It tends to happen that between the Adtomic platform and Google Analytics there will be differences as far as attributed sales for each channel. This has to do with the attribution models for different sources of information that we use for our app and the one Analytics uses.
👉The attribution models on the Facebook and Google platforms are focused on what the impact of their ads are and they take into account when the user interacted with them.
👉Analytics, on the other hand, reports all the e-commerce sales by focusing on the last channel with which the person interacted before completing the purchase.
Analytics is a very complete and useful tool for analyzing metrics in relation to the website, the traffic, and generated transactions. It takes into account all the channels that traffic is generated from to your site:
References (links that lead to the website)
Direct (when the person visited the site directly without going through any other channel)
+ Other paid or organic means
In this sense, Analytics supplies a vision that allows you to understand how the different channels interact and how they impact your website. Nevertheless, Analytics can be a double-edged sword when it comes time to analyze each channel separately.
Why don’t we use Analytics for reports with our tool?
🔺The problem with that tool is that its attribution model isn’t right for all channels. Analytics measures last click, which means, the last click that the user carried out before purchasing.
➡️An example that is usually used for explaining the unfairness that Analytics commits with Facebook is a typical situation: when you go to work, you look at your social media and you probably click on ads for things that interest you.
At that moment, although you had the intention of buying, it’s impossible to do so because it implies pulling out the credit card and going through the whole payment process which, while in public transportation, is hard to do.
So, it’s probable that when you get to work, you will search for it on Google from the computer and end up buying it. In that case, since Analytics attributes the sales by the last click, this sale would be credited to Google (which could be due to a paid ad or an organic one), even when if it wouldn’t have been for the Facebook ad, you would never have gone in and searched for it with the intention of purchasing on the website.
In this sense, it often happens that sales that should be credited to Facebook are adjudicated to channels that are focused on the lower part of the purchasing process, such as Google. And if we’re going to take the Analytics results into account to redistribute the budget, we would start from a mistaken diagnosis because we overvalue Google and underestimate Facebook. Put simply, if it were up to Analytics, far less would be invested in Facebook.
For the person to get in through Google, they have to first create the need. And that need is generated from the Facebook and Instagram platforms.
🔺Another inconvenience that this tool has is its deficiency in a sales adjudication based on the Cross Channel. What does that mean? That it doesn’t always unite the information from a single user if they interacted from two different devices.
This means that if a person sees an ad from their cell phone and then purchases on the computer, Analytics won’t always take them as a single user that used several channels to reach the conversion moment, but rather it often assumes that they are two different people. This happens because not everyone has their Google user open on all their devices.
This also impairs the Facebook and Instagram results because in this case, Analytics would have assumed that Facebook had not intervened at any time in the conversion process.
So, although Analytics is a platform for following up on traffic and for having a holistic view of your site, we don’t think that it is reliable enough to attribute sales to the different channels. It doesn’t allow measuring and valuing the different channels involved in the purchasing process in an exact way. That’s why in our app’s control panel, we glean all the information from the Facebook and Google platforms.
What are the Facebook and Google attribution models like?
👉Facebook attributes by default with the 28 day click, 1 day view model. This means that sales that are generated within the 28 days after the ad click are credited, or if the user had an ad impression on the last day before the purchase.
👉Google attributes, by default, with the 30 day after click model. This means that if a sale is completed up to 30 days after a person clicks on a Google ad, this platform will give credit as its own.
On Facebook there are also other attribution models that come from different combinations days of click and days of visualization.
📌The choice of attribution model will depend on the type of business, of product, on the purchasing process, and the size of the website.
📌On Google you can also measure by last click, which means by the last click before the purchase. This is done through linking the Analytics account.
❗️Attention: We only recommend models outside the standard for websites with lots of traffic and conversion and with long and complex purchase processes.
Learn more about attribution models for these platforms in this article.
Is it possible to modify the attribution model in the Adtomic report?
Yes, it’s possible. Just as we said before, there are several types of attribution models for Facebook and one alternative for Google outside the standard format.
In this sense, it’s essential to know what the purchase process is for the products you sell in order to understand how long it takes the user to make a decision to complete the transaction. By knowing that, we can define what attribution model is the one that matches your report.