Get your head around the advancements of AI and SEO with this webinar session that includes an overview of how AI is powering new search engine results. Jun 5, 2024 01:00 PM EST Register Now.


Attribution Modeling

Do you know which of your campaigns or funnels is yielding the best ROI? No? Well, you’re in good company. Countless organizations generate leads and achieve conversions without gaining insight into the source of these leads or which campaign produces the best outcomes. The cure for this lack of insight is attribution modeling. Attribution modeling is a method developed to analyze which marketing channels or touchpoints are responsible for a particular conversion.


Examples of attribution modeling include:

    • Last-click attribution:  In this model, credit is given to the consumer’s final touchpoint just prior to converting. Unfortunately, this one-touch framework ignores other contacts the user might have had with the business’s marketing efforts that led up to the last “click”.
    • First-click attribution:  First-click attribution (another one-touch model, like last-click attribution) gives full credit to whichever first action a customer takes on their journey to conversion. It disregards any subsequent touchpoints the customer might have with further marketing efforts prior to converting.
    • Linear attribution. A multi-touch attribution framework, linear attribution assigns equal credit for the conversion to all touchpoints along a customer’s path.
    • Time decay attribution. This model assigns more conversion credit to the touchpoints that happened nearer to the instant of conversion relative to earlier touchpoints in the conversion journey.
    • U-shaped attribution. Here, the customer’s first and final engagements get the majority of the credit, and the remaining credit is assigned evenly to whichever touchpoints happened in between. Google Analytics gives 40% credit to the first and latest engagements, while the remaining 20% is allocated equally to the interactions in the middle.
    • Algorithmic, or data-driven attribution. In algorithmic attribution, there is not a predetermined rule set for assigning credit. This method utilizes machine learning in order to study each marketing touchpoint and design a model for attribution based on the resulting data. Algorithms (and results) vary from vendor to vendor, and methods are generally not disclosed to partners. Google’s data-driven attribution is an example of this model.
    • Custom attribution. With this model, you can make your own rules for how to assign credit to various touchpoints in the conversion journey.

While attribution modeling can be (as it looks) quite complex, it’s a critical component to tracking and understanding how your customers are flowing through the journey you’ve crafted for them and at what point they’re converting. Let us here at Overdrive, help you build attribution models, and drive more successful conversions. Click to contact us.


What is attribution in digital marketing?

Attribution, (multi-touch attribution), is when you identify certain user actions that resemble an established “goal” or conversion”, and then the assign a certain value to these goals.

How is attribution measured?

One of the most effective ways to measure attribution is to create a custom model – allowing your teams to assign varying amounts of credit to specific touchpoints, based off the metrics you consider to be important.

What does marketing attribution mean?

Essentially, marketing attribution is the method used to determine the value / return on investment (ROI) of the various platforms and channels that bring your organization potential customers.

This definition is part of the Demand Ops Map. Get your PDF map with links here.

Click here to get the Demand Ops Map

Knowledge Center

Learn. Share. Succeed.

Benefit from our Demand Gen Expertise

Take action. Achieve your goals.

Dev Tool:

Request: demand-gen-ops/attribution-modeling
Matched Rewrite Rule: demand-gen-ops/([^/]+)(?:/([0-9]+))?/?$
Matched Rewrite Query: demand_gen_ops=attribution-modeling&page=
Loaded Template: single-demand_gen_ops.php