aMER and the 'Spending More, Making the Same' Scaling Trap
You add 30% to the budget. Platform-ROAS barely flinches — still a healthy multiple in the ad manager, still green. But the actual revenue line, the money that lands, is flat. You’re spending more and making the same, and the dashboard insists everything is fine. That gap between what the platform reports and what the business banks is the most expensive blind spot in paid acquisition. The metric that closes it is the acquisition marketing efficiency ratio.
For the neighboring economics, compare Contribution Margin CM2 vs CM3: The Real Meta Ads Scaling Gate and use The Winback Flow to Run Before You Re-Acquire on Meta to validate the measurement decision.
Why platform-ROAS keeps reporting “fine”
Platform-ROAS measures what the channel can claim, not what your business netted. Several mechanics let it hold steady while real growth stalls:
- Attribution credit drifts upward as you scale. A widening or click-and-view attribution window lets the platform take credit for purchases that would have happened anyway — returning customers, brand searchers, people already mid-consideration.
- Returning customers get counted as conversions. The pixel doesn’t care whether a buyer is brand-new or their fifth order this quarter. Both inflate reported ROAS.
- Optimization harvests the easy demand first. Early spend reaches the highest-intent pockets. As budget climbs, delivery reaches further into lower-intent inventory, but the average ROAS the platform reports barely moves because the cheap conversions are still in the blend.
The result: reported ROAS can sit at a flat 4x for weeks while the incremental revenue from each new increment of spend decays toward zero. The number isn’t lying about the average. It’s just structurally incapable of showing you the margin — the value of the next unit of spend.
aMER: the ratio attribution can’t game
Start with blended MER (marketing efficiency ratio): total revenue divided by total ad spend across all channels. It’s honest in a way platform-ROAS isn’t, because it uses the revenue you actually collected and the spend you actually paid — no pixel, no attribution model, no double-counting between ad sets.
But blended MER carries its own blind spot: a loyal repeat base. A strong returning-customer cohort can prop up MER even while new-customer acquisition quietly stalls. You look efficient in aggregate and miss that the acquisition engine has seized.
The acquisition marketing efficiency ratio — aMER — removes that cover:
aMER = new-customer revenue ÷ total ad spend
It strips out repeat revenue and forces the only question that matters when you scale: is paid media still buying new customers efficiently, or am I paying to re-harvest people who would have come back on their own? Two accounts with an identical 4x platform-ROAS can have wildly different aMER. The one running mostly on repeat demand is a retention business wearing an acquisition costume — and it will break the moment you try to scale on it.
The plateau, in one view
Here’s the trap as a worked example. Treat the figures as illustrative planning numbers, not a benchmark to copy.
| Spend tier | New-customer revenue | aMER | Platform-ROAS |
|---|---|---|---|
| Baseline | $160k | 4.0 | 4.2x |
| +50% budget | $204k | 3.4 | 4.1x |
| +100% budget | $228k | 2.85 | 4.0x |
Platform-ROAS looks rock-steady — a story of healthy, controlled scaling. aMER tells the opposite story: efficiency is bleeding out as you add budget. But even the average aMER understates the problem. Look at what each increment actually bought:
- From baseline to +50%, an extra slice of spend bought +$44k new revenue → marginal aMER 2.2.
- From +50% to +100%, the same-sized extra slice bought only +$24k → marginal aMER 1.2.
The last increment is where the business is. If your breakeven sits near a 2.0 ratio (more on that below), that final budget bump is underwater — you spent more to acquire customers at a loss — while the ad manager still proudly reports a ~4x return. That is the “spending more, making the same” trap rendered in numbers.
Read the marginal, not the average
The single most important move is to stop reading aMER as a single blended number and start reading it as a curve. Average aMER is a lagging, flattering summary; marginal aMER — the change in new-customer revenue divided by the change in spend — is the leading signal of saturation.
Three signatures tell you you’re in the trap:
- Platform-ROAS flat, aMER falling. Classic saturation. The channel’s average looks fine; your acquisition economics are eroding.
- New-customer count flat while spend rises. You’re paying more for the same people. CPA is inflating even if blended efficiency hides it.
- Marginal aMER below your breakeven. The decisive one. The last increment of spend is destroying contribution, independent of what the average says.
Tie the target to contribution margin
aMER is only meaningful against a threshold you set from unit economics, not from a vanity ROAS goal. The honest version is simple:
Breakeven aMER ≈ 1 ÷ contribution margin
If your contribution margin after cost of goods, shipping, fulfillment, and payment fees is 50%, you need new-customer revenue worth roughly twice your spend — a 2.0 — just to break even on acquisition before any profit. Thinner margins demand a higher aMER; you cannot scale your way out of weak unit economics. This is also why a “good” platform-ROAS is a meaningless target in isolation: it’s disconnected from the only number that decides whether growth is profitable.
If you also count a new customer’s expected repeat value, you can justify a lower first-order aMER — but only if your retention data is real, not aspirational. Borrowing against future repeat purchases that haven’t shown up in the cohorts yet is how brands talk themselves into scaling at a loss.
What to do when aMER stalls
- Hold spend and re-baseline. Freeze budget for a clean window and measure new-customer volume and aMER at steady state. You can’t diagnose a curve you’re still pushing.
- Run a holdout to check incrementality. Suppress a randomized slice of audience and measure the lift in actual new customers. If revenue barely dips when you withhold spend, much of your reported ROAS was harvest, not incrementality — and your true aMER is worse than the dashboard’s.
- Find the saturation lever. Climbing frequency on the same audience, rising CPA at flat conversion volume, and creative fatigue all read identically on platform-ROAS but show up cleanly when you watch the marginal curve. Refresh creative and open genuinely new audiences before adding budget, not after.
- Set a marginal-aMER floor as a scaling gate. Add the next budget increment only while marginal aMER stays above breakeven. The moment it crosses under, the next increment is contribution-negative — stop, fix the lever, then resume.
This is exactly the kind of monitoring that’s easy to specify and tedious to do by hand. A read-only operator like Bach AI can watch the marginal curve continuously and flag the tier where the last increment of spend dropped below your breakeven aMER, then propose the budget hold or audience expansion for you to approve before anything changes.
The takeaway
Platform-ROAS answers “what can the channel claim?” aMER answers “is paid media still buying new customers profitably?” — and its marginal version answers the only question that governs scaling: “is the next increment of spend worth it?” Build the aMER curve from banked new-customer revenue, anchor your threshold to contribution margin, and gate every budget increase on marginal aMER staying above breakeven. Do that, and the plateau where spending more makes the same stops being a surprise on the bank statement. It becomes a line you can see coming — and a line you choose not to cross.