Meta’s Attribution Overhaul Is Hitting Remarketing Hardest: What Ecommerce Brands Must Adjust

If your remarketing campaigns on Meta suddenly look like they fell off a cliff — congratulations, you’re reading the data correctly. You just don’t know what you’re looking at yet.

We covered Meta’s new attribution model when the changes were first announced. Click-through narrowed to link clicks only. Engage-through introduced as a new category. Video engagement threshold halved.

But now that the rollout is live and the data is flowing, one thing is becoming very clear: remarketing campaigns are absorbing the biggest hit. And it’s not even close.

Here’s why — and what to do about it.

Why Remarketing Gets Disproportionately Hammered

To understand why retargeting campaigns are taking the brunt of this change, you need to understand how remarketing audiences actually behave on social platforms.

Remarketing Audiences Engage Differently Than Cold Traffic

When someone already knows your brand — they’ve visited your site, added to cart, browsed your product pages — they interact with your ads differently than a stranger seeing you for the first time.

Cold traffic clicks links. They have to. They don’t know who you are, so if your ad hooks them, their next step is clicking through to your site to learn more.

Warm audiences? They save. They like. They share with a friend who also shops your brand. They see your retargeting ad, think “oh yeah, I need to go back and buy that,” and then open a new tab, type your URL directly, and purchase.

Under the old attribution model, that purchase got counted as a click-through conversion on your remarketing campaign — even though the “click” was just a like or a save. Your remarketing ROAS looked incredible because it was getting credit for high-intent purchases from people who were going to buy anyway, with social interactions masquerading as clicks.

Now that facade is gone.

With click-through limited to actual link clicks, remarketing campaigns lose a massive chunk of their attributed conversions. Those conversions shift into engage-through — or worse, they fall into view-through territory where they may not even appear in your default reporting columns.

The Numbers Tell the Story

Across the ecommerce accounts we manage, here’s the pattern we’re seeing after the attribution rollout:

Read that last row again. Actual revenue hasn’t changed. The products are still selling. The customers are still buying. But the remarketing dashboard looks like someone pulled the plug.

This is a reporting problem, not a performance problem. But reporting problems become performance problems fast when people make bad decisions based on bad reads.

The Three Mistakes Ecommerce Brands Are Making Right Now

Mistake #1: Gutting Remarketing Budgets

This is the most common — and most expensive — mistake we’re seeing.

Brand sees remarketing ROAS drop from 8x to 4x overnight. CFO asks what happened. Media buyer panics and slashes the remarketing budget by 50%. Revenue drops two weeks later.

What actually happened: the same campaigns are driving the same revenue. The attribution label changed. Cutting budget based on a label change is like firing your best salesperson because someone renamed their commission category in the accounting software.

Mistake #2: Shifting Budget From Retargeting to Prospecting

The second-order version of Mistake #1. Remarketing “looks worse,” so brands reallocate to prospecting where the numbers look relatively stable.

The problem: remarketing exists for a reason. Those warm audiences convert at 3-5x the rate of cold traffic. They’re the most efficient dollar you spend. Moving money away from them because of a reporting shift is like closing your best-performing retail location because you changed how you count foot traffic.

Mistake #3: Adding Engage-Through to Click-Through and Calling It a Day

Some media buyers are simply combining click-through + engage-through into one “total” column and going back to business as usual. This defeats the entire purpose of the split.

The whole point is that these represent different behaviors and different intent levels. A link click from a retargeting ad is a high-intent signal. A save is meaningful but softer. Combining them erases the insight Meta is now giving you.

The Remarketing Attribution Playbook: 5 Adjustments to Make Now

1. Rebuild Your Remarketing Reporting View

Your old remarketing report is lying to you — or rather, telling you a story that no longer matches reality. Here’s the new view you need:

Columns to add in Ads Manager:

• Click-through conversions (purchases)
• Engage-through conversions (purchases)
• View-through conversions (purchases)
• Total conversions (all attribution types)
• CPA by attribution type
• ROAS by attribution type

Now look at your remarketing campaigns through this lens. You’ll likely see that the “lost” click-through conversions didn’t disappear — they moved to engage-through. Your total should be roughly the same. The composition just shifted.

2. Create Weighted ROAS Metrics

Not all conversions carry the same weight. A click-through conversion on a remarketing ad represents a direct, high-intent action. An engage-through conversion represents a softer signal that still correlates with purchase behavior.

Here’s a weighting framework we’re using with our clients:

Weighted ROAS formula:

(Click-through revenue × 1.0) + (Engage-through revenue × 0.6) + (View-through revenue × 0.25) ÷ Total spend

This gives you a single number that respects the hierarchy of intent while still crediting softer interactions. It’s not perfect — no single metric is — but it’s dramatically better than either ignoring engage-through or treating it as equal to click-through.

3. Segment Remarketing by Audience Temperature

The attribution shift doesn’t affect all remarketing audiences equally. Here’s what we’re seeing:

Hot audiences (cart abandoners, 1-3 day website visitors): These still generate strong click-through numbers. These people are ready to buy — they click links, they complete purchases. Attribution shift is minimal here.

Warm audiences (7-30 day visitors, past purchasers, engagers): This is where the biggest shift happens. These audiences interact with your ads socially — saves, likes, shares — and then purchase through direct or organic channels. Their click-through numbers dropped hard, but engage-through picked most of it up.

Tepid audiences (30-90 day visitors, broad engagers): The engage-through shift is less significant here because these audiences were already contributing mostly view-through conversions. Not much changed.

The implication: your warm remarketing audiences are the ones that need the most reporting adjustment. Don’t treat all retargeting the same way.

4. Use Incrementality Testing to Validate Remarketing Value

Here’s the move that separates sophisticated ecommerce operators from everyone else: stop arguing about attribution and measure incrementality instead.

Attribution tells you which ad touched the customer before they bought. Incrementality tells you which ad caused the purchase.

How to run a basic incrementality test on remarketing:

1. Pick one remarketing audience segment (e.g., 7-14 day site visitors)
2. Run a holdout test — show ads to 80% of the audience, hold back 20%
3. After 2-3 weeks, compare purchase rates between the exposed and holdout groups
4. The difference is your incremental lift — the revenue your remarketing actually caused

Meta’s native Conversion Lift tool can automate this. If you’re spending $5K+/month on remarketing, there’s no excuse not to run incrementality tests. They’ll tell you more about the true value of your retargeting than any attribution model ever will — especially now that the attribution models are in flux.

5. Recalibrate Your Automation Rules

This is the silent killer. If you’re running automated rules in Ads Manager — rules that pause ads below a certain ROAS, increase budgets above a certain CPA, or adjust bids based on conversion volume — those rules are now using different inputs.

A rule that says “pause any ad set with ROAS below 3x” may now be killing your best remarketing campaigns because they dropped from a reported 7x to a reported 3.2x overnight. The campaigns are performing identically. The rule just doesn’t know that.

Action items:

• Audit every automated rule in your account
• Identify which rules reference click-through metrics
• Either pause those rules temporarily or adjust thresholds to account for the attribution shift
• Consider rebuilding rules to use total conversions (click + engage + view) as the trigger instead of click-through alone

The Bigger Strategic Question: Is Remarketing Still Worth It?

Let’s zoom out. Beyond the attribution mechanics, this change is surfacing a question that smart ecommerce operators have been quietly asking: how much of remarketing’s reported value was ever real?

Remarketing has always had an attribution problem. The people in your retargeting audiences are — by definition — your warmest prospects. Many of them would have purchased whether they saw your retargeting ad or not. Every attribution model, including Meta’s old one, overstated remarketing’s contribution because it couldn’t separate “the ad caused the purchase” from “the ad was seen before a purchase that was going to happen anyway.”

The new attribution split makes this visible. When you see that 55% of your remarketing conversions are now engage-through (driven by saves and likes rather than link clicks), it’s worth asking: how many of those were purchases your ads influenced versus purchases your ads happened to be near?

The answer is almost certainly “both.” Remarketing absolutely drives incremental revenue. The brand reinforcement, the reminder, the social proof of seeing your ad repeatedly — those things work. But they probably don’t work as well as the old 8x ROAS suggested.

The honest answer is probably a 3-5x weighted ROAS once you account for the attribution shift and natural purchasing behavior. Which is still excellent. Just not the fairy tale number some brands were reporting.

How This Changes Full-Funnel Strategy

If remarketing’s reported contribution drops (even if real contribution stays the same), it changes how your full-funnel metrics look on paper.

Suddenly, the blended ROAS across prospecting + remarketing looks lower. The CFO or the founder asks why overall Meta performance is declining. You have to explain that nothing changed except the counting method — and that’s a harder sell than it sounds.

Here’s how to frame it:

Before attribution change:

• Prospecting: 2.5x ROAS
• Remarketing: 8x ROAS
• Blended: 4.2x ROAS

After attribution change (same performance):

• Prospecting: 2.1x ROAS (slight click-through drop)
• Remarketing: 4x ROAS (major click-through drop)
• Blended: 3.0x ROAS

Reality:

• Prospecting: 2.5x weighted ROAS (add engage-through credit)
• Remarketing: 5.5x weighted ROAS (add engage-through credit, weighted)
• Blended: 3.8x weighted ROAS

The weighted view is closest to truth. Present all three to stakeholders so they can see the gap between raw click-through (understated), old reporting (overstated), and weighted reality (best estimate).

The Bottom Line

Meta’s attribution overhaul is a net positive for the industry. Cleaner data. Better signal separation. More honest reporting.

But it’s hitting remarketing disproportionately hard because retargeting audiences were the biggest beneficiaries of the old, inflated click-through counting. The brands that come out ahead are the ones who:

1. Don’t panic — revenue hasn’t changed, labels have
2. Rebuild reporting — separate click-through, engage-through, and view-through into distinct columns
3. Weight conversions by intent — not all attributed revenue carries the same confidence level
4. Test incrementality — stop relying on attribution alone to value remarketing
5. Fix automation rules — your rules are using inputs that just changed

The worst thing you can do is cut remarketing budget because the dashboard looks different. The second worst thing is pretend nothing changed and keep reporting the old way.

The answer — as it usually is — is somewhere in the middle. Read the data. Weight it honestly. And keep doing what’s actually driving revenue.

Need help adjusting your remarketing strategy to Meta’s new attribution model? Book a strategy session — we’ll walk through your specific numbers and build a reporting framework that tells the truth.