Insight into Facebook Ads and Google Analytics Difference
September 26, 2017
Something that we are asked about time and time again is:
“Why is there a difference between the metrics that Facebook and Google Analytics reports?”
Firstly, don’t worry. This can be explained. You are not alone.
The very short answer is that there are key differences in how both platforms attribute conversions that us Facebook advertisers should be aware of. And the long answer is coming soon…but first some house-keeping…
It sounds obvious. Maybe even a teensy bit patronising. But the first step is to check that your Facebook pixel is installed correctly. If you’re not already using it, then install the Google Chrome Pixel Helper add-on. This is a great little time-saver that will tell you if your Facebook pixel is installed correctly.
Once you’ve done that, and all looks good, it’s then also worth checking your Google Analytics code is installed correctly.
However, even with this all set-up correctly, it is still unlikely you will fix the discrepancies. Sorry. But we’ll go over why shortly. However, it’s always good to check things over on your side of the fence first :)
So let’s take a quick look at how Facebook and Google handle conversions:
1. Conversion attribution date
First up, both Facebook and Google attribution conversion dates differently. Of course they do.
How Facebook attributes conversions
Facebook shows conversions based on the time of the ad impression, and not the time of the conversion. Let’s say someone is served an ad on 6th October (n.b. This is a random date and bears no significance. And it definitely isn’t my birthday). The person then makes a purchase on 10th October. Facebook attributes this purchase in it’s reporting to 6th October (i.e. the day the ad was served).
With this in mind, it is worth noting that anytime you pull stats from Facebook for a certain date or period less than 28 days ago, the data is likely to continue to change overtime. No, it’s not magic. It is because as more and more people convert, up to 28 days since the ad view/click, the data is going to change for the date of the impression. Hey, but this is a good thing - it means you’re getting conversions further down the line. It also indicates that there is a consideration period for purchaser’s of your products. Useful stuff.
How Google Analytics attributes conversions
Conversely, Google Analytics handles conversions differently. For the same Facebook campaign above, for the person that purchased on 10th October, in Google Analytics the conversion would be attributed to 10th October (i.e. the day of purchase, and not the day the ad was served).
Hopefully you are already doing this, but it’s worth noting, that in order to view stats for a particular Facebook campaign in Google Analytics, you should tag the ads with ‘utm tags’. This then passes through to Google Analytics reporting, so that you are able to attribute where your conversions came from.
So, to summarise, when you are comparing your stats, always be aware of the attribution model used. You really don’t want to be comparing apples with cookies…
2. People BASED MARKETING VS Cookies
One big advantage of Facebook advertising is that they can attribute actions to users, rather than cookies. Facebook can track across different browsers, and different devices (as long as the user is signed into Facebook) - they know when and how users are logged into Facebook.
With Google Analytics, rather than having user profiles to track, they use cookie tracking. Quite simply, all tracking happens in the same browser that the cookie was dropped into. If someone browses the same website from their mobile and desktop on the same day, Google Analytics will view them as two separate people. Facebook on the other hand, would know they were the same person.
So if someone sees an ad at work in their Facebook news feed but doesn’t click, then later when they get home they decide to visit the advertiser's website on their iMac, and purchases said product, Facebook is able to attribute this purchase to the ad the person saw (as long as they are logged into Facebook on both devices). Whereas, Google Analytics is unable to make this distinction, as the browser that the cookie was dropped into wasn’t the same one that the conversion occurred on.
This cross-device tracking is crucial in today’s super-connected world. Mobile is so entrenched in our daily lives, that it is always going to play a part in our purchasing habits, particularly in the research/consideration stage. The actual conversion on the other hand may happen when we are on our computers at lunch time, or at home on desktop when we have time think about it. According to Facebook, more than 65% of conversions start on one device and are completed on another. So with this in mind, Facebook reporting is going to be more representative of the conversions you get from your Facebook ad campaigns.
"According to Google, consumers tend to browse and perform pre-conversion research on mobile devices but are not ready to convert, whereas desktop visitors are already informed and ready to convert. Consumers may perform micro-conversion actions, such as email sign-ups on mobile devices, but tend to ultimately perform that final conversion action on a desktop computer.”
- Source: https://www.marketing-mojo.com
3. Last click attribution
Like many channels, Google Analytics attributes conversions to the last click. So for instance, someone may click an ad on Facebook but doesn’t convert, and then they later see an ad on Google Adwords, and click again, then purchase. In this instance, even though the Facebook ad made the person aware of the product, Google Analytics would give attribution to Adwords, and not attribute anything to Facebook, while Facebook would attribute the conversion to the campaign.
So if you are running a multi-channel campaign (Facebook, Adwords, email etc), then whichever of them receives the last click (and the user converts), then that will get the attribution in Google Analytics. Whereas in Facebook, regardless of whether they viewed or clicked the ad, you will see the conversion.
4. Google Analytics cannot track view-throughs
As we’ve seen with the last click attribution model, someone needs to actually click an ad for the conversion to be attributed.
But consider a user that sees an ad in their Facebook timeline, but doesn’t click it. They then go straight to the businesses website, and manually type in the domain name. They then purchase. Google Analytics would view this as an organic referral, whereas Facebook know that the person saw an ad, and therefore attributes the conversion to it. Their view is that Facebook has contributed to the sale in some way, through the user’s awareness of the product.
Taking this a step further - imagine a user has been retargeted with a Dynamic Product Ad campaign. This ad format is displaying them an ad for something they already have an interest in, and had previously added to cart. They do not click, but instead go straight to their Basket on the website and complete the transaction. This of course would make sense that they would do that, as they have no reason to click the ad, and are already very much product aware. The ad alone has reminded them that they had previously added a product to their basket, but not yet purchased it. Therefore, the Facebook ad should by rights take credit for having reminded them.
As Google Analytics has no insight as to which ads someone has viewed on Facebook, they are unable to report on view-through conversions. Only clicks. Therefore, Facebook is usually going to report higher metrics, for this very reason. But this isn’t a case of conversions being over-inflated. It can be argued that by viewing an ad in your Facebook news feed, regardless of whether you clicked it, you have been made aware of the brand, and therefore Facebook has had some influence in the purchase decision.
Of course, it is impossible to analyse exactly how much influence the ad had on the purchase decision!
5. Attribution window
Facebook reports on a 1-day view through and 28-day click through attribution window. Most 3rd parties track only on a click through window, which causes discrepancies because as mentioned above, Facebook also counts the view through.
So before you pull together any reports, make sure that the conversion attribution windows match for both platforms. By default, Facebook conversion reports are set to a 1 day view or 28 day click window. Google Analytics has a default 30 day window.
6. Moving from HTTPS to HTTP
Google Analytics gets referral information from the HTTP header, and uses these to credit conversions back to ads. However, if there isn’t one present, it will treat such referrals as direct. Without a referrer field, Google Analytics doesn’t know where someone has come from.
In addition, whenever someone moves from HTTPS to HTTP, the referrer cannot be recorded. So if a website doesn’t have a secure version (i.e. https), then the referrer may be lost. This can result in third party trackers under-reporting Facebook conversions by as much as 40%.
7. Cookie/Ad Blockers
Quite simply, if someone is using a cookie blocker, then Google Analytics isn’t going to work in tracking their activity. Similarly, if a person is using an ad blocker, you may run into issues with your Facebook pixel being loaded, and therefore not triggering for a conversion.
Therefore as a result, these stats may under-report compared to what you see in your internal metrics.
8. Site visit limits
Google Analytics limits a site visit to once every 30 minutes per unique visitor. This is in order to prevent spam visits being tracked. However, this also doesn’t take into account any users that leave and return to the website within that time-frame.
Facebook on the other hand doesn’t have this kind of limit, and can report on multiple visits per unique user within the 30 minute timeframe.
So for a user that clicked to your website multiple times in a 30 minute period, Facebook would count this as multiple clicks, whereas Google Analytics would see this as a single visit.
It’s clear that there is no uniform manner in which 3rd party trackers and ad networks track activity. Primarily this is down to the data they have available to them, and the manner in which they are able to track people online. Facebook on the one hand has the benefit of knowing exactly who a user is and what their Facebook id is, and therefore can track cross-device conversions. While Google Analytics is limited to relying on cookie data.
Ideally, you will have internal data that you are able to use to match up your conversions, and attribute referrals where you can. If this isn’t possible, then choose the attribution model that works best for you. If you are confident that Facebook view-throughs contribute to your conversions, then this is a good measure. Keep an eye on the difference between purely clicks driven conversions versus including view-throughs, and see how much of an impact this has.