Hey {{ First Name | Founder }},
Your ROAS is up. Your agency is celebrating.
But your Shopify revenue hasn’t changed.
Meta has changed how it attributes, here's the low-down.
Meta has overhauled how it counts conversions. Multiple times. In the past year.
Let me break it down because this stuff genuinely impacts your budget decisions.
What changed #1: "Engaged-View" attribution is now ON by default and the threshold dropped
Previously, Meta introduced engaged-view attribution as an optional setting - it would count a conversion if someone watched at least 10 seconds of your video ad and then purchased within 24 hours (without ever clicking). Engaged-view is available for all placements except Facebook in-stream video ads that cannot be skipped.
In late 2025, two things changed: Engaged-view attribution is now on by default previously it was off and had to be enabled manually.
The engaged-view threshold for video dropped from 10 seconds to 5 seconds. So Meta now counts a conversion if someone watches just 5 seconds of your video, then buys within 24 hours even without ever clicking your ad.
Why does this matter? Because it means your conversion count likely went up not because performance improved, but because the definition of a conversion got wider. If you haven't noticed, you might be celebrating results that are partly inflated.
What changed #2: "Click-through" now means something different
This one is huge. Historically, Meta counted all click types toward click-through conversions likes, shares, saves, and link clicks. Most third-party tools like Google Analytics only counted actual link clicks to a website. This discrepancy is why your Meta Ads Manager numbers never matched your GA4 or Triple Whale numbers.
Meta has now fixed this going forward, only link clicks count toward click-through attribution. Conversions previously attributed to non-link interactions (likes, saves, shares) now fall under a renamed category: "engage-through attribution" (formerly known as engaged-view attribution).
So if your reported click-through conversions suddenly dropped and you're panicking, this might be why. Your ads didn't stop working. The measuring tape just got more accurate.
What changed #3: The new category Incremental Attribution
Meta introduced Incremental Attribution in 2025. This feature isolates conversions that would not have happened without your ad helping you separate correlation from causation.
This is the thing I have been banging on about with geo hold-out tests for years and now Meta is starting to build it natively into the platform. This could be useful.
So what should you actually do?
Step 1: Check your current attribution setting
You'll find attribution settings at the ad set level in Meta Ads Manager.
Open your ad set → scroll down to Cost per Result Goal → click Show More Options.
Your current default is almost certainly 7-day click + 1-day view + 1-day engaged-view.
Step 2: Audit what's inflating your numbers
Use Meta's Compare Attribution Settings feature inside Ads Manager, along with the Breakdown by Attribution Setting option in custom reports. These tools let you see conversion data segmented by click-through, engage-through, and view-through so you can understand the composition of your results.
To find it: go to Ads Manager → click the Columns dropdown → select Compare Attribution Settings at the bottom.
Step 3: Apply the right settings for your account
Here's my take:
Generally default settings are fine, just know what they are comprised of.
If you have a lot of conversions coming from non-click conversions, you might want to consider testing a different attribution setting and seeing how it impacts your CAC, your total site revenue and the overall performance of your campaigns.
Make sure any test or change you make, you do in a clean environment - not during a new product launch, sale, or just after you reskinned your website!
The bottom line
Meta keeps expanding what counts as a conversion. That's not always in your interest as an advertiser. You now have three ways to look at Meta performance in-platform: click-through, view-through, and engaged-view.
While additional data can be useful, you must question whether reported improvements represent real business impact.
Your job is to understand what you're actually measuring and hold your agency accountable for knowing the difference.
Jessie
PS. This is exactly the type of insight I drop inside the Ecomm Rockets community to help founders stay ahead. If you'd like more info, just reply “community” or book a quick call and we can chat: https://calendly.com/jessiehealy/coffee_with_jessie
