Facebook Ads vs Google Ads: Compare ROAS and CPA Fairly

Each platform reports its own ROAS using its own attribution. Putting the two side by side tells you less than it seems to, and adding them together tells you less still.

By Ed · Published October 25, 2026 · 9 min read

Twenty-five years starting companies, consulting, owning a multi-million-dollar P&L and leading business intelligence for large enterprises.

Facebook Ads and Google Ads ROAS cannot be compared directly because each platform reports conversion value under its own attribution rules, and both can claim the same sale. The fair method is to compute each platform's ROAS as a ratio of sums, compare the total they claim with your actual revenue, and use blended ROAS from your own revenue to make budget decisions.

Why the two ROAS figures do not compare

ROAS is conversion value divided by spend. The spend side is solid: both platforms bill you, and you can check it. The value side is where they diverge.

Different attribution. Each platform credits a purchase to an ad interaction within its own window and rules. A purchase that follows a click in one place and a view in the other can be credited to both.
Different roles in the funnel. Search ads often catch people who were already looking, partly because of something they saw elsewhere. Social ads often create the interest. Reported ROAS rewards whoever is closest to the sale.
Double counting. A customer who saw a Meta ad, then clicked a Google search ad and bought, is claimed by both. Add the two claimed values and you will often exceed actual revenue.
Different counting of what is a conversion. A purchase event, a lead, an add-to-cart. Check the definitions before comparing.

The result: a Meta ROAS of 3.0 and a Google Ads ROAS of 4.0 do not tell you Google Ads is better. They tell you what each platform believes about itself.

The arithmetic, with an example

All numbers below are an illustration, not real data.

             Spend    Claimed value    ROAS (value / spend)
Google Ads   10,000       40,000       4.00
Meta          8,000       24,000       3.00
Total        18,000       64,000       3.56

The combined 3.56 is 64,000 divided by 18,000. It is not the average of 4.00 and 3.00, which would be 3.50. The two differ because Google Ads spent more, so it carries more weight. This is the ratio-of-sums rule, explained in CPA vs CPL.

Now compare with reality. Suppose your store recorded 45,000 of revenue from all sources in the same period. The platforms together claim 64,000, which is 19,000 more than the store sold in total. The blended ROAS from your own revenue is 45,000 divided by 18,000, which is 2.50. That is the number that cannot be inflated by overlap.

Claimed by platforms   64,000
Actual store revenue   45,000
Overlap or non-ad      19,000 (at minimum)
Blended ROAS           45,000 / 18,000 = 2.50

The excess is not just double counting. Organic and direct sales are also in the store figure, so the true overlap is larger than the 19,000 difference. Nothing in an export lets you split it exactly.

CPA has the same problem

Cost per acquisition is spend over conversions per platform. The same caution applies: a conversion on one platform is "a purchase within its window", and on the other might include a different event or a longer window. A CPA gap between platforms is a lead to investigate, not a ranking. Our guide to ROAS by channel walks the by-hand version, and combining Google, Meta and LinkedIn exports covers lining up the columns.

What to use for decisions

Within a platform, use that platform's ROAS to rank campaigns against each other. The same rules apply to all of them, so the ranking is meaningful.
Across platforms, use blended ROAS from your revenue, and look at how it moves when you shift budget.
For the incrementality question, only a test answers it: pause or cut a channel in a region or period and watch total revenue. Reported ROAS cannot.

Doing the comparison with Quiriz

Quiriz has Google Ads and Meta Ads connectors, both in beta, read-only, refreshed daily and included in paid plans. Both produce campaign performance datasets with the same column names (date, campaign, cost, conversions, conversion value), and Google Ads ROAS and Meta Ads ROAS are each defined once as conversion value over spend. Google Ads conversions are as Google Ads counts them; Meta Ads conversions are purchases under the ad account's attribution setting. So you can ask:

"By month, what did Google Ads and Meta Ads each spend, and what conversion value did each report?"
"What was the combined spend and combined reported value last quarter, and the ratio?"
"What was ROAS for each platform last quarter, side by side?"

Quiriz does the ratio of sums for you. To get blended ROAS from real revenue you also need your store or accounting data in Quiriz, for example a Shopify or other revenue dataset, and then you can ask for revenue over total spend across both.

Three comparison mistakes to avoid

Comparing unlike conversions. A Meta figure that is purchases and a Google Ads figure that includes leads or add-to-cart actions are not the same event. Restrict both to purchases before you compare.
Comparing unlike windows. A seven-day window and a one-day window will produce different ROAS on identical sales. Note each window beside the number.
Moving budget on the gap alone. Shifting spend from the lower-reported platform to the higher-reported one assumes the reported difference is real. It may be an attribution difference. Move a modest share, then watch blended ROAS and total revenue for a few weeks. If total revenue holds and spend falls, the move was right whatever either dashboard said.

Limits

The connectors report each platform's own attribution. They cannot remove the overlap between them.
Meta's value is purchase value; Google Ads' value is conversion value for whichever conversion actions your account counts, so check they refer to the same event.
Both datasets cover 13 months; recent conversions can be restated by either platform.
Currencies are the account's currency. If the two accounts use different currencies, convert before combining.

For ecommerce brands reconciling to store revenue, see Quiriz for ecommerce; for the platform pages, Google Ads AI and Meta Ads AI.

Compare your two ad platforms fairly

Connect Google Ads and Meta Ads and ask for spend, reported value and ROAS by month for both.

Try Quiriz free →

Frequently asked questions

Is Facebook Ads or Google Ads better for ROAS?

Neither is better in general. Each reports ROAS under its own attribution, and Google Ads search captures people already looking while Meta often introduces a product to people who were not. Reported ROAS therefore reflects the different jobs as much as performance. Compare using your own revenue and testing, not the two dashboards.

Why is the ROAS from Meta and Google Ads more than my actual revenue?

Because each platform credits itself for conversions it influenced, and the same customer can have clicked both. When you add the conversion values the platforms report, you often exceed actual revenue. The excess is double counting and no export tells you its size.

How do I calculate blended ROAS?

Divide total revenue from your store or accounting system by total ad spend across all platforms for the period. It does not depend on any platform's attribution, so it is the stable number, though it does not say which channel deserves the credit.

Can I average Meta ROAS and Google Ads ROAS?

Not by averaging the two ratios. A ROAS of 4.0 on 10,000 of spend and 3.0 on 8,000 combine to the total value over total spend, not 3.5. Add the numerators and denominators first, then divide.

What about CPA across the two platforms?

Same rule: total spend over total conversions per platform. But the conversions are defined differently, so a CPA of 30 on Meta and 45 on Google Ads is not a like-for-like gap. Check what each counts as a conversion first.