Attribution is the process/act of apportioning value between traffic attraction channels, clicks, advertising messages or other factors with which the user interacted on the way to conversion.
The path to conversion can be short, consisting of one visit (site or resource) during which the conversion occurred, or long. On the way to the conversion, the user can visit the site several times from different channels. These channels can be a social network, advertising, organic search, e-mail, direct site visit, and others.
A conversion on the site can be any important action on the site. It can be a purchase, filling out a form, clicking on a specific button, clicking on a specific link, etc.
Also, multi-channel sequences play an important aspect in determining the impact of different marketing channels on conversion. What is the difference between attribution and multi-channel sequences is another topic.
Continuing on attribution, it is important to understand that there are different models of attribution.
Attribution is needed to understand which traffic channels are most profitable. This is especially useful for paid advertising traffic. If we see in the attribution report that some paid traffic channel brings the most benefit, then perhaps increasing the spend on that paid channel will improve the results. Why is it possible? Because if the attribution is incorrectly configured, then the data in the reports about it may not be displayed correctly.
The more traffic acquisition channels are used, the more carefully you need to analyze the data before allocating the advertising budget. It’s much easier with one traffic channel.
Let’s continue about attribution models. There are different attribution models. The attribution model is the principle by which value is distributed between traffic channels on the way to conversion.
Currently, the most popular attribution model is the data-driven model. In this model, value is allocated to channels based on the algorithms of the tracking and analytics system used.
There are different analytics systems and they all use different attribution models.
Perhaps the most popular system today is Google Analytics 4. Advertising services such as Google Ads, Meta Ads, Microsoft Ads and others also have their own attribution models. Therefore, data in reports in different tools, advertising services and tracking systems may differ.
It is important to note that attribution models are divided into types of attribution models. There are two main types of attribution models, rule-based and data-based.
Take rule-based typing as an example. It includes such models as Last Click – when all the value is given to the channel from which the last click and visit occurred, before the conversion.
As for the type of data-driven attribution. This includes the data-driven attribution model.
How does a data-driven attribution model work?
This model uses machine learning algorithms to allocate the value of the contribution of each channel (touch point) on the way to conversion.
What factors does a data-driven attribution model consider when allocating value across touch points?
In the context of the paid channel, these factors can be: device type, number of interactions with the ad, time between interaction and conversion, order of their display and types of ad objects.
Let’s briefly summarize.
Attribution is the distribution of the value of each channel’s contribution to the conversion path.
An attribution model is a way to allocate this contribution.
Types of attribution models. Based on simple rules or based on machine algorithm data.
Leave a Reply