ROAS stands for Return on Ad Spend. The calculation is simple: revenue generated divided by the amount you spent on advertising.
£10,000 in sales from £2,500 in ad spend = a ROAS of 4. Also written as “4:1” or “4x.” All the same thing.
What’s a good ROAS? The honest answer is: it depends entirely on your margins, and anyone giving you a single number without knowing your business is guessing.
The 4x Myth
You’ll often see 4:1 cited as the benchmark for “good” ROAS — four pounds of revenue for every pound spent on ads. This figure has spread through the marketing world because it’s been repeated enough times to feel authoritative. It isn’t.
A 4x ROAS is profitable for some businesses and catastrophically loss-making for others, depending entirely on your margins.
Example: If you sell a product for £100 that costs you £80 to produce and deliver, your gross margin is 20%. A 4x ROAS means you spent £25 to generate £100 in revenue — which leaves you with £20 in gross profit before paying for the ad. You’ve spent £25 to make £20. That’s a loss.
Example two: If you sell a digital product for £100 with a 90% margin, a 4x ROAS means you spent £25 to make £90 in gross profit. That’s a very good outcome.
The 4x rule is useless without knowing your margin.
How to Calculate the ROAS You Actually Need
The ROAS you need to break even on ad spend is:
Break-even ROAS = 1 ÷ Gross Margin
If your gross margin is 40%, your break-even ROAS is 1 ÷ 0.4 = 2.5. Anything above 2.5x and you’re covering your cost of goods and your ad spend. Anything below and you’re losing money on every sale.
To make a profit on top of that, your ROAS needs to exceed your break-even figure by enough to cover your other operating costs. Most e-commerce businesses need to be well above break-even ROAS to account for returns, overheads, and payment processing fees.
Blended ROAS vs Channel ROAS
Most businesses run ads on multiple platforms — Google, Meta, TikTok, maybe others. Each platform reports its own ROAS, and those numbers are almost certainly inflated, because each platform takes credit for sales that the others also influenced.
A customer who saw your Instagram ad twice, clicked a Google Shopping ad, and then came back direct to purchase — that sale will show up as a conversion in Meta’s reporting and in Google’s reporting. Both platforms count it as theirs.
Blended ROAS sidesteps this problem. Instead of trusting any single platform’s attribution, you take your total revenue and divide it by your total ad spend across all channels.
Total revenue last month: £50,000. Total ad spend: £12,500. Blended ROAS: 4x.
This is less granular — you can’t immediately tell which channel is performing — but it’s honest. Use it as your primary health metric, and use individual channel ROAS for directional decisions rather than gospel.
When ROAS Is the Wrong Metric
ROAS measures revenue, not profit. High-revenue, low-margin businesses can have impressive ROAS numbers that translate to very thin profits. If you’re running promotions or selling loss-leaders, ROAS looks great while profitability suffers.
For businesses with variable margins across products or customer segments, consider MER (Marketing Efficiency Ratio) or cost per acquisition alongside ROAS. A sale isn’t equally valuable if one product has 60% margin and another has 10%.
And for subscription businesses or anything with meaningful repeat purchase, pure ROAS misses the customer lifetime value entirely. Acquiring a customer at a “bad” ROAS might be completely sensible if that customer buys from you ten more times over the next two years.
The Bottom Line
ROAS = Revenue ÷ Ad Spend. Good ROAS = whatever generates profit given your margin, not a number you heard on a marketing podcast. Calculate your break-even ROAS from your gross margin. Set your target above that to cover operating costs. Track blended ROAS rather than trusting individual platform attribution. And if your margins vary, or you’re in a subscription model, ROAS alone doesn’t tell the full story. The formula is easy. The thinking behind it requires knowing your own numbers.
