Every agency and every course talks about ROAS. Almost none of them answer the only question that matters: at what multiple do I actually make money? Let's settle it — the formula, your break-even line, and the budget rules that follow from it.
ROAS is revenue divided by ad spend. Spend $1,000, get $3,000 of revenue — that's 3x. The arithmetic is trivial; the trouble starts with interpretation, because whether 3x is good news depends entirely on your margins.
A 3x ROAS can be a quiet loss
Take a $20 t-shirt: $9 landed cost, $3 shipping and packaging, ~3% payment fees, and the sales-tax slice that was never yours. Roughly $5.60 remains before ads — a ~28% contribution margin. At 3x ROAS you spent $6.70 of ads to sell that shirt. You lost about a dollar per unit while the revenue chart looked great. This exact math quietly drains thin-margin stores everywhere.
Your break-even ROAS: one formula
Break-even ROAS = 1 ÷ contribution marginmargin = (price − product cost − shipping − fees − tax slice) ÷ price
| Margin | Break-even ROAS | Comfortable target (+30%) |
|---|
| 50% | 2.0x | 2.6x |
| 40% | 2.5x | 3.3x |
| 30% | 3.3x | 4.3x |
| 20% | 5.0x | 6.5x |
Thin margins demand brutal ROAS. Sometimes the winning ad move is not better ads — it is fixing the margin structure first.
Want these numbers for your own store?DashBooster computes revenue, net profit and ROAS in real time · 7-day free trial
Start free →Four ways sellers misread ROAS
- Counting tax as revenue. The VAT/sales-tax slice inflates ROAS by its rate. Compute on net.
- Ignoring pixel loss. Post-iOS-14 tracking misses 20-35% of purchases, so true ROAS runs above what Ads Manager shows — the classic error is killing winners. Fix the measurement with the Conversions API before judging campaigns.
- Judging on yesterday. Daily ROAS is noisy. Decide on 3- and 7-day windows.
- Reading only the account average. A 3x account often hides a 5x winner subsidizing a 1.5x loser. Judge per ad set.
Budget rules that follow
- 7-day ROAS ≥ break-even × 1.3: scale — but at most ±20% every 48 hours, or you reset the learning phase and manufacture volatility.
- Hovering near break-even: hold budget; fix creative or landing pages instead. Scaling here just scales the loss.
- 3-day below 7-day and both trending down: suspect creative fatigue first; cut only if a fresh creative doesn't recover it.
One more nuance: ROAS measures efficiency, not growth. An 8x campaign doing 2 orders a day builds less business than a 4x campaign doing 40. Past validation, deliberately trading some ROAS for volume — down to the break-even floor — is how accounts scale. And if your product earns repeat purchases, first-order ROAS slightly below break-even can still be profitable on
lifetime value.
FAQ
Q. Should I benchmark against industry-average ROAS?
No — industry averages describe other people's margins. Your target is your own break-even plus a safety buffer. Borrowed targets produce borrowed losses.
Q. Ads Manager ROAS never matches my store revenue. Which is right?
Both are partially wrong in opposite directions: attribution windows overcount, pixel loss undercounts, refunds arrive late. That is why serious operators reconcile ad data against actual order data.
Q. My ROAS swings wildly day to day. Broken?
Probably just small numbers — a handful of daily conversions cannot produce a stable ratio. Read weekly, and consolidate ad sets so learning accumulates.
🧷 Key takeaways
- ROAS = revenue ÷ spend, but the judgment line is break-even ROAS = 1 ÷ margin.
- Compute on net revenue, correct for pixel loss, judge on 3/7-day windows, per ad set.
- Scale winners ±20% per 48h; near break-even, fix creative — don't scale.
- Efficiency isn't growth: trade ROAS for volume deliberately, never below the floor.
Stop computing break-even by hand
DashBooster calculates your break-even ROAS from real margins and shows live ROAS against it, every day.
Start your 7-day free trial →