It's been a common question asked by many clients Wildfire Digital has worked with over the years - how accurate is Google Analytics social data? Is everything being correctly attributed? How come Facebook says X and Google Analytics say Y? They are common questions and also very valid questions. Oftentimes the agency reporting on overall website performance is not the same agency reporting on social activity and if there's a discrepancy, well, nobody wants to be wrong and with the figures sometimes painting wildly different stories what is the client to think? Answers are needed... Firstly we'll address some short, sharp reasons why there are some discrepancies, often, if not always shown between platforms. Then we'll talk about the common comparison we get asked about, Facebook VS Google Analytics. Finally we'll show you how to improve the social data within Google Analytics as it can miss out on plenty of activity - to the social agency's despair.
Top 5 Reasons For Analytics Program Discrepancies
- How they collect data can differ
- How much data they collect and analyse (sample sizing)
- The metrics they report on, which can differ even though may sound the same, e.g. clicks vs visits vs views
- Unique vs non-unique visitor counts – due to different cookie specifics
- Attribution – should a channel get the value if they played a part in reaching the visitor initially? Some programs think yes others think no.
Facebook VS Google Analytics
One of the most common discrepancies we've come across is with Facebook and Google Analytics. Both Facebook Insights and Facebook Ads Manager show you how many people clicked an ad, post, or page link. Google Analytics similarly shows you how many people were referred to your site by Facebook in the form of sessions, and allows you to drill down by landing page resource. However, ‘clicks’ are not the same as ‘sessions’, and conversions aren't triggered the same either - see below.
Discrepancy 1: If a user clicks your Facebook post more than once in 30 minutes (let's say twice for simplicity), Google Analytics will as standard record only a single session. In this case, 2 Facebook clicks = 1 Google Analytics session. If you've got an uncapped frequency set in your Facebook campaign; say it's a short campaign where you want to inundate visitors for mass awareness and recall, this 2:1 ratio can tally up to a big discrepancy.
Discrepancy 2: When a user accidentally clicks your Facebook post and immediately clicks out of the still-loading landing page, Google Analytics may not have time to record a session. In this case, 1 Facebook clicks = 0 Google Analytics sessions.
Discrepancy 3: Facebook takes credit for a conversion even if a Facebook user only sees your Facebook ad and doesn't click it, then visits your website through other means later on (assuming that conversion has been adding to the Facebook Ad Manager). These are called view-through conversions. Google won't attribute these conversions to Facebook. Although Facebook may have played a part in the conversion, if it wasn't the direct source at the point of the conversion happening then it's giving credit where it probably isn't due. Tip: This can be edited in Facebook settings.
So, What Do We Do About It?
In short, we're not comparing apples with apples! Google Analytics can be shown as the unbiased entity, most of the time, but it isn’t perfect either. You just have to be careful with programs ‘marking their own homework’, so it is wise to take numbers from all programs being used to get a roundabout estimate of performance - keeping in mind the metrics they use and how they are interpreted. By comparing Facebook-reported clicks with Google-reported Facebook referral sessions, we can better assess whether or not those clicks in Facebook are truly valuable – this is where Google’s engagement metrics come in and helps with that end-to-end view.
How To Improve Google Analytics Social Attribution
Let's get straight into it. If you go into Google Analytics and then Acquisition & All Traffic & Channels and view the Default Channel Groupings, you might see a source that says (Other) where Google Analytics does not have the information to put that activity into one of it's default channels. If you click on (Other) and see any Source that is social activity (as we do here), then this should be attributed to the Social channel.
You can add Medium as a secondary dimension to investigate further and to use these details to help group this activity. Below we can see 'organic-social' can be used to set a rule.
Once you have identified the information to use to set a rule go to Admin & Channel Settings (in the Google Analytics View you are dealing with) & Channel Grouping. Here by default you will have one called Social. Click into it, click OR as you want to set another rule to add to the Social Channel, and then use the options. In the below example we've used the Medium should contain 'organic-social' - then click 'Done'.
It will take around 24 hours for the data to be amended to cater for this change.
How To Correctly Report On Paid Social Activity VS Organic Social Activity In Google Analytics
You always have your social network ad platforms/managers to help understand your social performance and all the types of posts and paid activity you are doing. If your activity is leading to your website then to get a full end-to-end view, you'll need to involve your analytics platform. In this article we're just dealing with the common Google Analytics platform but other platforms may have the same kinds of options or settings to achieve the same kind of thing.
So, Google Analytics, where do i find my social activity? Well it's in the All Channels report, it's in the Social report and it could even be in the Campaigns report. The main problem we have here is how Google Analytics attributes social activity by default. If your social activity does not exactly match the preset parameters made by Google Analytics it won't be counted - even if the Source is correctly identified, if the Medium is anything other than 'referral or 'social' that activity won't be found in the Social channel.
What's the solution?! UTMs, UTMs, UTMs! Adding UTM parameters to both your organic social posts and paid social activity will allow you to capture everything. Because of the nature of appending UTM parameters to your landing pages, you'll also get much better granularity on exactly what activity resulted in what performance, so great for side-by-side comparisons and adapting that learning for future activity.
Where will this new tagged activity be found? It will all show in the campaigns report if you have added a campaign parameter to your activity, otherwise you'll need to follow the same steps as above, creating rules in your Channel Settings within the Admin section of Google Analytics. It's a good idea regardless as that All Channels report is arguably the most common report marketing manager, strategists and data analysts use to gain that holistic view on channel performance.
Splitting Paid Social and Organic Social Something we were asked to investigate from a client was around paid social activity and organic social activity. The client wanted to understand how they performed separately, to compare them. This was a great suggestion as your organic social activity gives you a kind of benchmark to what your social performance is - you can then compare paid amplification of that same post or just paid posts in general, to see if it is worth the spend.
For this to work it's a two system, or two agency approach. The agency responsible for the social activity are tasked with setting up UTM parameters using a standardised format. The critical component was to ALWAYS use for the Medium parameter 'organic_social' or 'paid_social'. We then added a new Channel Grouping called 'Paid Social' where all 'paid_social' activity moved into. We amended the Social channel to also attribute activity with the Medium 'organic_social'.