Table of Contents Hide
Attribution is one of the topics that often comes up when discussing conversions and return on advertising investments. Given the enormous stakes for companies that spend budgets on multiple marketing platforms, this is easily explained. They must understand how the user journey is composed to rationalize their expenses effectively. Let’s explore this essential topic in this article.
I – Attribution models
If you work in a marketing department, you have probably heard this question: “What is the best attribution model?”
Answer: there isn’t one. To conduct a good analysis, it is necessary to compare the different models and see which channels stand out and which lag in all models.
Here are some examples of attribution models:
Last click: the most commonly used. It attributes the conversion to the previous acquisition channel used before the conversion. However, before the conversion, this model does not consider the user’s interactions with the different marketing channels.
First click: it attributes the conversion to the first point of contact with the prospect. This model considers first impressions and can help companies understand which marketing sources are most effective in attracting prospects.
Linear: it attributes a portion of the conversion to each marketing source used before the conversion.
Algorithmic: advanced method that uses algorithms to assign a value to each visitor interaction with different marketing sources. This allows companies to understand how each marketing channel contributes to converting prospects into customers.
II- Visualizing Attribution
For more advanced companies, there are attribution-specific solutions like Easyence or Eulerian. However, they come at a higher cost, though they can be amortized through smarter advertising spending decisions.
Building your attribution model by aggregating all your acquisition data and working with your dataset using Python is also possible.
We purposely exclude attribution in marketing platforms like Google Ads and Facebook Ads, which offer a biased view of reality. Every interaction that leads to a conversion in the defined time window is counted. Adding sales from Google and Facebook may yield a higher number than reality.
III- The Limits of Attribution
Regardless of the solution chosen, it’s impossible to determine 100% attribution for several reasons:
Post-view attribution: Visual interactions sometimes don’t lead to clicks in a user’s decision-making process. While some solutions can measure these touchpoints, they may have reliability limits. It isn’t easy to accurately measure how many times a user viewed an ad. Furthermore, it doesn’t consider viewing time and user intent, so it’s difficult to claim that a simple view impacted the final conversion.
GDPR: Users who don’t consent won’t have their data and interactions measured on the site. This will create more or less significant attrition depending on the acceptance rate.
Ad blockers: They are marketers’ and ad platforms’ nightmares. However, this isn’t an unsolvable problem since server-side tracking can help bypass them.
Modern browsers: Like ad blockers, some browsers block and restrict tracking. For example, Safari deletes cookies left by your site 24 hours after a visitor with a tracking domain source (e.g., Google Ads, Meta Ads, and anything containing UTM parameters).
IV- The Value of Server-Side Tracking
Implementing a server-side tracking solution won’t solve the first two limits but will allow you to address the last two.
Through Google Tag Manager Server-Side, you can bypass ad blockers (made more accessible by using Addingwell as a server hosting solution) and rewrite your cookies, extending their lifespan to the maximum set by GDPR (13 months). This setup allows you to track user interactions with your site over time. It also allows you to maintain a user’s consent choice between visits. If you want to implement this in your container, don’t hesitate to read our dedicated article on the subject.
As you can see, attribution is a critical topic in digital marketing. While understanding the stakes is easy, the active part raises questions regularly. Nonetheless, using a server-side tracking solution improves your data quality, particularly if your on-site conversions require more than a day for the user.