[← Measurement & attribution](/category/marketing-measurement)

# Marketing efficiency ratio (MER) vs ROAS: when each one lies

[David Pombar](/about-me) · October 26, 2026 · 9 min read

The marketing efficiency ratio (MER) is total revenue divided by total marketing spend, with no attribution at all. It works as your main metric when three things hold: revenue arrives soon after the spend, channel attribution is unreliable, and the decision is how much to spend in total, not where. If any of the three fails, MER becomes a sanity check and another metric takes the lead.

ROAS stays the tool for deciding where the money goes within each channel. You need both, because each one hides the error the other one exposes.

**How I compared them.** Five criteria: the question each metric answers, the data it needs, what it cannot see, how a tracking failure affects it, and how fast it reacts. Both are ratios that each team defines, not standards, so I state the definitions I use below. MER has no standard benchmark: a "good MER" depends on your margin, so the post shows how to calculate your own.

## MER and ROAS side by side

|     | MER | ROAS |
| --- | --- | --- |
| Formula | Total revenue / total marketing spend | Revenue attributed to the ads / cost of the ads |
| Question it answers | Is marketing, as a whole, paying for itself? | Which campaign or channel is pulling its weight? |
| Attribution needed | None | Yes: a model and a look-back window |
| Revenue source | The order system or the general ledger | Analytics, the ad platform or the CRM |
| Scope | The whole business, or one market | A campaign, ad set, channel or audience |
| Breaks when tracking breaks | No  | Yes: a missing or duplicated event moves it |
| Credits organic and brand demand | Yes, to all spend | No, only what the model attributes |
| Useful for | Total budget, scaling decisions, board reporting | Bids, creative, channel mix within a budget |

## The formula, and your break-even MER

The base calculation:

```
MER = Total revenue / Total marketing spend
```

For example, a store with $1,200,000 of revenue in a quarter and $300,000 of marketing spend has a MER of 4. Every dollar of marketing came with four dollars of revenue, from every source, paid or not.

Both sides need a written definition before the number means anything:

-   **Revenue:** net of refunds and cancellations, from the system that records orders. Decide whether it includes tax and shipping, and keep it that way.
-   **Spend:** media at minimum. Decide whether agency fees, creative production and tools are in. A "media MER" and a "fully loaded MER" are both fine; mixing them across months is not.

A MER of 4 means very different things at different margins. With a 30% gross margin it barely covers the media; with a 70% margin it leaves plenty. So the useful threshold is your own break-even, expressed in contribution margin:

```
Break-even MER = 1 / contribution margin before marketing
```

Contribution margin before marketing is revenue minus product cost, shipping, payment fees and returns, as a share of revenue. For example, with a 40% contribution margin, break-even MER is 2.5. Below 2.5, marketing spends more than the orders contribute. The [ROAS and break-even calculator](/how-to-calculate-return-on-ad-spend#roas-and-break-even-roas-calculator) does the same calculation if you enter total revenue, total spend and that margin.

## The differences that decide which one to trust

### MER cannot double count, platform ROAS usually does

Each ad platform attributes conversions with its own model and window. Since Apple's [App Tracking Transparency](https://developer.apple.com/documentation/apptrackingtransparency) and with consent and browser limits on tracking, platforms model more of the conversions they claim. Add up the ROAS revenue from Meta, Google and TikTok and you can end up with more sales than you had.

MER cannot inflate that way: it is money that came in divided by money that went out.

### MER credits everything to marketing, including what marketing did not cause

Revenue from returning customers, word of mouth, organic search and seasonality all lands in the numerator. A brand with strong organic demand can show a high MER while its paid media adds little.

That is why MER cannot tell you whether a specific channel works. For that you need ROAS, and for the causal answer you need an [incrementality test](/incrementality-testing).

### ROAS breaks when tracking breaks; MER does not

ROAS depends on events firing correctly: one purchase per order, the right value, the right currency, the right UTM. A broken tag after a site release moves ROAS overnight.

MER only needs the order total and the spend invoices. It does not depend on a pixel, consent or a click ID. So if ROAS drops sharply and MER does not, suspect the tracking before you suspect the campaigns.

### ROAS reacts in hours, MER in weeks

You can read ROAS for a campaign the day after it launches. MER moves slowly, because one campaign is a small share of total spend and revenue. Use ROAS for daily optimization and MER for monthly or quarterly budget decisions.

## Separate new customers: aMER

Total MER mixes acquisition with repeat purchases. **New-customer MER** divides only first-order revenue by marketing spend. I call it aMER (acquisition MER), and in conversation "new-customer MER", so nobody has to guess the acronym. You will also see nMER and NC-MER for the same idea.

Do not confuse it with new-customer ROAS by channel. Northbeam's [ROAS (New)](https://docs.northbeam.io/docs/northbeam-metrics-101), for example, divides attributed new-customer revenue by spend, so it depends on the attribution model. aMER is blended and does not. Triple Whale's [NC ROAS](https://triplewhale.readme.io/docs/nc-roas) uses blended ad spend, which brings it close to aMER: check the definition, not the label.

The definition matters more than the name, because it varies a lot between teams. Write down four decisions:

1.  **Numerator:** only the revenue of each new customer's first order, without repeat orders in the same period.
2.  **Gross or net:** whether returns, discounts and taxes are subtracted.
3.  **Denominator:** media spend only, or also agency, creators, tools and samples. Media only is the most common, but then it is not comparable with a fully loaded CAC.
4.  **Who counts as new:** first order ever, or new in the last 12 or 24 months. Reactivated customers are a gray zone.

If two people on the same team say "aMER" and answer any of these differently, they will compare numbers that are not comparable. I put the formula next to the metric on the dashboard rather than worry about the label.

## Read MER and ROAS together

The two metrics moving in different directions tell you more than either one alone. A reading guide for a monthly review:

| MER | Platform ROAS | Likely explanation | What to check first |
| --- | --- | --- | --- |
| Up  | Up  | Paid media and demand are both improving | Whether a promotion or seasonality explains it |
| Down | Down | Paid media is getting less efficient | Saturation: did spend grow faster than revenue? |
| Flat | Down sharply | A tracking problem, not a performance problem | Purchase event, consent, UTMs after the last release |
| Down | Up or flat | Platforms are claiming credit for sales that would have happened anyway | Retargeting and branded search share of spend |
| Up  | Down | Organic or brand demand is carrying the business | Whether paid spend can be cut without losing revenue |

The fourth row is the expensive one. Platform ROAS looks healthy, so budget keeps going up, while total revenue per dollar of marketing falls. MER shows it because it does not depend on what the platforms claim.

## Switching your reporting to MER

Adding MER to a dashboard is cheap: two numbers you already have, and a division. The hard part is agreeing the definitions with finance, so the revenue and spend in MER match the figures the CFO sees.

What usually changes for the team:

1.  **The weekly review starts with MER** against break-even, then drills into channel ROAS.
2.  **Channel targets become ranges, not single numbers,** because channel ROAS is only a model's estimate.
3.  **Budget increases need two signals:** ROAS on the channel, and MER holding up as spend grows.

It is still new. Some of the agencies that partner with Trackingplan, where I work, have started offering MER-based reporting, and adoption among their clients is uneven.

## My verdict: it depends on how fast revenue follows spend

### Multi-channel e-commerce: yes, but not alone

This is where MER fits best. Platforms over-claim, and the purchase cycle is usually short, so spend and revenue in the same period are reasonably connected.

I would use MER as the north star for the total budget, with two adjustments: track aMER next to it, and express it in contribution margin, not revenue. Within each channel I would keep optimizing with platform ROAS, calibrated with incrementality tests (geo holdouts, controlled pauses) or [marketing mix modeling](/marketing-mix-modeling-complete-guide-analysts).

### B2B with a long cycle: not as the main metric

If an opportunity takes 6, 12 or 18 months to close, this quarter's revenue answers to last year's spend. Period MER then measures a lag: you can cut budget, see a spectacular MER for months, and run out of pipeline afterwards.

Here I would lead with qualified pipeline generated per dollar (or cost per real opportunity, not per MQL), CAC payback and LTV:CAC. I would only look at MER by cohort and with the lag: a year's spend against the revenue it ended up producing, as an annual review rather than a management tool.

### Apps: only as a global check

Two problems rule it out. Revenue arrives over months, through subscriptions and in-app purchases. And a large share of installs is organic (store search, featuring), which MER mixes with paid.

What usually works is cohort ROAS at D7, D30 and D90 with a projected LTV curve, plus cost per payer or per trial instead of CPI. Company-wide MER is there to check that the projections have not drifted from reality.

### The weakness that worries me in each metric

-   **MER:** looking healthy while acquisition deteriorates. A large base of repeat customers can hold the ratio up for months while the cost of new customers climbs. By the time repeat revenue stops compensating, the problem is old. That is why I insist on aMER.
-   **Platform ROAS:** over-attribution, above all in retargeting and branded search. It rewards channels that harvest demand over those that create it, and optimizing on it alone shifts budget toward sales that would have happened anyway.
-   **CAC:** how elastic its definition is. Does it include salaries, tools, agency, discounts? Blended or paid only? I have seen the same business with very different CACs depending on who calculated it.
-   **Projected LTV**, key in apps: it is a prediction that ends up treated as data. It fails exactly when you need it: a new channel with different users, a price change, a new retention strategy. An optimistic LTV justifies a CAC that never pays back.
-   **Pipeline metrics** in B2B: they can be gamed. If marketing is measured on opportunities created, opportunities appear that nobody closes. Measure them by close rate, even if that brings the lag back.
-   **Incrementality tests:** the closest thing to causal truth, but expensive, noisy on small budgets, and a snapshot. A result from six months ago, with other creative and other seasonality, is worth less today.

If I had to pick one, it is MER's. The other metrics have obvious flaws, and people distrust them. MER has a reputation as the "honest" metric, and that is exactly why nobody questions it when it starts hiding an acquisition problem.

## What to check this week

-   **Write down your contribution margin before marketing** and turn it into a break-even MER. Put it on the same dashboard as total MER and aMER.
-   **Add up the revenue your ad platforms claim** for last month and compare it with the revenue in your order system. The gap is how much attribution is double counting.
-   **Write the four aMER decisions** (numerator, gross or net, denominator, who counts as new) next to the metric, before anyone compares numbers.

*Disclosure: I am the Chief Revenue Officer at Trackingplan.*

![](/images/about-me/davidpombar_profile.jpg)

Written by [David Pombar](/about-me). Chief Revenue Officer and Swiss Army knife at Trackingplan. Lives in A Coruña, Galicia.