Your Meta ROAS says 4.2. Your Google ROAS says 3.8. Your client says sales went down. Somebody is lying. Spoiler: all three. Getting more sales with AI starts with measuring what is really happening.
What ROAS is and how to calculate it correctly in 2026
The formula is simple: ROAS = Revenue / Spend. If you invest €1,000 and generate €4,000 in sales, your ROAS is 4. But that is where the simplicity ends.
The problem is that “revenue generated” depends on how conversions are attributed. And Meta, Google and GA4 have three different answers to the same sale.
Why Meta, Google and GA4 never say the same thing
Real scenario
A user sees an Instagram ad on Monday. On Thursday they search for the product on Google and click a search ad. On Friday they go directly to the website and buy. Meta claims the sale (7-day window). Google claims the sale (last paid click). GA4 says the source was “direct”. All three are half right. None is entirely right.
Result: if you add Meta’s ROAS to Google’s ROAS, the number you present to the client is between 40% and 70% higher than the real ROAS [1]. Not because you are lying. Because the tools are designed to claim the credit.
Siloed ROAS vs cross-platform ROAS: the table nobody wants to see
| Metric | Siloed ROAS | Cross-platform ROAS (real) |
|---|---|---|
| How it is calculated | Each platform divides its attributed revenue by its spend | Unique revenue (GA4) divided by total spend across every platform |
| Double-counting | Yes. The same sale appears in Meta AND in Google | No. GA4 records the conversion only once |
| Accuracy | Optimistic. Can inflate ROAS by 40% to 70% | Real. Reflects what the ad spend actually generated |
| Decisions it drives | Keep investing in channels that look profitable but aren’t | Identify which channel generates the most real incremental sales |
| What you tell the client | “Our combined ROAS is 7.8” | “Our real ROAS is 4.2” |
How to get more sales with AI by measuring real ROAS
With boostiq, this analysis happens automatically: you connect Meta Ads, Google Ads and GA4; boostiq normalises the data with GA4 as the source of truth; you calculate real cross-platform ROAS in seconds; you identify which channel generates incremental sales; and you reallocate budget on real data.
The result according to internal pilot data: an average 22% increase in real ROAS within the first 60 days [2]. Not because the campaigns magically improve. Because the money stops going where it seemed to work and starts going where it really works.
The most common mistakes when measuring ROAS
- Trusting Meta’s ROAS without cross-referencing it with GA4. Meta will always claim more than its share. That is its financial incentive.
- Comparing Meta and Google ROAS as if they were the same metric. They are not. The attribution models are different.
- Not including management costs. If you pay an agency €2,000/month and €5,000 in advertising, your real investment is €7,000, not €5,000.
- Budget decisions without normalised data. This week’s ROAS may be a mirage of late attributions from last week.
Discover your real ROAS. In seconds.
boostiq cross-references Meta, Google and GA4 and calculates it for you automatically. No spreadsheets, no double-counting.
Start your free 14-day trial →No card · No technical setup · Set up in 5 min
Frequently asked questions about ROAS and more sales with AI
What ROAS is considered good in digital advertising?
It depends on the sector and the product margin. What matters is not the absolute number, but whether it covers costs and generates net profit.
How do I calculate real ROAS if I have Meta and Google active at the same time?
Use GA4 as the source of truth for unique conversions. Divide revenue by the sum of Meta spend plus Google spend. boostiq calculates it automatically.
Why does Meta always have a higher ROAS than Google in my reports?
Because Meta uses a wider attribution window and claims conversions that would have happened anyway without the ad. It doesn’t mean Meta performs better: it means it claims more.
Can AI improve ROAS or only measure it better?
Both. Measuring ROAS correctly already improves budget decisions. In addition, boostiq detects optimisation opportunities and can apply changes with your approval.
What if my real ROAS is much lower than I thought?
It is uncomfortable, but it is good news: it means you have a real diagnosis and can make the right decisions. With real ROAS you can identify which channel genuinely generates sales and concentrate the budget there.
Sources
- boostiq internal pilot data (2025-2026). Double-counting analysis in accounts with Meta and Google active simultaneously.
- boostiq internal pilot data (2025-2026). Sample: 18 accounts with combined spend above €5,000/month.
- Google Analytics Help, “About attribution models”. Available at support.google.com/analytics.
