What’s Changed With Facebook Event Tracking Post iOS14.5
Driving yourself crazy trying to unpack all of the changes to event tracking in Facebook post iOS14.5? Learn how to evaluate the accuracy of your tracking.
If you’re a D2C brand then chances are the months of May – July 2021 have been quite stressful trying to get your bearings post-iOS14.5.
Specifically when adapting to the new world of Facebook Reporting in Ads Manager and Events Manager.
Facebook has put out a substantial amount of material, webinars, and warning signals about delays (or decreases) in reporting.
But there isn’t a simple Shopify 1-pager that explains exactly what’s changed, why, and how to compare the new FB reporting to the old.
Plus, there are nuanced areas that our customers rely on that we have to infer the change and subsequent impact.
In this article I’m going to share what we are seeing from our customers, the why behind these changes, and ways that you should consider adapting.
What Changed In Facebook Events Manager
In June, we saw more signups that were looking for this:
Unfortunately Elevar (and no company for that matter) can truly fix this.
It’s the post iOS14.5 reality that many are catching up to.
Evaluating Purchase Events
Let’s break down exactly what the ask above is related to validating events in Facebook.
The scenario might go something like this:
Go to Facebook Events Manager and check the total number of purchase events received in last x days:
540 – got it.
Then you’d go to Shopify to check purchases in time range like this:
WTF? Something must be broken.
540/847 = 63%
“Facebook is not receiving over 1/3 of my orders – this is a big issue!
Even if I remove draft and loop orders, it’s still a big gap.”
Now to do a sanity check, what does Google Analytics say:
Google Analytics is pretty much an exact match to Shopify (shameless plug for Elevar’s server side Google Analytics tracking that gets you 99% match 🙂 ).
“So if GA is correct, but Facebook is not, and both are set up pretty much the same way, then something must be broken with my Facebook pixel.”
If this sounds familiar then you’re in the right place.
Here’s what is happening now.
Facebook’s Events Manager that you’re used to validating data here:
Is now extremely limited in the events that it displays.
By default Facebook is excluding all events from opted-out users, or users that never opted-in, that happen on iOS14.5+ (which they confirm here).
So in theory you can go to your Google Analytics operating system report to view your count of transactions that may be excluded:
In the image above I have 336 transactions that might be excluded since they are all 14.5 iOS or above.
So if we assume 10% of users are opted-in and then add to our Facebook Purchases then we can get our back of napkin total:
336 purchases from iOS14.5+ x .9 = 302 purchases
302 purchases + 540 visible inside Facebook = 842
842 is kind of close enough to total orders in Shopify (847) to make a reasonable conclusion that these indeed are the missing events.
What About Aggregated Events Measurement Tab?
What we’ve seen many do, instead of looking at the GA iOS reports, is something like this:
This doesn’t work either.
Here’s the Facebook description about the AEM tab:
The Aggregated Event Measurement tab in Events Manager counts an event if all the following are true:
- The event is attributed to someone who clicked on an ad and then took an action on your website
- The event is prioritized in your web event configuration and is the highest priority event that occurred during the 24-hour conversion window
- The event is from a person on an iOS 14.5 or later device
The key points here when measuring events are:
- They are attributed purchases ONLY
- They are on 1-2 day delay
Does this make your day starting out feeling like this?
You are not alone feeling your data is everywhere.
Event Matching Quality
This is another common question we receive:
“How can I get my event matching quality up from 3-4 to at least 8/10 for my events?”
This isn’t likely to happen for non-purchase events for majority of brands.
The reason is because most users on your website are guests.
So you don’t have email, phone number, etc that Facebook uses in their point total calculation.
Our general feedback now is:
- If non-purchase events are ~ 4
- Purchase events are ~ 7-8
Then you’re good / shouldn’t stress too much about this rating.
What Elevar Is Doing To Improve Transparency
If you are an Advanced or PRO plan customer using our server side integration for Facebook CAPI and/or Google Analytics then you may have seen this type of data studio report we’ve shared with you:
Essentially what we’re doing is comparing the total purchase conversion events sent to your server side channels to your total Shopify orders in near real time.
We created this type of monitoring to answer the question:
“Can you confirm that Facebook is receiving all my purchase conversions?”
Since there is no longer a way to validate this inside of Facebook, you can now do this with Elevar.
We’re able to do this by monitoring the responses from each API – Facebook, GA, etc.
You’ll begin to see these reports inside your dashboard soon to:
- Quantify the accuracy of data sent to your critical marketing channels
- Alert you when there is a big drop
Updates to our monitoring emails:
Tracking transparency is only going to continue to get more difficult as more privacy laws are enacted which trickle down to platform reporting.
We strongly believe that building monitoring automation around your tracking is going to be critical for the remainder of the year.
If only to save you time that is likely better spent on more strategic analysis or initiatives.
Ads Manager Conversions
If there is one learning lesson here it is this:
The days of monitoring Facebook Ads Manager for near real time performance is gone.
We understand many small brands may not have the cash runway to just “let it ride” for a few weeks/months and see how overall revenue and aggregate ROAS plays out.
But the ability to view same-day purchases attributed to a campaign in Facebook at 2pm and make adjustments just isn’t reliable anymore.
Here are a few data points from our friends at Common Thread Collective:
Here’s the link to view the general methodology that Andrew outlines leveraging Google Analytics data.
How To Learn More
We’ll continue to share more resources as tracking evolves with Facebook and all of the other channels.