Meta's Incremental Attribution Explained

Over the past year, Meta has been rolling out updates that go far beyond Andromeda.

One of the biggest — but least understood — is incremental attribution.

Meta knows what advertisers have been feeling for years:

So Meta is doing what Google Ads and other ad platforms will eventually do when signal quality drops:

It builds a data-driven model to fill the gaps.

Incremental attribution is Meta’s attempt to show whether your ad actually caused a conversion — even if tracking can’t prove it directly.

It’s a bold promise. But it also raises a bigger question:

Is this the insight advertisers needed…

or another layer of algorithmic guesswork pretending to be certainty?

Let’s break down what it actually is, how Meta calculates it, and how (and if) you should use it.

Incremental attribution tries to answer one question:

“Did this ad create a conversion that would  have happened without it?”

Because traditional attribution can’t see:

Meta fills in these blind spots with modeled impact, not direct tracking.

Can this machine learning model be trusted blindly?

Absolutely not!

Instead of crediting everything to a click or a view, Meta now uses “modeling” to:

This matters because traditional attribution is getting less reliable every year.

Incrementality is Meta’s attempt to rebuild measurement in a fragmented world.

But always remember:

It’s modeled data — not observed truth.

It’s Meta’s best guess. It’s not proven “data-driven”.

Treat it accordingly.

In other words:

Incremental attribution = modeled lift, not measured event tracking!

This is both powerful… and potentially misleading if misunderstood.

Most advertisers don’t know where Meta hides this setting, because it wasn’t hyped like Andromeda.

It lives inside your reporting columns.

This is for reporting only. It won't change optimization unless you activate it at the campaign level.

Here's how to find it:

Step 1 — Open Columns → Compare Attribution Settings

In Ads Manager, click the Columns dropdown. Scroll down and select Compare attribution settings.

(Or you can click Purchases column and select Compare Attribution Settings in the dropdown)

Step 2 — Enable Incremental Attribution

A pop-up will open showing various attribution windows.

Under Advanced options, toggle on Incremental attribution.

Important notes:

Here’s the truth most marketers won’t say out loud:

Incremental attribution is interesting but…

Use incremental attribution as context, not direction.

It’s helpful for:

It’s not helpful for:

Use it the same way you use blended metrics, ROAS, POAS, CAC, or MER:

As a signal — not a steering wheel.

Meta says Incremental Attribution is powered by “sophisticated models”…

Whatever that means. Thanks, Meta.

⬇️⬇️⬇️

Incremental attribution is smart and useful — but it isn’t a replacement for real conversion data, and it definitely shouldn’t guide your optimization on its own.

If you rely only on incrementality:

The less data you feed Meta, the worse it performs.

So don’t jump into incremental-only decision-making.

Monitor it, compare it, learn from it — but keep your primary optimization on stable, high-volume conversion events.

In other words:

Use incremental attribution as insight, not instruction.

Have you checked out incremental attribution yet?

Did it actually tell you anything new?

Help me write better content for you. Answer these 2 questions:

♥️ Tools we love to use at my agency:

* If you click the links above, we may receive a tiny commission that helps keep Digital Ad Snack running—thank you!