Google Analytics and Facebook Ads: Why is the data so different?
On many forums and online spaces people are talking about the enormous differences between Facebook Ads click data and the sessions attributed to these campaigns in Google Analytics (via UTM tagging).
Let’s examine this issue a piece at a time. Depending on the type of Facebook Ads campaign (and the corresponding objective), the main causes of discrepancies between Facebook click data and Google Analytics sessions are the following:
CLICKS TO WEBSITE-type FB Ads campaigns
1.Errors when entering UTM parameters in FB Ads (parameters ignored -> everything ends up in References).
There are two options for using UTM link tagging in FB Ads:
1.1. Include the tagged link with all its parameters in the Target Website field (as shown above).
1.2. Use the URL Tags field (Advanced Options; can be pulled down at the end). IMPORTANT: You must enter the chain of parameters without a question mark (?) at the start. For example: utm_medium=AAA&utm_source=BBB&utm_campaign=CCC…
Note: These parameters are added to those indicated above if the initial URL also has them (e.g. parameters for the website itself, such as ?search=XYZ, ?orderby=price, etc.). URL tag values are also given priority. If a parameter is repeated in both fields, the value entered beneath will have priority.
2. Not looking at the right click statistic in FB Ads:
Here’s the best answer to this question posed on Quora:
“This perhaps on thing I hate about Facebook Ads. In Facebook terms’ a “click” a confusing term and can refer to any of the below:
- Expanding the description
- Clicking to read the comments
- Post Likes
- Post comments
- Post shares
- Page Likes
- And of course, clicking on your ads
For current FB Ads reports, the metric we should take into consideration (i.e. the one that’s most similar) is Link Clicks, although we should bear in mind that as well as clicks on ads, it also includes the clicks made on the FB page (in other words, all of the clicks towards the target website).
With this in mind, it would be strange if Analytics and FB data actually did match!
3. FB subdomains such as m.facebook.com and l.facebook.com sometimes delete the parameters (redirect, iFrame, etc.). We’ve even come across cases in which tagging has been lost in clicks from the top-level domain facebook.com:
To test this, Aukera conducted a small experiment. We created a parameter to distinguish the URL of the landing page for FB Ads from that of the normal URL.
The outcome: we saw that for certain clicks on FB, sessions would go to the referral medium in Google Analytics (albeit a minority of around 1%).
This all suggests that the parameters are lost (possibly due to a redirect), because otherwise, if the ad’s URL was copied and shared, it would still have the tag, right? Meaning it would go to Facebook / Display, not to referral.
4. The common technical causes and explanations for the differences between clicks and sessions.
APP INSTALL-type FB Ads campaigns
Unfortunately, even if you tag Play Store / iTunes links via URL Builder, FB will ignore the parameters and all traffic will be sent to Google Analytics as “direct”.
Notes: According to FB’s own support, links to apps that are located on facebook.com are the only ones that can be tagged via the URL Tags field. Google Analytics’ link-tagging tools for each ecosystem (URL Builder for iTunes and URL Builder for Play Store) are of no use in this instance. In fact, the link generated by the Google Analytics link creator for iTunes isn’t even accepted by FB Ads:
OFF TOPIC: Above all, we should use URL Builders to link from our website (or other websites) to our apps in the app stores. If we don’t tag our links to the Play Store and iTunes, all the referrals will be sent to Direct Traffic in Google Analytics’ Acquisition Reports.
The concept of “acquisition” in app analytics is different to that of web analytics. The only data we can get is on source of install (with app opening). However, as we’ve just seen, it’s not possible to get this data for FB campaigns (nor for Twitter). Google Analytics for apps is, above all, for contents analytics, and only for “organic” sources at that.
As a result, unfortunately measuring FB Ads campaigns directly via Google Analytics acquisition reports for mobile apps is not a valid option 🙁
Alternatives for measuring app install campaigns via FB Ads
Or we could include/embed the app within FB itself (“Canvas is a frame in which to put your app or game directly on Facebook.com on desktops and laptops”).
Finally, there’s the solution that’s currently the most common: use paid-for services to measure installs, such as AppsFlyer, which charges per install measured.