Meta Ads Attribution Settings: The Complete 2026 Guide

Meta's attribution system determines which conversions get credited to your ads and which don't. In 2026, it changed twice. If your reported ROAS dropped in January or March without any campaign changes, attribution is almost certainly why.

This guide covers both changes, what the current attribution options actually mean, which settings work for which business types, and how to stop letting the reporting layer confuse your optimization decisions.

What Is Meta Ads Attribution?

Attribution is Meta's answer to the question: did this conversion happen because of this ad? The answer depends entirely on which rules you set.

When a user clicks your ad and buys three days later, does your ad get credit? What if they scrolled past it without clicking, then bought the next day? What if they liked it, and bought the same afternoon? Attribution settings define the rules for all of these scenarios.

Meta tracks four types of user interactions with your ads: link clicks, social interactions (likes, comments, shares, saves), video views of five seconds or more, and passive impressions. Each interaction type can trigger an attribution window, and each window has a different duration. Your attribution setting determines which combinations count.

The reason this matters beyond reporting: Meta's algorithm uses attributed conversions as its optimization signal. Change the attribution window, and you change what Meta's learning phase is optimizing toward. Narrower attribution means fewer conversions feeding the algorithm, which affects delivery quality and CPMs as a downstream effect.

The 2026 Attribution Overhaul: What Changed and When

Two separate changes hit in 2026. Most advertisers conflated them. Understanding each one separately is how you figure out which change affected your account and what to do about it.

January 12, 2026: The View Window Removal

On January 12, 2026, Meta permanently removed two attribution windows from the Ads Insights API: 7-day view and 28-day view. These windows had allowed advertisers to credit conversions to ads that users had only seen, without clicking or interacting.

Accounts using these windows saw reported conversions drop 15 to 40 percent overnight. No campaign changes. No budget cuts. No targeting shifts. The conversions didn't disappear from the business. They disappeared from Meta's reporting.

Remarketing campaigns were hit hardest. View-through attribution had historically inflated how much credit remarketing got for conversions that email, organic, or direct touchpoints actually drove. Removing it exposed that inflation.

March 2026: Click Redefinition and Engage-Through

The second change was bigger and more structural. In March 2026, Meta redefined what counts as a click for attribution purposes.

Previously, every interaction with your ad triggered the 7-day click attribution window. Likes, shares, saves, comments, and link clicks all counted as clicks. If someone liked your ad on Monday and bought your product on Saturday, that counted as a 7-day click conversion.

Now only link clicks count as clicks. Likes, shares, saves, comments, and video views of five seconds or more moved to a new attribution type called engage-through, with a fixed 1-day window.

This matters most for brands with high social engagement. If your ads generate lots of likes and shares, you were previously getting generous attribution credit for that activity. Post-March 2026, that credit is capped at one day.

The Four Attribution Types Explained

Meta's 2026 attribution system has four distinct types. Understanding what each one tracks (and what it doesn't) is the foundation for setting your windows correctly.

1. Click-Through Attribution

Click-through attribution credits conversions that happen after a user clicks a link in your ad. The link click could go to your website, a lead form, or a Messenger conversation. Reactions, shares, and video plays no longer trigger click-through attribution as of March 2026.

You choose between a 1-day window and a 7-day window. The 7-day window is standard for most e-commerce brands because the typical DTC purchase decision cycle runs 1 to 7 days. Switching from 7-day to 1-day typically reduces reported conversion volume by 30 to 40 percent for the same campaign. That's not because your ads got worse. It's because you're excluding genuine consideration purchases that happen in the 2-7 day window after the click.

2. Engage-Through Attribution (New in March 2026)

Engage-through is the newest attribution type. It tracks conversions that happen after a user interacts with your ad through social gestures: likes, comments, shares, saves, or video views of five or more seconds.

The window is fixed at 1-day. If someone liked your ad today and bought something tomorrow, engage-through captures that. If they wait three days, it doesn't.

Engage-through fills a real measurement gap for video-heavy creative. If your ads generate significant social engagement, you now have a dedicated attribution bucket for those interactions rather than having them inflated inside click-through or ignored entirely.

3. View-Through Attribution

View-through attribution credits conversions that happen after a user sees your ad but doesn't click or interact at all. The window is fixed at 1-day.

This is the most contested attribution type. A user could have scrolled past your ad for half a second and bought your product for completely unrelated reasons, and view-through would count it. We use view-through data for brand awareness visibility, not as a primary optimization signal. If you're running performance campaigns, the noise it adds can mislead your ROAS reporting significantly.

4. Incremental Attribution (Advanced)

Incremental attribution uses holdout testing to determine which conversions your ads actually caused rather than just correlated with. Meta runs a controlled experiment, shows some users your ads and withholds them from a holdout group, then measures the conversion rate difference.

This is the most accurate method Meta offers. Seer Interactive's testing across a $1.05 million spend found Meta self-reported 87 percent of conversions as incremental. GA4 cross-reference showed only 67 percent were incremental. That 20-percentage-point gap matters a lot when making budget decisions.

Which Meta Attribution Setting Should You Use?

There is no universal answer, but there is a framework. The right setting depends on your average order value, your typical purchase consideration cycle, and what conversion event you're optimizing toward.

Low-AOV E-Commerce (Under $100)

Use 7-day click + 1-day engage-through + 1-day view. The 7-day window captures the full impulse-to-purchase cycle for most low-ticket products. Engage-through adds signal for the social sharing and video activity that drives DTC brand discovery. View-through gives you passive impression visibility, though treat it as supplementary data rather than primary ROAS.

This is also Meta's default setting as of March 2026. For low-AOV brands, the default generally works well as a starting point.

High-Ticket Products ($200+)

Use 7-day click as your primary window. Consider turning off engage-through if social interactions represent less than 5 percent of your reported conversions. High-ticket purchases require deliberation. A like on your ad rarely translates to a same-day purchase. Including engage-through for high-ticket products adds noise without proportional signal.

The exception: if you're running video-heavy upper-funnel content for a high-AOV brand and seeing meaningful video-driven conversion activity, keep engage-through active and measure it as a separate column in your reports.

Lead Generation

Use 1-day click for lead gen. Social interactions don't drive form completions the way they drive product purchases. Someone who liked your lead magnet ad rarely fills out the form that same day. The 1-day window keeps your optimization signal clean.

Disable engage-through entirely for lead gen. Its 1-day window adds noise and can inflate your reported lead count with interactions that have no meaningful connection to form completion intent.

B2B and Long-Cycle Products

Meta's attribution windows are too short for B2B with a 30-to-90-day sales cycle. Use 7-day click as your Meta attribution setting, but treat it as a partial signal only. Your real attribution for B2B happens in your CRM, not in Ads Manager.

The practical implication: don't optimize B2B Meta campaigns toward purchase events with long consideration cycles. Optimize toward earlier-funnel signals like content engagement or landing page views, which happen closer to the ad impression.

The Attribution Setting That Usually Backfires

We often see accounts that switch to 1-day click to get attribution data that aligns more closely with Shopify. The intent is reasonable. The execution creates a different problem.

When you switch to 1-day click, your reported ROAS drops significantly (typically 30-50 percent for the same campaigns) because you're no longer counting conversions in the 2-7 day window. Meta's algorithm then sees fewer conversions per day and may interpret that as reduced demand, adjusting delivery accordingly.

The result can be lower ROAS in both Ads Manager AND Shopify, because the algorithm's optimization quality degraded. [NEEDS REAL DATA: add verified ROAS and CPA deltas from an actual account where this occurred to replace this placeholder]

If you want Shopify alignment without disrupting delivery, use the Compare Attribution Settings feature (covered below) rather than changing your live attribution window.

How Attribution Settings Affect Campaign Delivery

This is the part most guides skip. Attribution settings don't just change what appears in your reports. They change what Meta's algorithm optimizes toward.

Meta's delivery system works by analyzing the characteristics of users who convert, then finding more people who match that profile. The broader your attribution window, the more conversion events feed that learning process. Narrow the window, and you're training the algorithm on a smaller, more restrictive dataset.

The practical implication: switching from 7-day click to 1-day click mid-campaign isn't just a reporting change. It restarts the learning phase and tells the algorithm to optimize based on a different, smaller set of signals. Meta requires 50 conversion events per week per ad set to exit the learning phase. Accounts that were comfortably above that threshold on 7-day click may fall below it on 1-day click.

The Attribution and Audience Delivery Loop

When you set a 1-day click window, Meta learns from users who buy within 24 hours of clicking an ad. That's a specific type of buyer: high-intent, probably already familiar with your brand or product category. Meta then finds more people like them.

When you run 7-day click, the algorithm learns from a broader set that includes deliberate purchasers who research and return. It finds more of that audience. Neither is inherently better. But they're different audiences, and switching between them mid-flight confuses the learning system and can take two weeks to restabilize.

The Compare Attribution Settings Tool

Meta's Compare Attribution Settings tool is one of the most underused features in Ads Manager. It lets you see how your conversion data looks under different attribution windows simultaneously, without changing anything about your active campaigns.

You can compare 1-day click, 7-day click, and view-through side by side for any campaign, ad set, or ad. The data is purely diagnostic. It doesn't change the attribution setting your campaign is actively optimizing against.

This is the right way to understand your Shopify discrepancy. Instead of changing your live attribution setting and triggering a learning phase reset, you run the comparison report and see exactly where the gaps are.

How to Access Compare Attribution Settings

How to Change Your Meta Attribution Settings

Attribution settings are set at the ad set level, not the campaign level. You can run different attribution windows across ad sets within the same campaign, though we recommend keeping them consistent unless you're running a controlled test.

New Ad Set

When creating a new ad set, you'll see the Attribution Setting section in the Budget and Schedule panel. The 2026 default is 7-day click + 1-day engage-through + 1-day view. You can modify each component independently.

Active Ad Set

To change on a live ad set: navigate to the ad set, click Edit, scroll to Attribution Setting, and make your change. Meta will flag the change if it's likely to affect your learning phase.

The safest time to make attribution changes is at the start of a new budget period, not mid-flight. You get a cleaner before-and-after comparison without the mid-campaign noise contaminating your read.

Meta Attribution vs. Third-Party Tracking Tools

The discrepancy between Meta Ads Manager and Shopify (or Google Analytics, or Triple Whale) is not a bug. It's a feature of how attribution works differently across systems.

Meta uses person-level data tied to Facebook and Instagram account IDs. It tracks across devices and sessions within the attribution window. Shopify counts the last-click session source. GA4 defaults to data-driven attribution that distributes credit across all touchpoints. These three systems will almost never agree, and that's expected.

Why the Numbers Diverge

Cross-device attribution: A user sees your Meta ad on mobile, then buys on desktop through a Google search. Meta counts it as a Meta conversion (they clicked the ad). GA4 counts it as Google organic. Shopify counts it as direct.

View-through inflation: Meta can credit a conversion to an ad the user saw but never clicked. Shopify never does this. If view-through is on, expect Meta's numbers to run higher than Shopify's.

Pixel signal loss: iOS privacy changes and cookie restrictions mean Meta's pixel doesn't capture every conversion. Pixel-only tracking captures roughly 40 to 60 percent of actual conversions. CAPI (Conversions API) improves this by 25 to 40 percent by matching server-side purchase data to Meta user profiles via email and phone number hashing.

The Right Way to Think About Discrepancies

Meta's numbers tell you about Meta's contribution within its attribution rules. Shopify's numbers tell you about the last session. Neither is complete on its own.

When we audit attribution for accounts, the first thing we look for is the direction of the discrepancy. If Meta is reporting significantly higher conversions than Shopify, view-through or engage-through inflation is usually why. If Meta is reporting lower, signal loss from iOS and cookie blocking is the likely cause, and CAPI is the fix.

Switching from combined attribution (7-day click + 1-day view) to 7-day click only typically reduces the Meta vs. Shopify discrepancy. But it comes with a corresponding drop in Meta's reported ROAS because the in-platform numbers get smaller. You're not actually performing worse. You're just measuring more conservatively. [NEEDS REAL DATA: add verified delta percentages from a real account that made this switch]

Attribution Recovery Playbook

If you saw your Meta reported conversions drop in January or March 2026 without any campaign changes, here's the five-step process we use to diagnose and recover.

Meta Attribution in 2026: The Bottom Line

Two significant changes in three months reshuffled how Meta counts conversions. Most advertisers either didn't notice or didn't update their benchmarks and reporting accordingly.

The practical priorities: establish a new performance baseline using post-change data, implement CAPI if you haven't yet, and use the Compare Attribution Settings tool before making any window changes. Don't let a reporting drop drive budget cuts before you've confirmed whether it's a real performance issue or an attribution reclassification.

Attribution is one of the first things we audit when we take on a new Meta account. If your numbers feel murky and you want a clear read on where your real discrepancies are and what to fix first, that's exactly the kind of work we do.