MER vs ROAS: Which Efficiency Metric Runs Your Budget?
Same account, same money, two completely different stories. Meta's dashboard told me 4.0 ROAS. My blended MER, run across the whole business that quarter, came back at 1.8. Nothing was broken and nobody was lying. The two numbers were just measuring from different altitudes.
Here's the short version you can steal. In the MER vs ROAS debate, MER (Marketing Efficiency Ratio) measures total revenue divided by total marketing spend — a top-down, blended read. ROAS measures return at the campaign or channel level. They aren't rivals fighting over the same throne. You report MER to the CFO and channel ROAS to the media buyer, and if you get those audiences backwards, you'll spend your Monday defending a number to the wrong room.
The 30-Second Answer
If you asked me to sketch this on a whiteboard, it looks like this:
| MER | ROAS | Blended ROAS | |
|---|---|---|---|
| Definition | Total revenue ÷ total marketing spend | Revenue ÷ spend for one campaign/channel | Same math as MER, different name |
| Formula | All revenue / all marketing cost | Channel revenue / channel spend | Total revenue / total spend |
| Scope | Whole business, top-down | Single channel, bottom-up | Whole business, top-down |
| What it answers | "Is marketing efficient overall?" | "Where does the next dollar go?" | Same as MER |
| Who reads it | CFO, board | Media buyer | CFO, growth lead |
That last column isn't a typo. Blended ROAS is MER wearing a different lanyard. Northbeam notes the same thing in its writeup — "this broad scope is why MER is sometimes referred to as blended ROAS or eROAS." The blended version captures halo and cross-channel influence that a single platform can never see. So when someone in a meeting says "blended ROAS," mentally file it next to MER and move on.
Napkin Math: Why Platform ROAS Says 4.0 and MER Says 1.8
Let me show you the gap with deliberately round numbers, because that's the fastest way to see where the truth leaks out.
Meta reports $100k in spend and claims $400k in revenue. Clean 4.0 ROAS. You screenshot it, you feel great, you almost put it in a QBR. (I have done exactly this. More on that in a second.)
Now zoom out to the whole business for the same period:
| Line item | Spend | Revenue |
|---|---|---|
| Meta (as reported) | $100k | $400k (claimed) |
| Brand/search spend Meta ignores | $80k | — |
| Agency retainer + creative | $30k | — |
| Other channels | $40k | — |
| Total marketing spend | $250k | — |
| Actual booked revenue (all sources) | — | $450k |
| Blended MER | 1.8 |
The 4.0 didn't lie about its own math. It counted claimed conversions inside its own walls and ignored every dollar of spend that didn't run through its ad account. It also can't tell you how many of those $400k were people who'd have bought anyway.
The gap between 4.0 and 1.8 comes from three things: spend the platform never saw, revenue it double-counted or claimed credit for, and halo it can't measure either direction. None of those are bugs. They're just outside the platform's field of view.
Now the confession I owe you. Years ago I stood up in a quarterly review and bragged about a campaign posting a monster ROAS. What I hadn't checked was that most of that "conversion volume" was branded search — people typing our name into Google after seeing a billboard we'd paid for separately. I was taking credit for demand another budget line had already bought. The blended number would have caught it in about four seconds. I did not run the blended number. Lesson filed permanently.
Why Platform ROAS Stopped Being the Scoreboard
There's a specific before-and-after here, and it's Apple's iOS 14.5 App Tracking Transparency rollout. Before that, platform-reported conversions were rough but usable as a scoreboard. After it, the reporting cratered. Adligator's writeup on Facebook attribution in 2026 puts it plainly: since ATT, "Facebook has consistently underreported conversions — typically by 15-30%, sometimes more."
So the dashboard number went from "approximately what happened" to "a fraction of what happened, biased in ways that shift with your traffic mix." That's a different kind of number. It's still excellent for one job, telling the algorithm which of your ads to show more often, and increasingly useless for a second, which is reporting the truth upward.
The Google side has its own mechanics worth knowing. As Search Engine Land explains, the column you actually see in Google Ads isn't labeled ROAS at all — it's "Conv. value / Cost," calculated by dividing conversion value by cost, and Target ROAS gets entered as a percentage (300% for a 3.0 target). The catch sits underneath it. Google's own documentation says its bidding "predicts future conversions and associated values using your reported conversion values, which you report through conversion tracking." Read that twice. The platform's ROAS is only ever as good as the conversion values you feed it. Garbage in, confidently-optimized garbage out.
Jordan Glickman, writing about post-iOS 14 attribution, lands where most practitioners have: "platform ROAS became a directional input for within-platform optimization decisions, not a primary performance signal." That's the whole shift in one sentence. Platform ROAS didn't get worse at its actual job. It got worse at a job we'd wrongly assigned it — being the scoreboard.
What MER Gets Right, and Where It Goes Blind
MER's superpower is that it refuses to play the attribution game. It doesn't care whether Meta or Google or your email list "deserves" a given sale. It sums all your marketing spend, sums all your revenue, divides, and hands you one honest ratio. No last-click argument. No modeled-conversion asterisk. The attribution wars simply don't apply, because MER never picks a side.
That same refusal is its blind spot. HubSpot puts it directly: MER "cannot diagnose which individual campaigns or channels are driving performance." It'll tell you efficiency slipped from 1.8 to 1.5 this quarter. It will not tell you that TikTok tanked while search held steady. MER flags the fire. It can't point at the room. HubSpot's own framing is that the two metrics are "complementary rather than interchangeable" — MER only becomes actionable when you pair it with channel ROAS, CAC, and LTV underneath it.
One grounding note on how the number gets expressed, because people trip on this constantly. Triple Whale's first-party data across its customer base found that in 2025 "the median MER for our Triple Whale customers was 41%." That's MER expressed as a spend-to-revenue percentage — spend as a share of revenue, so lower is better. Flip it the other way and you'd talk about revenue per dollar of spend, where higher is better. Same reality, inverted fraction. Before you benchmark yourself against anyone, confirm which direction they're pointing the ratio, or you'll congratulate yourself for a number that's actually bad.
Which Metric Goes to Which Room
Here's the framework I actually use to decide what number goes in front of which face. The mistake isn't picking the "wrong" metric — both are correct. The mistake is delivering the right metric to the wrong audience.
| Audience | Lead metric | Why it fits | Decision it drives |
|---|---|---|---|
| CFO / board | MER (blended) | Attribution-proof, ties all spend to all revenue | "Is marketing efficient overall — keep or cut budget?" |
| Head of growth | MER + channel ROAS deltas | Reads mix health, spots the drift | Reallocation across channels |
| Media buyer | Channel / campaign ROAS | Within-platform optimization signal | Where the next dollar bids |
Here's the take I'd fight for. Reporting raw channel ROAS up to a CFO is walking into an attribution fight you will lose. The moment you say "Meta returned 4.0," a sharp finance person asks how much of that overlaps with your Google spend, and now you're defending modeled conversions instead of talking about whether the business is healthy. MER closes that door. One number, whole-business, nothing to argue about.
The reverse fails just as hard. Hand a media buyer your blended 1.8 and you've given them nothing to act on. They can't bid a blended ratio. They need the channel-level signal, warts and under-reporting and all, because within a single platform that number still steers the algorithm correctly. Right altitude, right room. If you want the layer that sits beneath both — the CAC and LTV mechanics that make either ratio mean something — we built a unit economics dashboard walkthrough for exactly that.
Watching Both at Once, Live
The practical annoyance is that MER and channel ROAS usually live in different places. Blended math wants your billing data, your organic revenue, your retainer costs. Channel ROAS lives in ad platforms. Stitching them together by hand every Monday is how weekly reporting quietly eats a person's morning.
You want both on one view, ideally something you can interrogate in plain language instead of rebuilding a pivot table each week. Tools like GA4, Triple Whale, and Kixo let you ask for a blended figure conversationally rather than hand-assembling it from four exports. Whatever you use, the goal is the same: the health metric and the steering metric visible side by side, so the moment MER drifts you can drop a level and find the channel that moved it. If you're still deciding how to structure that measurement stack, that's the question to answer before you pick a dashboard.
The Mature Answer: Both, for Different Rooms
So which metric runs your budget? Neither, alone. MER is your health metric — the resting heart rate of the whole marketing function. Channel ROAS is your steering wheel, useless for diagnosing overall fitness, essential for deciding which direction to turn next.
Come back to the number I opened with. The 4.0 and the 1.8 weren't in conflict, and neither was wrong. The distance between them was the insight. It measured exactly how much brand spend, double-counted conversion, and unmeasured halo the platform view was quietly ignoring. Read the gap. Don't try to eliminate it.
And on the attribution wars that have burned more analyst hours than any of us will admit — I'm not declaring a winner. Last-click isn't dead, modeled conversions aren't gospel, and blended math has its own blind spots. What MER buys you is the option to stop fighting at the executive level entirely, and to keep the skirmish contained to the media buyer's screen where it belongs. Call it a truce, not a victory.
FAQ
Is blended ROAS the same as MER? Effectively yes. Blended ROAS and eROAS are alternate names for the same top-down calculation — total revenue divided by total marketing spend. If someone uses them interchangeably, they're not wrong.
Why is my platform ROAS so much higher than my MER? Because the platform counts only spend inside its own account and claims credit for conversions it may have merely influenced. MER adds the spend it ignores (brand, retainers, other channels) and uses your actual booked revenue. Post-iOS 14.5 under-reporting, cited by Adligator at 15–30%, complicates it further in both directions.
Should I stop trusting platform ROAS entirely? No. Just re-cast its job. Treat it as a within-platform optimization signal for bidding, not as your scoreboard. It's still good at telling the algorithm which ads to favor. It's no longer reliable as a source of business truth.
What's a good MER number? It depends how you express the ratio. Triple Whale's 2025 first-party data put the median at 41% when measured as spend-to-revenue, where lower is better. Always confirm the direction of the fraction before comparing yourself to anyone else's figure.