The tracking switch, formerly often referred to as the cookie switch,
is one of the essentials in e-commerce. TL;DR: A switch controls
Tags & scripts based on a user’s origin.
In the background, the structure is much more complex: The basic construct
The basic construct is usually a tag manager – i.e. a container with the help of which
scripts and codes relatively uncomplicated on the individual pages of a website.
website. In contrast to hard coding, scripts can be placed on many pages simultaneously in a single step. Triggers and exceptions can be used to determine exactly where a script is loaded and where it is not. The more data layer variables are included, the more specific tags can be implemented.
To turn this into a switch, the origin of the user is integrated as an additional trigger. This opens up the possibility of having certain scripts played only if a user had a very specific point of contact.
Why does this matter?
A customer can’t buy anything in any online store if he doesn’t even know that it exists. Whether new or existing customer – everyone became aware of the online store in one way or another. Not infrequently, this contact takes place via advertising, and advertising costs money. Accordingly, it is indispensable to check and optimize the success of advertising measures in terms of their cost-benefit factor. To do this, I need to know which paid advertising contact led to the customer’s final purchase.
This is where the aforementioned origin trigger comes into play: the tracking script that matches the customer’s last contact is always executed. In this way, it is recorded exactly which advertising campaign generated the purchase, while all other advertising trackers remain on the sidelines.
There are different logics according to which the final channel is defined. Currently, the most widespread method is “last-cookie-wins”. Means as much as: The last channel involved immediately before the purchase is attributed the sale. The opposite of this would be “first-cookie wins” – in this case, the salestracker of the very first point of contact is played out. Increasingly, online retailers are choosing to implement a so-called “basket freeze” within the switch. This logic is a hybrid form of both previously mentioned variants: It applies “last-cookie-wins” until the customer has added an item to the shopping cart. After that, the strategy changes to “first-cookie-wins”. This freezes the channel in the switch that led to the shopping cart being filled and then attributes the order to it.
Customer-journey and attribution
In most cases, a user has more than one touchpoint with an online store. Each of these touchpoints has an influence on the potential conclusion of a purchase. It does not matter whether they are paid or unpaid contacts. Since we determine the origin of users for the functionality of the switch, the logical extension is to also record the journey. Analyzing the customer journey then allows us to attribute value to all touchpoints. This means that we attribute proportionate importance to the individual touch points that contributed to the conclusion of the purchase. The logic of the evaluation must be adapted to the retailer’s marketing strategy. For example, if there are a lot of touch points in the customer journey, a bathtub model is a good solution. Here, the first and last touchpoints are given a high share, while everything in between has to share the rest. If the marketing mix is rather thinly spread, it makes more sense, for example, to set half of each of the first and last touchpoints in the evaluations. Everything in between is not taken into account here. It is important that the analysis of the customer journey is used for the internal evaluation of the individual marketing channels. The respective salt trackers of the marketing campaign should still be played out independently of this on the basis of the previously defined cookie logic.
The effects of the Consent on the switch
In recent years, the requirements for data protection have increased rapidly. In this context, tools for consent management have also made their way into e-commerce. The consideration of user consent has a direct impact on the functionality of the switch as well as on the attributed evaluation of the customer journey.
Unless the user consents, the switch is not allowed to mark the user’s touchpoints, nor is it allowed to play out third-party tracking scripts. This means: the user’s revenue cannot be statistically considered in any channel and no origin-based tracking script can be triggered.
At the moment, there are no reliable numbers about how high the rate of Consent Refusers really is. Nevertheless, it is important to think about how to deal with the expected statistical gaps at an early stage. It is more likely that further restrictions will be added here, which will require alternative technical solutions for tracking and tag management in the medium term.
