Attribution models

What are attribution models?

Attribution models analyze different userTouchpoints and marketing channels up to the point of conversion. The Main function of an attribution model is to analyze the value contribution of different Touchpoints or marketing channels. These models are crucial to ensure transparency and understanding of the user path to conversion. For manufacturers and producers, this analysis plays an important role in determining the influence of each individual channel on the overall conversion rate.

Value propositions and marketing channels

Attribution models help marketers to evaluate the contributions of different marketing channels to the achievement of objectives. Different Attribution models assign different Touchpoints or channels, which can lead to differences in evaluation and analysis. These models provide a more accurate insight to plan future budget decisions and optimize marketing strategies. The evaluation is based on the assumption that each Touchpoint along the Customer Journey has a specific influence on conversion.

Intended use and benefits

The Main purpose of attribution models is to generate valuable insights for the allocation of marketing budgets. This not only increases efficiency, but also the effectiveness of marketing efforts. By analyzing the entire Customer Journey the strengths and weaknesses of the existing marketing strategy and can identify specific areas that need to be optimized. In addition, a comprehensive understanding of the Attribution modelsto create clear and detailed reports that support strategic decisions.

Types of attribution models

Attribution models are available in various forms, each with its own advantages and disadvantages. One of the best known methods is the Last interaction Attributionwhere 100 % of the value of the last interaction before the conversion is assigned. This method is easy to implement and often more accurate for short purchase cycles. The First interaction Attributionwhere 100 % of the value corresponds to the first Touchpoint which is useful if the first contact is particularly important.

More complex attribution models

The Model Last non-direct click only takes into account the last non-direct interaction before the conversion and eliminates direct access such as a direct visit to a website. This model is helpful to specify the effectiveness of specific marketing strategies. Another example is the Linear attributionwhere the value is distributed evenly across all interactions. This method provides a more balanced overview and is useful for determining the value of each interaction. Touchpoints to be displayed.

The Time course attribution gives more weight to later interactions, which is useful for longer sales cycles and for building relationships. Finally, there is the Position-based attributiona hybrid model in which the value is split between first and last contact and the interactions in between. This method recognizes the contributions of all Touchpoints and is particularly useful for business models with multiple Touchpoints before a conversion.

Example models for attribution models

There are various models that are used in attribution modeling. One model that is easy to understand is the Model Last interaction. Here, 100 % of the value is assigned to the last Touchpoint that precedes a conversion. This is particularly useful for short sales cycles, in which the last contact is often the decisive impulse for the purchase.

Diverse approaches to value assignment

Another approach is the Model First interaction. In this model, the entire value is assigned to the Touchpoint who started the conversion path. This model is particularly suitable when the first contact is decisive for the further course of the purchase process. There is also the Model Last indirect clickwhere the last channel before the conversion is taken into account, while direct accesses are ignored.

An alternative approach is offered by the Model Last Google Ads-clickwhere 100 % of the value corresponds to the last Google Ads-click can be attributed. This is particularly valuable when Google Ads plays an important role in the marketing strategy. A balanced approach is the Linear modelwhere the value is evenly distributed Touchpoints so that each channel carries an equal share of the value.

Other models such as the time-weighted model and the Position-based model offer even more differentiated approaches. In the time-weighted model Touchpoints closer to the conversion has a higher value. The position-based model typically distributes 40 % of the value to the first and last contact and the remaining 20 % evenly to the contacts in between Touchpointswhich is a combination of the two models mentioned above.

Comparison of attribution models

When comparing the different Attribution models shows that each model offers different insights into the value of the marketing channels. The fundamental difference lies in the way in which the individual models analyze the contributions of the Touchpoints evaluate for conversion. Models like the Last interaction provide a clear, albeit one-dimensional, view by attributing the entire value to the last interaction. These models are easy to use and are well suited to short buying cycles.

Different weightings and their effects

In contrast, the Linear model an even overview of all Touchpointsby distributing the value evenly. This allows marketing managers to better recognize which channels along the Customer Journey are used particularly often, without a specific Touchpoint to be preferred. Also the time-weighted model stands out by valuing later interactions more highly and is therefore better suited to longer purchase cycles.

A flexible approach is the Position-based modelwhich uses a hybrid methodology and assigns value to both the first and the last Touchpoint As a rule, 40 % of the value is allocated to the first and last contact and the remaining 20 % to the interactions in between. This differentiated evaluation enables a more comprehensive analysis of the Customer Journey and can be particularly useful for complex sales cycles in which multiple Touchpoints play a role.

The comparison shows that the choice of a suitable attribution model depends heavily on the company's objectives and the specific purchasing behavior of the customer. While simple models such as the Last interaction can be implemented quickly, advanced models such as the Position-based attribution The new tools provide more detailed insights and support strategic planning through a more nuanced perspective on the value of individual marketing channels.

Example of attribution modeling

An example of practical attribution modeling shows how different customer interactions can be evaluated before a purchase is made. Assume a customer interacts via several channels before making a purchase. Purchasing concludes a contract. First, he visits the company's website via a Google search. Shortly afterwards, he clicks on a Facebook ad and visits the website again. Finally, he returns directly via the URL and makes the conversion.

Comparison of different models

Applying the Model Last interaction 100 % of the value is assigned to the last Touchpoint before the purchase, in this case direct access. In contrast, the Model Last indirect click ignore the direct interaction and attribute the entire value to the previous channel, in this case the Facebook ad.

At the First interaction-modeling, the entire value is assigned to the first interaction, i.e. the Google search. In the Linear model on the other hand, the value is evenly distributed Touchpoints so that each channel receives a share of 33.3 %.

The Position-based model would typically distribute 40 % of the value between the first (Google search) and last interaction (direct access), and the remaining 20 % evenly between the intervening Touchpoints in this case the Facebook ad.

Another example is the time-weighted modelwhere later interactions have a greater weighting. Here, direct access would receive the largest share of the value, followed by the Facebook ad, and Google search would have the smallest share.