that converts doesn't show up as anything except direct revenue.
As the attribution gap grows, and I think it will grow significantly over the next few years as agentic commerce matures, MER becomes more important, not less. It's the only measure that doesn't have a blind spot for the traffic that came through channels nobody was tracking.
The tin is the tin. It counts everything. The poster person's dashboard doesn't.
That's not an argument for ignoring channel metrics. It's an argument for not letting channel metrics be the only thing in the room when you're deciding whether the marketing budget is earning its keep.
Check the tin first.
Summary Questions and Answers
Q: What is Marketing Efficiency Ratio (MER) and how do you calculate it?
MER is total revenue divided by total marketing spend over the same period. If your store made £1,334 in a month and spent £18 on marketing, your MER is 74. So, £74 came back for every £1 spent. Unlike ROAS, which measures returns from a single channel or campaign, MER gives a blended view of how efficiently your entire marketing investment is turning into revenue. It captures all revenue sources including paid, organic, word-of-mouth, email, and AI-driven traffic in a single number that no individual channel can inflate.
Q: How is MER different from ROAS?
ROAS measures the return from a specific ad campaign or platform. It divides the revenue that platform claims it drove by the spend on that platform. MER divides all revenue by all marketing spend, with no attribution required. The difference matters because platforms have a commercial incentive to claim as much revenue as possible. A paid social channel might claim £270,000 in attributed revenue in a month where total business revenue only grew by £30,000. MER catches this gap because it only cares what went into the tin, not what any individual channel claims to have put there.
Q: What are the limitations of MER?
Two worth knowing before you start. First, MER only means something when compared to a benchmark. For example, he same period last year, or a baseline before a spend change. Because a strong sales month will always produce a high MER regardless of whether marketing caused it. Second, MER is a whole-business metric. It tells you whether all your marketing together earned its keep, but it cannot tell you which channel or campaign did the work. Think of it as a smoke alarm, it tells you there is smoke but not how the fire was started. It’s an indicator.
Q: Why is MER more important for fashion brands now than it was five years ago?
Two reasons. iOS and privacy changes have degraded the tracking signal that paid social platforms use for attribution, meaning their reported numbers are increasingly modelled estimates rather than measured results and those estimates consistently run high. At the same time, AI shopping agents are driving an increasing share of consideration and purchase decisions without leaving any attribution trail. That traffic arrives with no UTM parameter, no claimable click, and no impression. It shows up as direct or organic revenue. MER captures all of it because it counts everything that goes into the tin, regardless of how it got there.