Meta incremental attribution model

The Facebook (Meta) incremental attribution model is a modern approach to measuring the real impact of ads by estimating how many conversions are truly caused by your advertising—rather than just credited due to exposure or clicks. This model addresses the limitations of traditional attribution methods by isolating “incremental” conversions: those that would not have occurred without seeing an ad.

👍

Good to know

Because AnyTrack sends all conversions to Facebook Ads (Meta Ads) (not just the conversions that had an interaction with your ads) via the Conversion API, it enables Meta Ads to apply the Incremental Model.

What Is Incremental Attribution?

Incremental attribution evaluates the additional value your ads deliver beyond what would have happened organically. It relies on advanced machine learning and statistical techniques, including holdout testing and lift studies. In holdout testing, a segment of your target audience is deliberately excluded from seeing your ads (the “holdout group”), while other users continue to see ads as usual (the “test group”). The conversion rates between these groups are compared to determine the true incremental lift, i.e., what additional conversions were driven by the ads themselves.

How Does It Differ from Traditional Attribution?

Traditional Meta/Facebook attribution models are rules-based, assigning credit to ads within a set time window (like 7-day click or 1-day view), even if the user might have converted anyway. Incremental attribution, instead, focuses on measuring causal impact: conversions that likely wouldn’t have occurred without ad exposure.

How Is the Model Used in Practice?

• The model is activated in Facebook Ads Manager via attribution settings (“Incremental Attribution” checkbox).
• Advertisers can compare performance across standard and incremental metrics, such as Incremental Conversions, Incremental ROAS, and Incremental Purchase Value.
• This helps you identify which campaigns are truly driving new customer actions—and not just taking credit for behavior that would have happened anyway.

Meta’s data shows that using incremental attribution can reveal over 20% more incremental conversions in some cases.

Why Is It Important?

Incremental attribution allows marketers to:

• Optimize budget allocation by prioritizing campaigns and ads that drive true growth.
• Avoid over-investing in retargeting or re-marketing, as these often capture customers who would convert organically.
• Future-proof measurement as browser and OS privacy changes reduce the reliability of channel-based click/view attribution.

Enable the Incremental attribution model

During your campaign creation, open the Attribution Model settings and select Incremental.

❗️

Attention:

Campaigns with the Traffic and Awareness objective do not have the Incremental Attribution model option.

FAQ: Meta Incremental Attribution Model

• What is incremental attribution in Meta (Facebook) Ads?

  Incremental attribution measures the true impact of your Facebook ads by identifying conversions that occur as a direct result of ad exposure—not those that would have happened organically.

• How does incremental attribution differ from traditional attribution models?

  Traditional models credit all conversions within an attribution window, even if users would have converted without seeing your ad. Incremental attribution isolates the conversions genuinely driven by your advertising, providing more accurate reporting.

• Why should I use incremental attribution for my ad campaigns?

  Using incremental attribution helps you optimize your ad spend, improve ROAS, and make smarter budget decisions based on actual campaign effectiveness.

• How can I enable Meta’s incremental attribution in AnyTrack?

  Connect your Facebook Ads account to AnyTrack and activate incremental attribution in the advanced attribution settings of your Ads Manager.

• Are there limitations to incremental attribution?

  While incremental attribution offers better accuracy, its results depend on correct campaign setup and the performance of holdout testing. Data privacy changes may also affect results.