Platform ROAS or real margin: read the right number before reallocating a budget

Each platform measures its own ROAS, with its own attribution rules. Added together, these figures often exceed actual revenue.

By Ahmat Kard3 min read

Why ROAS figures do not add up

The same customer can see a Meta ad, later click a Google ad, then buy. Each platform applies its own attribution model and conversion window; both can claim the sale.

As a result, the sum of revenue attributed in each interface often exceeds the revenue actually collected.

What ROAS does not see

ROAS compares attributed revenue to ad spend. It ignores returns, discounts, shipping costs and cost of goods. Two campaigns with the same ROAS can leave very different margins.

Build a shared view

A shared view fits in three columns per source:

  • Spend, taken from each platform.
  • Opportunities and sales, taken from the CRM with the source stored when the lead arrived.
  • Revenue or margin, taken from the CRM or accounting, after returns and discounts.

What about automated bidding?

Google Ads target ROAS bidding optimises the conversion value you send it. If that value is gross revenue, it maximises gross revenue. Sending a value closer to margin, when possible, aligns the algorithm with what matters to the business.

A worked example

Illustrative figures. Campaign A spends $1,000 and generates $4,000 in sales: a ROAS of 4. But it sells products at a 30% margin: $1,200 gross margin, minus $1,000 of advertising, so a $200 gain.

Campaign B spends $1,000 and generates $3,000 in sales: a ROAS of 3. It sells products at a 60% margin: $1,800 gross margin, minus $1,000 of advertising, so an $800 gain. The lower ROAS earns four times more.

Know your break-even point

Break-even ROAS is simple to compute: 1 divided by the margin rate. With a 40% margin, you need a ROAS of 2.5 not to lose money on advertising; with a 25% margin, you need 4.

This threshold changes from one product to another. A single ROAS target for the whole catalogue pushes the algorithm towards products that sell easily, not those that earn.

A blended ratio as a cross-check

To check that the whole picture holds, divide total revenue by total ad spend, all platforms combined. This ratio does not say which platform produced what, but it cannot count a sale twice. If it falls while platform ROAS figures rise, attribution is telling you too flattering a story.

How often to decide

Two distinct rhythms prevent overreacting:

  • Every week: read each platform’s numbers to adjust campaigns within that platform.
  • Every month: compare platforms on margin and CRM revenue, then reallocate budget.
  • At every attribution change on a platform: note the date and do not compare before and after.

Questions

Should platform ROAS be ignored?

No: it is still useful to compare campaigns within the same platform. It simply does not settle choices between platforms.

What can I do without a CRM?

Start by storing the source of each order or lead in a spreadsheet, then compare actual sales by source every month.

Sources