Why TripleWhale or Northbeam Data Doesn’t Match What Facebook Shows in-platform
This question (and answer) were recently posted in the Founders Membership and the answer was too helpful not to share.
To back up a quick second, we know that iOS14.5 was released in 2021 and people updated their iPhones to the newer operating service. Simple, right? However, these changes showed massive implications on advertisers’ ability to market, track, and understand data from potential and current customers due to opt-outs from tracking from iOS (Apple) devices.
Changes to that opted-out user data includes shorter attribution windows, eight standard and custom optimization events, no demographic data in reporting breakdowns, and likely smaller retargeting audiences due to only the highest prioritized events being noted.
What does this mean?
It means that in-platform reporting in Facebook Ads Manager isn’t accurate.
Enter: Third-Party Data Solutions
Two of the most common third-party data solutions are TripleWhale and Northbeam. These softwares use their own custom pixels that are embedded in a website to fully track user data that Facebook once could, and no longer can. They also combine data from multiple advertising platforms to show overlap in user data, such as if they saw and clicked on an ad on Instagram and then TikTok, or saw a TikTok ad and then Googled the company, you would be able to see that data. The attribution window is also 365 days — 51 weeks longer than Facebook’s 7-day (at most) attribution window.
Which brings us to the question:
Why does in-platform Facebook data not match what TripleWhale or Northbeam says?
(Quick shout out to advertiser Barry Hott in the Founders Membership for answering this question. Seriously, join us. This membership is SWEET)
Here’s the thing: using external tracking is going to look very different than what you see in Facebook due to attribution settings and cross-device tracking. As advertisers, we have to let FB optimize based off the data it gets (because in-platform Facebook data is what Facebook’s algorithm and pixel is learning from), but you also should be thoughtful about what data FB is getting/using (7-day click, 1-day view vs. 7-day click only, for example) and how that relates to actual real-life performance outside of Facebook, like in Northbeam or TripleWhale.
With that in mind, the third-party software can’t quite track impression-level tracking (like in-platform Facebook can), only click-through tracking, because their pixel is placed on the brand’s website. These solutions may also have a hard time tracking across devices, like if you click an ad on your phone, then later go to your computer to finish the purchase. (Also, Northbeam offers cross-device tracking and impression modeling, but no tracking will ever be perfect to account for social/viral or other dark actions.)
Because of all of this, the view-through tracking gets hairy due to that impression level discrepancy from what Facebook sees in-platform that a third-party solution cannot, which is why reporting can be incredibly inconsistent in terms of if something is bad on Facebook, it’s also bad on the third-party solution or vice versa. Even if the numbers are not the same, the general consensus of whether an ad/ad set is good or bad will be much more similar without the view-through data.
The TL;DR?
If you’re using a third-party tracking platform and haven’t yet compared a 7DC/1DV and a 7DC or 1DC ad set yet, we strongly recommend you do!
Love this type of stuff?
Need More Info?
Check out the Founders Membership for advertisers where we discuss things like this EVERY DAY with some of the world’s best digital advertisers. We’d love to have you!
If these tracking changes are still confusing, see below some great blog content that was released when the update happened that is still useful to know.