Meta’s Engage-Through Window Is Toggleable — Most WooCommerce Stores Don’t Know

What Meta Actually Changed on March 3#

Meta narrowed the definition of a click — and six days of attribution credit vanished from one column overnight.

On March 3, 2026, Meta announced a structural overhaul of its attribution framework. Click-through attribution now counts only link clicks — clicks that send a user to a website, app, lead form, or other destination. Every other interaction that previously counted as a click — likes, saves, shares, comments, video views, profile taps — was moved into a new category called engage-through attribution (Search Engine Land, 2026).

For WooCommerce stores running Meta Ads, reported click-through conversions dropped 15–40% overnight (Dataslayer.ai, 2026). Nothing changed about ad spend, targeting, or creative. Meta changed how it classifies the conversion, not whether the conversion happened.

This is the second attribution reset in three months. On January 12, 2026, Meta removed the 7-day view and 28-day view attribution windows from its Ads Insights API. The March change is mechanically different — it redefines what counts as a click, not how long the attribution window lasts — but the cumulative effect on WooCommerce dashboards is the same: numbers that looked stable in Q4 2025 now look dramatically worse.

Meta’s March 2026 click redefinition moved likes, saves, shares and comments out of click-through attribution and into a new engage-through bucket with a fixed 1-day conversion window — reported click-through conversions dropped 15–40% across WooCommerce stores overnight.

The diagnostic question is the same one Seresa has been asking since January: did your WooCommerce revenue drop by the same percentage Meta is reporting? If not, your ads are fine. Your reporting surface changed.

What Engage-Through Attribution Is and Where Your Conversions Went#

The conversions didn’t disappear — they moved into a new column with a shorter window.

Engage-through attribution replaces what Meta previously called engaged-view attribution, but with expanded scope. The old engaged-view only covered video views of 10 seconds or more. The new engage-through captures every non-link interaction: likes, saves, shares, comments, video views of 5 seconds or more, ad expansions, and swipes (Jon Loomer Digital, 2026).

The video engaged-view threshold also dropped from 10 seconds to 5 seconds in the same update. For videos shorter than 5 seconds, the threshold is 97% of the total video length (Search Engine Land, 2026). That means a 4-second Reel only needs 3.88 seconds of viewing to qualify.

Here’s the thing: the conversions that used to be counted under click-through because someone liked your ad and then purchased three days later still happened. The customer still bought. Your WooCommerce order still exists. But that conversion no longer sits in your click-through column. It sits in engage-through — with a 1-day window instead of the 7-day window it previously rode on.

Conversions that happened between day 2 and day 7 after a non-link interaction now fall outside engage-through’s 1-day window entirely. They don’t appear in any attribution column. They vanish from Meta’s reporting surface.

For WooCommerce stores with consideration cycles longer than 24 hours — which is most stores selling anything above impulse-buy pricing — this is where the real loss hides.

You may be interested in: Meta Killed Its 28-Day View Attribution Window on January 12, 2026

The Window Math That Eats Remarketing#

Remarketing campaigns are hit hardest because social engagement drives delayed conversions — and the new window cuts off at 24 hours.

Remarketing relies on familiarity. Someone who already knows your brand sees your ad, engages with it — maybe saves it, maybe watches the video — and purchases a few days later. Under the old attribution model, that purchase was credited to the campaign under click-through attribution, because Meta counted the save or video view as a “click.”

Under the new model, that same save triggers a 1-day engage-through window. If the customer purchases on day 3, the conversion disappears from attribution entirely (Dataslayer.ai, 2026).

The impact on reported ROAS is direct. Fewer conversions credited to the campaign means lower attributed revenue, which drags ROAS down. The reflex response — cutting remarketing budget — is exactly the wrong move. The campaigns are still working. The reporting surface can no longer see it.

Translation: your remarketing is probably performing the same as it was in February 2026. The dashboard just can’t prove it anymore.

Pixel-only WooCommerce tracking sees attribution accuracy drop to around 40% — meaning more than half of actual conversions never reach Meta at all (wetracked.io, 2026). Adding CAPI narrows that gap, with CAPI-enabled advertisers seeing 17.8% lower cost per result than pixel-only setups (Meta for Business, 2026).

The numbers tell a clear story. Every percentage point of attribution accuracy lost to the engage-through reclassification compounds the existing accuracy losses from iOS privacy changes, ad blockers, and consent banners. For a WooCommerce store already losing 30–40% of pixel events to browser-side interference, the engage-through split removes another layer of visibility from the remaining signal.

The Setting Most Stores Don’t Know Is Toggleable#

Engage-through isn’t a fixed constraint — it’s a campaign setting you can control.

Most WooCommerce store owners running Meta Ads don’t know that engage-through attribution is a toggleable setting. It ships as “on” with a 1-day window by default. But in your campaign’s attribution settings, you can turn it off by selecting “none” (Jon Loomer Digital, 2026).

Turning it off doesn’t change your ad delivery or billing — it changes what shows up in your reporting. Conversions that would have appeared under engage-through simply stop appearing. The algorithm still sees them internally for optimisation purposes, but your dashboard and export files won’t include them.

There are three paths:

Leave engage-through on (1-day default). You see engagement-driven conversions that happen within 24 hours. You lose visibility on anything after that. This is where most stores sit today because they don’t know the setting exists.

Turn engage-through off (“none”). Your click-through column shows only link-click conversions. Cleaner, simpler, but you lose all visibility on how social engagement contributes to conversions. Useful if you report MER (marketing efficiency ratio) or blended ROAS from your WooCommerce backend instead of relying on Ads Manager.

Keep engage-through on and build server-side deduplication. This is the path that preserves the most signal. Your pixel fires from the browser. Your server fires the same conversion to Meta CAPI from woocommerce_payment_complete, using the WooCommerce order ID as the event_id. Meta deduplicates against the server signal rather than relying solely on the browser signal — and the engage-through window captures the server event when the browser event is blocked.

The third option requires a CAPI implementation that most WooCommerce stores don’t have. But it’s the only option that doesn’t trade signal for simplicity.

What CAPI Should Optimise Toward Now#

The engage-through split makes CAPI deduplication more valuable, not less — because the 1-day window amplifies every missed event.

Before March 3, a WooCommerce store running CAPI had a comfortable 7-day window for non-link interactions to convert and still receive attribution credit. A save on Monday followed by a purchase on Thursday was captured. After March 3, that same sequence is invisible to the engage-through window — unless the server-side CAPI event fires within 24 hours of the engagement and carries a matching event_id for deduplication.

The mechanic is specific. Your pixel fires from the browser when the customer hits the WooCommerce thank-you page. Your server fires the same conversion to Meta CAPI from woocommerce_payment_complete, using the WooCommerce order ID as the event_id. When both events carry the same event_id, Meta deduplicates them and credits the conversion to the correct attribution bucket.

CAPI-enabled advertisers already see 17.8% lower cost per result than pixel-only setups (Meta for Business, 2026). That advantage widens in a world where the engage-through window shrinks the reporting surface for browser-only events.

What changes for CAPI optimisation after the engage-through split:

Event timing matters more. With a 1-day engage-through window, CAPI events need to fire as close to the conversion as possible. Delayed server events — from batch processing or nightly cron jobs — risk falling outside the 24-hour window entirely.

Event_id deduplication becomes non-optional. Without it, Meta counts both the pixel event and the CAPI event as separate conversions, inflating your numbers in one direction while the engage-through change deflates them in another. The noise compounds.

Purchase is the signal to optimise toward. The engage-through change affects how conversions are classified, not which events the algorithm uses for optimisation. Sending clean, deduplicated purchase events via CAPI gives the algorithm the highest-quality signal regardless of which attribution bucket the dashboard credits.

You may be interested in: Meta’s One-Click CAPI Just Went Live — Don’t Click It If You Already Have One

How This Changes Advantage+ Shopping Campaigns#

Advantage+ is partially insulated — but not immune — because it optimises on purchase signals rather than click classification.

Advantage+ Shopping campaigns optimise directly on purchase events, which partially insulates them from attribution reclassification (Growth Hackers, 2026). The algorithm doesn’t care whether a conversion is classified as click-through or engage-through — it cares that the conversion happened and that the signal is clean.

Where the engage-through change matters for Advantage+ is in reporting and budget decisions. If your Advantage+ campaign shows lower click-through ROAS because non-link interactions moved to a different column, the human decision-maker might reduce budget based on the wrong number. The algorithm didn’t get worse. The reporting got more granular.

For WooCommerce stores running Advantage+ Shopping, the engagement-driven conversion path — save the ad, browse later, purchase — is a real behaviour pattern. Cutting budget because the click-through column shrank means cutting budget on conversions that still happen but are now invisible to the default view.

The fix is the same: report blended ROAS from your WooCommerce backend alongside Meta’s attributed ROAS. When the two numbers diverge, the WooCommerce number is the one that matches your bank account. Social media has overtaken search as the world’s largest ad channel (WARC, 2026). Attribution models built for search-era click behaviour are being retrofitted for social-era engagement behaviour — and the engage-through bucket is Meta’s attempt to bridge that gap.

Attribution Type	What Triggers It	Window	Status After March 2026
Click-through	Link clicks (website, app, lead form)	7-day (default)	Narrowed — link clicks only
Engage-through	Likes, saves, shares, comments, video views (5s+), ad expand, swipe	1-day (fixed)	New — replaces engaged-view, expanded scope
View-through	Ad impression without interaction	1-day (default)	Unchanged
Incremental	Causal lift above organic baseline	Varies by test design	Available as alternative model

Key Takeaways#

• Reported click-through conversions dropped 15–40%: Meta moved non-link interactions (likes, saves, shares) out of click-through attribution and into a 1-day engage-through bucket — a reclassification, not a performance decline.
• Engage-through is toggleable: You can turn it off in campaign attribution settings. Most WooCommerce stores don’t know this setting exists and are running on the default 1-day window.
• Remarketing is hit hardest: Social engagement drives delayed purchases. The 1-day engage-through window means conversions after day 1 vanish from attribution entirely — remarketing campaigns lose credit for conversions that still happen.
• CAPI deduplication becomes critical: Server-side events with matching event_ids let Meta attribute conversions even when the browser signal is blocked or the engage-through window is too short. CAPI setups cost 17.8% less per result than pixel-only.
• Compare Meta ROAS against WooCommerce revenue: When the dashboard number diverges from your actual order revenue, the WooCommerce number is truth. Don’t cut budget based on a reporting reclassification.

References#

• Jon Loomer Digital — How Meta Ads Attribution Works in 2026 (March 2026)
• Dataslayer.ai — Meta Attribution Change 2026: What Engage-Through Attribution Means (April 2026)
• Search Engine Land — Meta Introduces Click and Engage-Through Attribution Updates (March 2026)
• Media Performance — Meta Engage-Through Attribution Explained (March 2026)
• Growth Hackers — Meta’s Click Attribution Update 2026: Impact by Campaign Objective (March 2026)
• wetracked.io — Attribution Accuracy Benchmarks for E-commerce (2026)
• Meta for Business — CAPI Performance Benchmarks via PPC Land (2026)
• WARC — Social Media Overtakes Search as Largest Ad Channel (2026)
• DOJO AI — Meta Ads Attribution in 2026: What Changed, Why It Matters (March 2026)

If your WooCommerce store is making budget decisions from post-March dashboards without comparing Meta-attributed revenue against actual order data, you’re optimising from the wrong number. Seresa builds first-party server-side pipelines that capture the complete conversion picture — from pixel and CAPI deduplication to BigQuery attribution truth tables — so the algorithm and your spreadsheet agree on what’s real.