Return on Ad Spend (ROAS) is often used as the primary indicator of campaign health, but it can be misleading. Platforms use attribution models (e.g., last-click, view-through) that frequently credit ads for conversions that would have happened anyway. As one specialist noted, Meta’s new “Incremental Attribution” reveals that classic ROAS “usually reports lower ROAS, but it’s far more realistic.”
🧠 The Problem with Traditional ROAS
-
Over-attribution by Platforms
Platforms like Meta and Google assign credit based on when users click or view ads, not on actual causation. This often inflates ROAS because conversions from retargeting or brand searches get credited repeatedly. -
Cannibalization and Duplication
Standard ROAS doesn’t account for cannibalization—where ads take credit for conversions consumers would have made organically. Non-experimental measurements can't untangle this effect. -
Skewed by Attribution Windows
Short attribution windows (like 1-day click/7-day view) can lead to inflated ROAS. Meta’s incremental attribution (available since April 2025) often reports significantly lower but more valid metrics.
✅ Better Alternatives: Measuring True Impact
A. Incrementality Testing
Also known as geo-holdouts or A/B tests, this method compares exposed vs. control groups to isolate the true lift generated by a campaign. It’s considered the gold standard, but it can be slow and resource-heavy.
- Geo-holdout: Run ads in some regions but not others to measure the incremental lift.
- Conversion lift studies: Use randomized A/B testing to quantify campaign-specific impact.
B. Incremental ROAS (iROAS)
iROAS calculates (Incremental Revenue – Control Revenue) ÷ Ad Spend. It reflects only the revenue directly driven by ads, excluding what would have happened organically.
📈 Key Metrics That Give Clarity
| Metric | What It Shows | Why It Matters |
|---|---|---|
| Incremental Lift | % improvement of a test vs. control group | Shows the true causal effect of your ads. |
| Incremental Conversions | The number of extra conversions caused by ads | Focuses on real, additional impact. |
| iROAS | (Incremental Revenue ÷ Ad Spend) | Shows true revenue per dollar spent, not vanity ROAS. |
🧩 Best Practices for 2025
-
Use Platform Incremental Attribution
Leverage Meta’s April 2025 “Incremental Attribution” metric instead of the default ROAS for a more realistic view. -
Run Incrementality Tests Regularly
Use geo-holdouts or A/B tests to uncover the true impact of your campaigns, especially before scaling your budget. -
Calculate iROAS
Move beyond vanity metrics by measuring incremental conversions and revenue to understand true profitability. -
Combine with Media Mix Modeling (MMM)
Pair incremental testing with MMM for a holistic, cross-channel view of your marketing performance.
✅ Final Recommendations
- Don’t rely solely on platform ROAS—it’s often inflated.
- Prioritize incrementality via experimentation and iROAS.
- Use platform and MMM insights to validate performance holistically.
- Align measurement with profitability, not vanity metrics.
By moving from attribution-driven ROAS to true incrementality, your campaigns will reflect actual impact—giving you the clarity needed to scale smarter in 2025.
