Open your Meta Ads Manager and Google Ads account side by side on a good week, and you'll likely see two numbers that both look great. Meta might report a 4.2x ROAS. Google might report 3.8x. Both dashboards are technically telling the truth about what they can see — and both are quietly overstating the picture, because neither knows what the other one already claimed credit for.
This is the single most common reason a founder or marketing manager feels like their numbers "don't add up" against their actual bank balance. The fix isn't a better dashboard. It's understanding which number each metric is actually built to answer.
01What each metric actually measures
ROAS (Return on Ad Spend) is revenue attributed to a specific platform, divided by spend on that platform. It's calculated inside each platform's own attribution model, using whatever tracking data that platform can see — which is the root of the problem.
MER (Marketing Efficiency Ratio), sometimes called blended ROAS, is total store revenue across every channel, divided by total marketing spend across every channel. It doesn't ask which platform gets credit for which sale. It only asks whether the whole marketing engine is profitable.
02Where the gap actually comes from
Attribution overlap is the biggest driver. A customer sees a Meta ad on Monday, searches your brand name on Google on Wednesday, then buys. Meta's pixel sees the click and claims the sale. Google's tracking sees the branded search click and also claims the sale. Your actual revenue only happened once — but two dashboards each recorded a full conversion for it.
Both platforms "look great" in isolation. The number that actually determines whether this was a good month — $50,000 in revenue against $20,000 in total spend — is meaningfully lower than either platform's self-reported figure. That 2.5x is what you'd use to decide whether to increase or cut total marketing budget next month.
03When to use which number
| Decision | Use This Number |
|---|---|
| Total marketing budget up or down | Blended MER |
| Reporting real performance to a client or investor | Blended MER |
| Which creative wins inside one campaign | Platform ROAS |
| Which ad set to scale within one platform | Platform ROAS |
| Comparing Meta's performance to Google's | Platform ROAS (with caution) |
Platform ROAS is still useful — it's just useful for a narrower question than most people ask of it. Inside Meta, comparing one ad set's ROAS to another's is a fair fight, because attribution overlap affects both roughly equally. The moment you start comparing Meta's number to Google's number, or using either one alone to decide total spend, you're comparing two numbers that were never designed to be compared.
04Turning MER into a spending decision: POAS
Knowing your blended MER tells you how the business performed last month. To decide how aggressively to spend next month, the number that matters is your target ROAS based on actual profit margin — sometimes called POAS (Profit on Ad Spend).
A brand running a 35% net margin needs a minimum ROAS of roughly 2.86x just to break even on the marketing spend itself, before accounting for any other overhead. Below that number, every "successful" campaign by a platform's own reporting is actually losing the business money in real terms.
- Pull total store revenue from Shopify/your storefront, not from any ad platform
- Add up total spend across every paid channel — Meta, Google, TikTok, everything
- Divide revenue by spend — this is your MER, recalculated weekly or monthly
- Compare against your POAS target (1 ÷ net margin) to know if you're actually profitable
05The honest way to report this
If you're reporting to a client, a partner, or your own leadership, lead with blended MER — it's the number that survives contact with the bank account. Platform ROAS belongs in the appendix, as a diagnostic tool for optimizing within a channel, not as the headline result.
The accounts that scale predictably aren't the ones with the highest platform-reported ROAS. They're the ones whose owners know the difference between the two numbers and have stopped making budget decisions based on the more flattering one.