Incremental attribution
- Updated October 31, 2026
Incremental attribution is one of Meta's attribution models. Instead of crediting every conversion that follows a click or view of your ad within a set window, it optimizes delivery for the conversions Meta's models predict the ad caused: the ones that would not have happened without it. You choose it before the campaign is published, and it can't be switched on later.
How does Meta define incremental attribution?
Meta's About incremental attribution page:
Incremental attribution is an attribution model that optimizes ad delivery for incremental conversions. It uses machine learning models that predict whether a conversion is caused by an ad.
The page lists two uses: "Focus campaigns on outcomes that are more likely to be directly driven by an ad" and, hedged, "Potentially drive more incremental conversions compared to your existing campaigns." Meta's October 28 explainer on its ads system describes what the model counts: "conversions that would not have happened without the ad."
What is the difference between standard and incremental attribution?
What each one counts. Meta's Peak Season 2026 guide gives each of Meta's three attribution models one line:
| Model | What it counts | How a conversion gets credit |
|---|---|---|
| Standard | "what the ads touched" | It falls inside the window you chose after a click, a view or an engagement |
| Incremental | "what the ads added" | Meta's models "predict whether a conversion is caused by an ad" |
| Custom | "the way your business already counts it" | You share credit from "your external analytics tool" |
The guide goes on: "Those are three different numbers. If the one you optimise against isn't the one your finance team uses, you're reading a different campaign from the one you're running."
The windows belong to the standard model. Meta's attribution models page: "If you select standard attribution, you can also choose attribution settings, which determine which conversions can be credited to your ads." Those are explained under attribution setting.
Is incremental attribution the same as a lift test?
No. Incremental attribution predicts, for each conversion, whether the ad caused it. A lift test measures the effect instead: Meta keeps a random control group from seeing your ads and compares the two groups. Meta's October 28 tips still send you to a study for proof of cause: "Judge results on cost per result or return on ad spend, and test incrementality. Platform-reported results show what happened; a Conversion Lift or Brand Lift study shows what happened because of your ads."
Can you switch an existing campaign to incremental attribution?
No. Meta's setup page: "You cannot change your attribution model after you publish your campaign." Custom attribution has the same rule in other words: it can be enabled "on a new campaign only. It cannot be added to existing campaigns."
Once both models are running, keep their numbers apart. The attribution models page warns that results "cannot be compared in the Campaign Overview table across ad sets with different attribution models".
What it means for your ads
- Pick the count before you build. The model is fixed once the campaign is published, and Meta's guide calls attribution "the decision most likely to be inherited rather than chosen."
- Read each model against its own history. An incremental ad set's cost per result and ROAS rest on a different count from a standard ad set's.
- Expect conversions that would have happened anyway to drop out. By Meta's definition they aren't incremental, so ads shown mostly to people already on their way to buy can earn less credit here than under a click or view window.
- Treat the model's credit as a prediction. When a budget decision rests on it, a lift test measures the effect, if the campaign meets Meta's minimums for one.
Sources
- About incremental attribution — Meta Business Help Center (accessed October 31, 2026)
- How Meta's Ads System Matches Your Ad to the Right Person at the Right Time — Meta for Business, October 28, 2026 (accessed October 31, 2026)
- Peak Season 2026: You Don't Beat The Auction. You Brief It. — Meta for Business, October 21, 2026 (accessed October 31, 2026)
- About custom attribution — Meta Business Help Center (accessed October 31, 2026)
- About attribution models and attribution settings — Meta Business Help Center (accessed October 24, 2026)
- How to use incremental attribution in Meta Ads Manager — Meta Business Help Center (accessed October 22, 2026)
Meta revises these pages without notice, so check the source before acting on any of it.
Terrads makes new Facebook and Instagram ads from your brand kit, so the campaign you build to try incremental attribution can start with creative of its own.