Why it matters
ROAS is the quickest read on whether your ads are paying their way. It puts campaigns, ad sets and creatives on one scale: how much revenue each rupee or dollar of spend brings back. On its own, though, it says nothing about profit. A ROAS of 3 can make money for one brand and lose it for another, which is why ROAS only means something next to your break-even ROAS.
How to calculate it
Divide the revenue attributed to your ads by what you spent to get it, or let the ROAS calculator do it — including the reverse sum of how much you can spend at a target ROAS. In Ads Manager, the Purchase ROAS column uses the purchase values sent by your pixel and Conversions API.
The catch is the word "attributed". Meta credits a purchase to an ad only when it happens inside the ad set's attribution window — for example within 1 day or 7 days of a click, and optionally within 1 day of a view. A wider window can only count the same purchases or more, so the same campaign reports a higher ROAS on 7-day click plus 1-day view than on 1-day click. Check which window a ROAS figure uses before comparing it with another.
Worked example: platform ROAS vs real ROAS
Example: you spend ₹50,000 in a month and Meta attributes ₹1,40,000 of purchases.
- Platform ROAS = 1,40,000 ÷ 50,000 = 2.8
- Your store shows only ₹1,10,000 of those orders delivered and paid for, after cancellations, returns and cash-on-delivery parcels refused at the door (return to origin, or RTO).
- Real ROAS = 1,10,000 ÷ 50,000 = 2.2
Same ads, two verdicts. If your break-even ROAS is 2.5, the platform number says scale and the real number says fix. Use platform ROAS to choose between ads. Use real ROAS, from your own order data, to judge whether advertising is making money.
What's a good ROAS?
There's no universal good number, because it depends on your margin, price and repeat rate. A brand keeping 70% of each sale after costs can thrive at a ROAS that would sink a brand keeping 30%. Work out your break-even point with the break-even ROAS calculator, then set a target above it that leaves the profit you need.
How to improve it
ROAS breaks down into four drivers, with CTR measured on link clicks and conversion rate as purchases per link click:
ROAS = CTR × conversion rate × average order value × 1,000 ÷ CPM
Example: 1.5% CTR × 2.5% conversion rate × ₹1,200 order value × 1,000 ÷ ₹180 CPM = a ROAS of 2.5. Improve any one driver — more clicks per impression, more orders per click, bigger baskets or cheaper reach — and ROAS rises in proportion. Fly ROAS's analytics splits ROAS into these same drivers, so when results change you can see which one moved.
Common mistakes
- Comparing ROAS figures that use different attribution windows.
- Treating platform ROAS as cash in the bank without checking returns and RTO.
- Judging ROAS on a few days of data or a handful of orders.
- Using one target ROAS for products with very different margins — POAS handles that better.
- Forgetting that ROAS across all channels, or MER, is what funds the business.
Frequently asked questions
What is a good ROAS for Facebook ads?
There's no single good ROAS, because it depends on your margins. Work out your break-even ROAS (selling price ÷ margin per order) and aim above it by the profit you need. A brand keeping 60% of each sale after costs breaks even at a ROAS of about 1.7, while one keeping 25% needs 4.0 just to break even.
What is the difference between ROAS and ROI?
ROAS compares revenue with ad spend only. ROI compares profit with the total investment, so it accounts for product costs, fees and overheads. A campaign can show a ROAS of 3 and still have a negative ROI if margins are thin.
Why is ROAS in Ads Manager different from my store's sales?
Meta counts purchases inside its attribution window, including some people who only viewed an ad, and records orders when they're placed. Your store sees every order whatever the source, plus later cancellations, returns and refused cash-on-delivery parcels. Tracking gaps also cause differences, so use Meta's figure to compare ads and your store's figure to judge profit.
