By Sofia Collins and Helen Udolloress

There is no questioning the value of Social media advertising in today’s digitally savvy world. It is in fact predicted that advertising spend on Social media will overtake TV spend in just a couple of years’ time, and that makes now as good a time as any to understand its true value.

Back to blog home

However, when it comes to attribution, Social remains somewhat “anti-social”.

When we inevitably try to understand how social is impacting other channels and generally performing within the wider marketing mix, Social often remains one of the most difficult channels to integrate. It has therefore become a recognised problem as to how we can determine its full worth, especially as the typical user journey to conversion is becoming ever more complex, and the product of many multi-channel touchpoints.

Social media

What is the challenge?

When you review your marketing across channels in Google Analytics (GA), DoubleClick Campaign Manager (DCM) or other media and analytics platforms, you will no doubt begin to question your Social activity’s true impact; the numbers just don’t match – and you can normally expect the role of Social to be significantly downplayed.  

If you’re left questioning: Why don’t Facebook’s conversions align with those reported in GA? Which of the two is more accurate? How can Social reporting become more integrated with other channels? And how, ultimately, can Social spend be best optimised? You are not alone. This is an industry-wide problem, so read on to find out more!

What is Attribution? A quick refresher

Example attribution model

Attribution is the science of assigning credit for a sale to all marketing touchpoints that a customer was exposed to prior to their purchase. The purpose is to credit marketing activity with a fair measure of its impact and influence on that customer's purchase.

This is especially important for marketing activity designed to engage customers early on in their research phase, rather than lead to a direct conversion, as their impact will be severely undervalued by the default last click reporting model. And, as you’ve guessed it, this applies to Paid Social as an upper funnel marketing channel; one that plays a large role in the awareness phase.

What does Paid Social reporting currently look like?

Most Paid Social reporting is done within bidding platforms, such as Facebook Business Manager (FB), Twitter Ads, and sometimes 4C. These platforms report on a ‘Last Social Click’ Attribution model in which Social takes 100% of the credit for all conversions where the customer viewed or interacted with a Social ad. This model tends to inflate Social credit as it can ignore the impact of other channels altogether. Conversions may also therefore be double-counted across channels as a result of this, as other bidding platforms will attribute the same credit to their own channels where they featured in the path alongside Social. This is why it is crucial to try and consolidate cross-channel measurement within one single source of truth. 

Why don’t Facebook’s figures match those in GA?

GA is a great place to start if you are looking to understand cross-channel customer experiences with your Brand and begin consolidating your marketing measurement. It will initially, however, raise a lot more questions than it answers as you will notice large data discrepancies between GA and FB. The reasons for this are not blockers to using GA but rather considerations to be aware of, and can be broken down into the following sections:

a) Discrepancies in the data

Social typically doesn’t lend itself well to sharing its data with third-party analytics platforms such as GA or DCM, limiting tracking of its activity to clicks through to site and providing no visibility over ad impressions served. This leads to Paid Social being undervalued in GA, especially as compared with Facebook which takes “view-through” conversions into account (i.e. conversions that occur after a user has viewed a social ad, and then converted after reaching site via other means).

Facebook ads

Another difference between the two data sets is FB’s ability to connect user behaviour across devices. GA currently reports on single device meaning that a user’s journey involving multiple devices will be fragmented in this view of reporting. To illustrate: if a user first reaches your site through a Facebook mobile ad, but then later purchases on their desktop (the last touch-point being a Search ad), GA will have no visibility of the fact this was the journey of a single user, but count them as two separate paths. 100% of the credit for the conversion will be given to Search, and the Social initiated journey will be viewed as one that did not convert.

b) Discrepancies in credit allocation

Another thing to bear in mind is that GA’s default model is “last non-Direct click”. This means Paid Social would receive 0 credit in GA when it is followed by an interaction with another channel – whilst in FB, it would get significant credit for that conversion. 

The default lookback window also varies between platforms. This corresponds to the period of time in which to attribute credit for a conversion to marketing activity. The FB window is 1 Day for Views and 28 Days for Clicks, whereas in GA the default is 30 days for both clicks and impressions (though as mentioned, Social impressions are typically not shared with GA by default).

Attribution examples

For these reasons, it is important to leverage both Paid Social’s bidding platforming and cross-channel measurement solutions to understand the true impact of this channel. 

To find out how Merkle | Periscopix can help you achieve this, please get in contact.

Share this article