Margin-Band Gating: What ROAS Target Your Margin Allows
A peer mentions they’re holding 4x ROAS and you feel behind. But 4x is a victory for one brand and a slow bankruptcy for another. The benchmark that matters is not the one a competitor quotes — it’s the one your margin will tolerate. Your margin sets a hard floor under your ROAS target, and that floor moves by a factor of four or more between a thin-margin brand and a fat-margin one. Chase someone else’s number and you’ll either leave growth on the table or scale yourself into a loss.
For the neighboring economics, compare Returns and Refunds: Building Net Revenue Into Your True ROAS and use ROAS Dropped Overnight: Real Signal or Tracking Artifact? to validate the measurement decision.
Break-even ROAS is just the inverse of your margin
Start with the one piece of math that isn’t up for debate. If your contribution margin is m (the fraction of revenue left after the variable cost of fulfilling an order), then ad spend breaks even when the gross profit it generates equals the spend itself.
Profit from a sale funds the ads. So you break even when:
revenue × margin = ad spend
Rearranged, that’s revenue ÷ ad spend = 1 ÷ margin. Revenue over ad spend is ROAS. So:
Break-even ROAS = 1 ÷ contribution margin.
A 20% margin needs 5.0 just to stand still. An 80% margin breaks even at 1.25. Same platform, same auction, two completely different definitions of “working.” This is the whole reason a single ROAS benchmark is useless advice.
The margin-to-target table
Here is the relationship laid out as a planning tool. Break-even is exact math. The working target adds a profit buffer on top, and those ranges are illustrative — set the exact number to your own profit goal and fixed-cost load, not as a assurance.
| Contribution margin | Break-even ROAS (1 ÷ margin) | Illustrative working target |
|---|---|---|
| 20% | 5.0 | 6.5 – 7.0 |
| 40% | 2.5 | 3.0 – 3.5 |
| 60% | 1.67 | 2.0 – 2.2 |
| 80% | 1.25 | 1.4 – 1.6 |
The working targets above leave roughly mid-single-digit to mid-teens profit on revenue after ad spend. You can derive your own: profit margin on revenue after ads equals margin − (1 ÷ ROAS). Pick the after-ads profit you need, add it to break-even, and you have your floor. That is the entire exercise behind “roas target by margin” — invert the margin, add the cushion, defend the line.
Notice the asymmetry in the spread. The thin-margin brand’s target sits far above its break-even in absolute ROAS terms because every point of efficiency is scarce. The fat-margin brand’s target barely clears break-even, because it has room to spare.
Gate on contribution margin, not the number on your P&L
The common way this goes wrong: people plug in headline gross margin — revenue minus cost of goods — and build a floor that’s too generous.
Headline gross margin ignores the variable costs that hit every single order: shipping and the share you don’t recover, payment processing, pick-and-pack, and the expected cost of returns. Strip those out and you get contribution margin, which is what’s actually available to fund acquisition.
The gap is large. A brand quoting 60% gross margin can easily sit at 45% contribution once shipping subsidy, processing, and a realistic return rate come out. That moves break-even ROAS from 1.67 to 2.22 — a third higher. Gate on the inflated number and you’ll greenlight campaigns that look profitable on the dashboard and quietly bleed contribution. Always invert the contribution margin, never the gross.
Platform ROAS reads high — set the floor above the math
There’s a second adjustment, and it’s about which ROAS you’re even looking at.
Platform-reported ROAS can read higher than the revenue you can actually attribute to the spend. Attribution windows count conversions that would have happened anyway, view-through credit is generous, and overlapping campaigns double-claim the same buyer. The blended view — total revenue divided by total ad spend, your MER — is the honest backstop, and it’s almost always lower than what any single campaign reports.
Two practical moves:
- Set your platform-ROAS floor above the pure margin math. If your break-even is 5.0, holding Meta-reported ROAS at exactly 5.0 likely means you’re underwater on a blended basis. Build in a haircut so the reported floor clears your real one.
- Reconcile against MER weekly. Pick a platform target that, given your typical over-attribution, lands your blended MER at or above the margin-derived floor. The platform number is the steering wheel; MER is the road.
The asymmetry nobody plans for
Once you internalize that the floor is a function of margin, the strategy writes itself differently for each band.
A high-margin brand (70%+) breaks even below 1.5. That means it can afford to buy growth at ROAS levels that would terminate a low-margin brand. It can fund more creative tests, sit through longer learning periods, and bid into colder, higher-intent-cost audiences — because there’s contribution to absorb the inefficiency. For these brands, an obsession with a high ROAS number is frequently the mistake: they’re under-spending against a floor they’re nowhere near.
A 20–30% margin brand lives on the opposite edge. Break-even is 3.3 to 5.0, the working floor is higher still, and there’s almost no slack for waste. Every dead ad set, every campaign left in a perpetual learning reset, every 20–40% of spend that delivers nothing (treat that as an illustrative planning range, not a fixed law) comes straight out of a margin that was never there. These brands win on surgical discipline, tight consolidation, and ruthless pruning — not on volume of experiments.
Same tactics, opposite prescriptions, entirely because of where the floor sits.
Holding the floor without strangling delivery
A margin-derived target is a destination, not a daily verdict. The quickest way to lose money with the right number is to enforce it too early.
- Don’t judge below the floor on day two. A new campaign needs enough recent optimization-event signal before delivery stabilizes — frequently on the order of a few dozen conversions, as a planning rule of thumb rather than a published threshold. Kill it before that and you’ve paid for the learning without collecting the return.
- Watch frequency as a leading indicator. When ROAS drifts below floor and frequency is climbing, you’re saturating the audience, not failing creative. The fix is reach, not a budget cut.
- Allow a payback window for genuine repeat buyers. If you have real, measured repeat purchase behavior, you can run first-order ROAS below the single-purchase floor and let lifetime contribution close the gap. Only do this with measured retention — not a hoped-for one.
- Re-derive the floor when costs move. Shipping rates, return rates, and processing fees all shift contribution margin. When they move, your break-even moves with them. A floor set last year against last year’s costs is a stale floor.
This is also where a read-only operator layer earns its keep: Bach reads the account, computes the floor from your actual contribution margin, and flags the campaigns running under it — surfacing the leak with its contribution impact before you approve any change, never acting on its own.
The takeaway
Stop benchmarking against other brands’ ROAS. Compute your contribution margin — after shipping, fees, and returns, not the headline number — invert it to get break-even, add the profit buffer you need, and that’s your floor. A 20% brand and an 80% brand are playing different games on the same board. One number was never going to fit all of them; the right number was inside your own margin the whole time.