What You Don't Know About Meta Attribution

In two prior episodes, I covered view-through and click attribution. Today, let’s talk about some of the most confusing, if not controversial, aspects of how Meta credits ads for conversions.

Some of these may seem obvious to experienced advertisers, but I guarantee that a few will be completely new information for most listeners.

1. Attribution Counts More Than Immediate Conversions

This might sound obvious, but it is worth repeating: attribution counts more than just conversions that happen immediately after a click.

When we imagine a typical purchase, we picture someone clicking an ad, going to the website, and converting right away. But real customer journeys are rarely that clean.

The default attribution setting is 7-day click and 1-day view, which opens the door to all kinds of paths to conversion. Someone can convert without clicking your ad, as long as it happens within a day of viewing it. And that conversion could happen across devices -- Meta can still credit your ad if the user is logged into Meta platforms on both devices.

This is one of the main reasons why Ads Manager results will rarely match third-party reporting. If you’re relying on URL parameters for attribution, they may be useless in these cases.

2. Results Don’t Always Reflect the Thing You Promoted

This is a commonly overlooked detail. Let’s say you run an ad promoting a baseball cap. A user clicks the ad and visits your site but does not like the hat. Instead, they browse and buy a t-shirt.

If your campaign is optimized for purchase conversions, that t-shirt purchase will be counted in your results.

So if Ads Manager shows 10 purchases from your hat ad but you only sold 8 hats, Meta is not inventing results. Some users simply bought something else.

3. A Single Person Can Trigger Multiple Conversion Events

Meta will de-duplicate conversion events when they happen at the same time and appear to be the same transaction. But beyond that, a single person can still be counted for multiple conversions.

For example, someone clicks your ad, makes a purchase, and then returns four days later to make another purchase. If both events happen within the 7-day click window, Meta will count both conversions.

You can isolate the first conversion using the Compare Attribution Settings feature and selecting First Attribution. This is one of my favorite tools because it often helps align Ads Manager data with backend reporting.

4. Click Attribution Includes Non-Link Clicks

This came up in the last episode, but it’s worth repeating. For more than a decade, I didn’t realize this, and I’m not alone.

Click attribution includes any kind of click. This could be:

• Opening a video

• Reacting to a post

• Leaving a comment

• Clicking the image

Let’s say someone already knows your brand. Maybe they’re on your email list, follow you on social, or visit your site regularly. Meta prioritizes them, and they see your ad.

They click, but not on a link. Maybe they like the post. Then, a few days later, they make a purchase, possibly unrelated to the ad they saw. Meta will still give your ad credit for that purchase.

The argument can be made that your ad contributed, just like with view-through conversions. But here’s the difference: view-through attribution allows only one day for the conversion. These low-intent clicks get seven days.

What’s frustrating is that you can’t separate link clicks from non-link clicks in Ads Manager reporting. There’s no way to tell which click led to the conversion. And you can’t verify that contribution through third-party tools either.

Attribution Is Messy

All of these scenarios create reporting challenges, especially if you or your client expect Ads Manager data to match third-party tools exactly. It won’t.

If you insist on perfect alignment, you’ll likely assume Meta is inflating your results. But that’s usually not the case. You might be able to confirm that a transaction happened, but you’ll have to trust Meta that your ad was part of that journey.

Maybe you already knew about these attribution quirks. If so, you’re ahead of most advertisers. But it’s still a valuable reminder:

The customer journey is rarely as clean as we want it to be.