Different types of attribution

Attribution is the insight in touchpoints a person or group of people has undergone in their journey to the desired result or conversion. In other words, which touchpoints have more or less contributed to the conversion?

In Billy Grace, we use three types of attribution. We always like to explain attribution via a football analogy.

In the above visual you see from left to right: Last Click attribution, Multi-touch attribution and Unified Marketing Measurement. Now this is true for a game of football where in reality a lot of factors are playing a role in the process that leads up to a goal. Now what happens when we replace the players for touchpoints in the customer journey?

Read on for an explanation per model on how Billy Grace leverages machine learning models to determine which channels are responsible for conversions.

The conversion is granted to the last touchpoint, based on last click.

Billy Grace offers its own deep learning-based multi-touch attribution model. At Billy Grace we call this Multi-touch Attribution (MTA).

A deep learning-based multi-touch attribution model uses machine learning algorithms to analyze data from various sources, such as user behavior data, campaign data, and conversion data. MTA will assign credit to each touchpoint along the customer journey based on its relative contribution to the final outcome.

This allows businesses to gain insights into how each touchpoint in the customer journey impacts conversions, sign-ups, or other goals.

For example, if a potential customer clicks on a social media ad, clicks on an email link, and then makes a purchase after visiting the business's website, the model will assign credit to each touchpoint along the customer journey. This approach provides businesses with a more accurate understanding of the value of each touchpoint and allows them to optimize their marketing strategies accordingly. Visually, this would look like this:

Important: In session mode, the value, or conversion, is attributed to the day that the touchpoint took place. Take this into account when you change the attribution window, as the window gets bigger the more conversions that took place in the current period could be attributed to sessions that happened outside the current period, resulting in lower numbers.

Overall, the deep learning-based multitouch attribution model is a powerful tool that helps businesses understand the complex and varied paths that customers take to convert, and provides businesses with insights to optimize their marketing strategies for maximum impact.

For clients that have sufficient data, the Unified Marketing Measurement model will be available. Unified Marketing Measurement (UMM) leverages both Marketing Mix Modeling (MMM) and Multi-Touch Attribution (MTA) methodologies, combining their unique advantages through the power of Machine Learning (ML) and Deep Learning (DL) technologies. In essence, UMM models if impressions have an effect on the starting of sessions via other channels (direct, Organic, Google, e-mail) if the model sees a significant effect part of the attribution value is distributed to impressions.

Unified Marketing Measurement will become available if all the following thresholds are met:

*The experience may vary for each use-case. Increasing your spending across various channels and days results in more data, enhancing the likelihood of achieving significant outcomes that differ from Multi-touch attribution. The specific threshold varies depending on the business case. If UMM is still unavailable after meeting the first two thresholds, please contact our team for assistance.

Unified marketing measurement allows marketers to understand the impact of their strategies at both an aggregate level (MMM) and at the level of individual customer touchpoints (MTA).

Visually, this would look like this:

Important: The attribution value of impressions is modelled, it is not possible to link impressions to specific users.

​Why should I use Unified Marketing Measurement?
In a digital marketing landscape, customer journeys can be complex, spanning multiple channels and touchpoints before leading to a conversion event. It's essential to have a holistic understanding of which marketing efforts are effective. The UMM method provides such an understanding.

This leads to: Increased accuracy in analyzing marketing performance and improved budget allocation.

How does this impact marketing performance?

By including the external factors that impact the start of different types of touchpoints, the model can allocate conversion value more accurately by estimating the impact of top of funnel channels and campaigns.
​

Insights by attribution

Attribution modelling helps marketeers to make the right choices while optimizing their advertising. The right attribution model can answer the following questions:

The Multi-touch attribution model of Billy Grace gives insight into which channels and campaigns really contribute to achieving your goals and helps to make the right decisions.

With the end of third party cookies insight and the coming of IOS 14.5 it becomes harder and harder to gain full insight in the conversion path. The Billy Grace pixel only uses first party data. Based on this first party data, Billy Grace gives accurate insight in the performance of your ads.

With the update of our latest attribution model that incorporates Marketing Mix Modelling techniques, you are even better equipped to deal with the problems that the post-cookie world brings in terms of analyzing marketing performance of your ads.