Most ad accounts we audit are reporting on a number their CFO doesn't trust. The headline says ROAS 4.2. The bank says break-even. Somewhere between the dashboard and the deposit, the money went missing.
It didn't. The metric was wrong.
ROAS vs MER - and then ROAS vs MER vs CAC - isn't a semantics debate. It's the difference between an account that looks healthy and one that actually is. Three numbers, three jobs, three audiences inside the business that need them. Get the stack wrong and you'll scale a profitable-looking account straight into a cash-flow hole.
The Three-Number Stack
Each metric answers a different question, for a different person, at a different altitude. They are not interchangeable, and that is the whole point of the framework.
| Metric | Question it answers | Who asks it | Reports to |
|---|---|---|---|
| ROAS | "Did this ad pay for itself?" | The media buyer | The dashboard |
| MER | "Did marketing pay for itself?" | The CMO | The CFO |
| CAC | "What did one new customer cost?" | The CEO | The board |
The trap is treating ROAS as the answer to all three questions. It can't carry that weight. It was never designed to.
ROAS vs MER: where platform ROAS keeps lying
Platform ROAS is what Meta or Google reports inside their own dashboard. Each platform takes credit for any conversion it touched, however briefly, so two ads from two channels can both claim the same sale. Email can claim it too. Stack the platform ROAS numbers side by side and they sum to a fantasy.
MER doesn't do that. MER is total revenue divided by total marketing spend, across every channel, no attribution gymnastics. If the business pulled $500k revenue last month against $100k of marketing spend, MER is 5. End of conversation.
That's the difference ROAS vs MER actually captures, and why marketing efficiency ratio explained correctly is the metric a finance team can use. It ignores who got the credit and tracks whether the spend, in aggregate, moved the top line. The CFO doesn't need to know which ad won. They need to know if the month worked.
If your reporting stack only shows platform ROAS, you are paying to be lied to. Politely. By an algorithm with a quota.
Two things make MER bite once you start using it:
- It rises and falls with brand spend, not just direct response. Big YouTube quarter? MER tells you whether the lift was real or imagined. Channel-level ROAS won't.
- It exposes channel cannibalisation. When platform ROAS in three channels all climb and MER doesn't move, the channels are taking credit for the same revenue. Now you know.
Most boards we sit in get a MER target before they get a ROAS target. That order isn't an accident.
Blended ROAS vs platform ROAS: the gap that punishes scale
Blended ROAS is the same idea as MER in slightly different clothes: total tracked revenue divided by total ad spend. It's what sits between the lie of platform ROAS and the honesty of MER.
The blended ROAS vs platform ROAS gap punishes scaling brands first. At small spend the noise is small; at $150k a month, the noise is a six-figure mirage. We've watched accounts where blended ROAS vs platform ROAS sat 2x apart - the in-platform buyer thought they were running 3.1x, the accountant said 1.6x. Both right, looking at different numbers, telling different stories.
The fix isn't exotic. It's having both numbers in the same row of the same report, every week, with the gap as its own column. The gap is the story.
CAC: how to calculate it without flattering yourself
CAC is the third leg, and the one founders most often calculate too kindly.
How to calculate CAC, honestly: total marketing and sales spend over a period, divided by the number of new customers acquired in that period. Not orders. Not sessions. If a returning buyer ends up in the denominator, you've just diluted CAC to a number no one can act on.
Then pair CAC with two contextual numbers it lives or dies by:
- LTV : CAC ratio. Healthy DTC sits around 3:1 over 12 months. Below 2:1 and you're buying revenue, not growing a business.
- Payback period. How many months until that CAC returns? Longer than your runway and the growth is borrowed time.
A CAC of $80 with a 9-month payback is a different business than a CAC of $80 with a 2-month payback. The headline number is the same. The cash-flow shape is opposite. Knowing how to calculate CAC is the easy part; reading it with payback and LTV next to it is what makes the number useful.
How to read the stack
Read the three numbers in this order. The lever you pull is different at each level.
| Level | Read | If it breaks, the lever is... |
|---|---|---|
| 1. MER | Direction of the whole engine | Channel mix and budget |
| 2. CAC | Health of new-customer economics | Creative, offer, audience match |
| 3. ROAS | Diagnostic per channel and campaign | Bid, creative, targeting |
Notice ROAS comes last. It's the most granular and the loudest, which is why it dominates Monday meetings - and why those meetings often optimise the wrong thing. When MER is healthy and CAC is in range, a soft platform ROAS isn't an emergency; it's a measurement artefact. When MER is sagging while platform ROAS climbs, something is wrong with the reporting, not the business.
If you want to pressure-test your own numbers, the break-even ROAS calculator is the right starting point - it forces you to write down the contribution margin most dashboards quietly hide.
The kill rule, applied to metrics
We run a 3x kill rule on creative testing: nothing scales past 3x its target cost per result. The same logic, applied to the metric stack:
- If platform ROAS is the only green number on the report, you have no read on the business.
- If MER drops two weeks in a row at flat spend, kill the optimisation that started two weeks ago.
- If CAC moves up faster than AOV, the creative isn't finding the buyer - go up the stack, not down.
This is where the always-on layer earns its keep. BAVai scans the stack every morning at 7am and flags when the three numbers stop agreeing. The human still makes the call. The machine makes sure no one is reading a polite lie for three weeks.
The takeaway
ROAS vs MER vs CAC isn't a debate about which is right. All three are right, at different altitudes. The mistake is letting the loudest one - platform ROAS - be the only number in the room.
Run the full stack. Show your CFO MER. Show your board CAC. Use ROAS to diagnose, not to declare victory.
If you can't show all three numbers, and explain why they're moving the way they are, which one are you actually optimising?