Attribution explained: What is the real ROI of Paid Social?

In 2018, companies in the United States spent close to 27 billion on social media spending alone. That ad spend will continue to grow – but what are we actually getting out of this effort? Sometimes, a paid ad strategy feels like throwing money into a big pit as a ritual sacrifice and hoping we will be rewarded. It can be so easy to choose to ignore exactly how social media platforms are charging you and what you are getting in return.

Why does Facebook show more conversions off paid social ads than Google Analytics?

Facebook results vs. Google results vs. Actual results

Because of “walled Gardens” – closed algorithms and data – a large portion of social ads data is not available to us and therefore accurately understanding ad results can be hard. So to verify ad results, we turn to third-party tracking platforms, like Google Analytics.

We have all experienced this situation – you see on Facebook manager that the results of an ad was a 5x ROI. You’re pretty happy with those results. Then, we go to Google Analytics to further understand the data, and GA tells us that zero conversions are attributed to this ad. You feel tricked and taken advantage of after spending money on this effort. In reality there is no trickery involved – Facebook is measuring apples while Google Analytics is measuring Oranges.

Facebook vs. Google Analytics Attribution Model

Attribution Model: An attribution model is the rule, or set of rules, that determines how credit for sales and conversions is assigned to touchpoints in conversion paths.

These measurement differences comes down the differences in attribution credit per platform. An attribution model, defined by Google Analytics, is the rule or set of rules, that determines how credit for sales and conversions is assigned to touchpoints in conversion paths.

Both platforms use a last-touch attribution model as the default model. This means that the credit of a conversion goes to the last interaction with a marketing effort before the conversion. It’s helpful to look at Google’s attribution data because it can give you a side-by-side look at the productivity of all of your marketing channels, and helps you compare the impact of each effort. Looking at Facebook conversion data only shows you one marketing channel, but it gives much more context and insight into your Facebook marketing efforts than Google Analytics can.

So what are the differences between Facebook and GA’s default attribution models?

Cross-device Conversion Tracking Discrepancies

How each platform tracks these conversions through a multi-touch purchase process. Facebook has an advantage of tracking Facebook users instead of cookies, which can track users across multiple platforms and devices. Facebook can follow users from phone to iPad to computer as long as they are logged in to Facebook. Google relies on cookies to track users, which means Analytics only tracks the browser where the cookies are dropped. If a user uses multiple devices during a purchase process, those will be tracked as separate conversion pathways.

For example: If a user sees an ad on their phone and then the next day uses their computer to visit the site directly and purchase a product, Facebook will attribute the conversion to the mobile ad, and Google will attribute the conversion to the direct website visit.

Click here to download the complete guide to timing your holiday campaigns.

Google can’t track conversions off impressions

While google can measure impressions off of Google ads, it doesn’t measure impressions from social ads. So according to Google, seeing an ad has no effect on the attribution during the purchase process.

Facebook can and does look at ad impressions when assigning attribution to a conversion. When a Facebook pixel can track data from ad impressions to website visits, Facebook will give credit to the last ad that was seen before a conversion – even if the viewer didn’t react to the ad.

Differing definitions of the ‘last click’

Facebook will grant credit to the last click of any Facebook effort. If no ads were clicked on before the purchase, the conversion will be attributed to the last ad seen before a conversion – coined a ‘post view conversion.’ This means even if the ‘last click’ before a conversion happens off of Facebook, with their attribution model a conversion can still be attributed to an ad.

Google on the other hand is more cut-and-dry. Google gives 100% of the credit to the last click regardless of platform. If Facebook is not the last click before a conversion, it will get zero credit for the conversion.

So, the conversion path shown on Google can have a completely different attribution than the very same conversion on Facebook.

Which results should I trust?

The true answer is that neither will be a 100% accurate on the ROI of your ad. That answer can be dissatisfying, I know. To get the most well-rounded results, look at the data on both platforms and use them to create your own holistic view on the amount of success. Consider Facebook results as the most optimistic view of the success of the ad, and Google Analytics as the pessimistic view of success. Neither can be dubbed as right or wrong – it’s more your perspective of the long-term influence ads can have.

Let’s say you track a specific customer purchase journey from start to finish. This customer journey lasted 1 month. During this, they visited the site, clicked on couple ads, checked some reviews, submitted a contact form and ultimately purchased. Did they purchase solely because of the ad? Probably not. Was the ad an integral part of the purchase journey? I would argue yes.

How can I accurately understand the success of paid ads?

If you want to measure the success of ads separate from the rest of your marketing efforts, there are a few ways to deepen your understanding:

Run a discount or special and create a custom promo code per platform. This will help you develop a much better understanding of the overall impact of ad impressions or engagement. Run this discount and look at the results – if Google says there are zero conversions attributed to ads, but the promo on the ads was used multiple times, that is valuable information. If Facebook attributes multiple conversions to an ad but the promo code was not used, that is also telling of the impact of impressions.

Dive into User explorer on GA to understand how and during what point of the purchase process people were interacting with your ads. User explorer can be really helpful in drilling down into and understanding a specific user’s purchase process. Let’s say you want to compare a conversion that has differing credit attribution on Facebook and Google. If the consumer purchase process shows multiple interactions with ads throughout the process, it could be argued the ad deserves the credit. If there are zero ad interactions and multiple interactions through other channels throughout the buyer journey, it’s harder to argue.

Use a third-party platform like DoubleClick by Google – alternative platforms have tools to help you work through the attribution grey areas. The tracking process can be time-consuming, but if you really want to optimize the attribution accuracy, these tracking platforms will get you as close as currently possible.

It’s easy to get wrapped up in the ambiguity of exactly which efforts are the most successful, but it’s important to take a step back and understand that we don’t have the technology to have a perfect measurement of our impact (yet).