On this pageJump to a section
If you had never run the ad, would that person still have bought?
That one question is the whole idea behind incrementality, and it is the question your ad platform has historically been the worst-positioned party to answer. Standard attribution counts a sale when someone saw or clicked and later converted. It does not ask whether they would have converted anyway. Touched is not the same as caused.
Meta now offers a setting that tries to answer the harder question. It is early, it is imperfect, and it is worth understanding before you decide what to believe from it.
What is Meta's incremental attribution setting?
It is an attribution setting in Ads Manager that estimates which conversions your ads actually caused. According to Meta's description, as relayed by Seer, it compares the conversion behavior of exposed and unexposed users to estimate how many conversions were ad-driven.1
The mechanism, as far as Meta has disclosed it: a machine learning model, trained on controlled lift tests, estimates how likely each conversion was caused by the ad rather than occurring organically. It compares ad-exposed users with similar unexposed users, forming something like a synthetic control group.3 Meta has not published the exact methodology, which means outside validation is not possible.3
It supplements standard attribution
Logical Position's explainer is clear on this point: it is a performance goal that estimates conversions Meta believes would not have happened without ad exposure, and it complements, rather than replaces, standard click and view attribution, which defaults to a 7-day click and 1-day view window.2
Who has access to it?
Not everyone, and the descriptions vary. Logical Position wrote that Meta began rolling it out in 2024, that the rollout is gradual and depends on account eligibility, and that Sales and Leads objectives qualified.2 A separate write-up listed Conversions, Product Catalog Sales and purchase-optimized Sales campaigns.3 Because the details differ between sources, the only reliable answer is your own account.
How to check in Ads Manager
When setting up a campaign with a qualifying objective, open the performance goal section, choose "Show more options" and look for "Incremental Conversions" in the dropdown.2 To see the numbers next to your normal ones, use the Columns menu, choose Custom, then Compare Attribution Settings, and under Advanced Options turn on Incremental Attribution.2
Seer also reported that advertisers can view incremental conversions retroactively from April 1, 2025, by customizing reporting columns.1
What did the biggest public test find?
On July 30, 2025, Seer Interactive published its results from testing the setting across six or more Meta accounts and about $1.05 million in ad spend in April, comparing standard and incremental optimization and cross-checking against GA4.1
- Meta: 87% of conversions were incremental on average, so 13% would have happened regardless.
- GA4: 67% were incremental, so 33% would have happened regardless.
- By targeting: Broad targeting and hyper-narrow segments, such as retargeting, had the lowest incremental performance. Refined mid-funnel audiences showed the clearest impact.1
Seer's interpretation: broad campaigns reach people who would have converted anyway, and retargeting reaches people already near the bottom of the funnel.1
How should you read the gap between Meta and GA4?
As a range with an honest explanation, not a contest.
Two systems looked at the same spend and disagreed by 20 points. Neither is neutral. Meta is measuring itself with a model it built. GA4 sees only what its own tracking can see. Seer's guidance was to treat Meta's self-reported incrementality as directional, cross-reference it with independent sources and assume the truth lies somewhere between.1
A critic's note is also worth keeping in mind: the model is a black box, it sees only Meta's own impact, and Meta benefits from looking impactful, so its definition of incremental may be generous.3
Barry Hott, who says he has advertised on Meta since 2008, was making this argument before the setting existed. In a video he posted on May 2, 2025 walking through the new column, he said:
the system isn't trying to get you incremental sales. The system is just trying to get you as many conversions as it can.
He also sells a course on this, and his example is one account, not a study. I cite him because his reasoning matches the mechanics: a platform paid on delivery has no reason to separate caused sales from coincident ones.
VideoBarry Hott
This 1 NEW Line in Ads Manager Could Save You Thousands
He lines up incremental attribution against one-day view, one-day click and seven-day click, split by new, engaged and existing customers.
What do you do with this on Monday?
Use it to ask better questions, not to produce a headline number.
- Check eligibility. See whether Incremental Conversions appears in your performance goal options and whether the comparison columns are available.
- Compare, do not replace. Add the incremental columns beside your standard ones. Look at the ratio campaign by campaign, not just the total.
- Look at retargeting first (the basics are in Warm Audiences After Cookie Whiplash). If a campaign mostly reaches warm audiences and its incremental share is low, that is the clearest candidate for a budget review.
- Run one holdout test. Withhold ads from a slice of an audience and compare sales. This is the independent check the platform cannot provide for itself.
- Test optimizing for incremental conversions. Run it as a separate campaign against your standard setup with the same budget, and judge by your sales data.
- Write down what you decided and why, so you can revisit it next quarter.
What should you not do?
Do not rebuild your whole account around a modeled number the week it appears. Reported conversion counts will look different, because you are measuring a different thing, and comparing them with last quarter's reports can mislead.
Also do not use incrementality as a reason to abandon retargeting entirely. Retargeting can still be the right way to close sales that are already in motion. The question is whether you are paying to take credit for them.
Over eleven years of buying Meta ads, I have watched the industry's measurement fashion swing from last-click to view-through to modeled everything. The constant is that the party that sells the ads should not be the only party grading them. Meta's willingness to ship a tool that reports lower numbers than standard attribution is, to its credit, a move in the right direction, as is the scoring tool I covered in Opportunity Score.
Meta's own Q2 report showed ad impressions up 11% and the average price per ad up 9%.4 I unpack it in State of Paid Acquisition: Q2 2025. When volume and prices both rise, knowing which conversions you actually caused stops being a nice-to-have.
The best measurement tool is the one that can tell you your favorite campaign isn't working.
If you want your measurement set up to answer that, see the system I build for clients.
Keep building.Ekai
Questions people ask
What is Meta incremental attribution?
It is an Ads Manager setting that aims to count conversions that would not have happened without your ads, estimated with a machine learning model that compares ad-exposed users with similar unexposed users.
How is incremental attribution different from standard attribution?
Standard attribution credits Meta when a person who saw or clicked an ad later converts, using click and view windows. Incremental attribution tries to remove conversions that likely would have happened anyway.
Who can use Meta's incremental attribution setting?
Availability has been gradual and depends on account eligibility. Reports describe it for Sales and Leads style campaigns, so check the performance goal options in your own Ads Manager.
Why do Meta and GA4 disagree on incremental conversions?
They measure different things with different data. In Seer Interactive's test, Meta put incremental conversions at 87% while GA4 put them at 67%. Treat the pair as a range, not a verdict.
Should I optimize campaigns for incremental conversions?
Test it on one campaign and compare against standard optimization and your own sales data. It is a modeled estimate from the platform being measured, so confirm it independently.
Sources
- 1
- 2
- 3
- 4
- 5
