In my last article, I gave an overview of cross-channel marketing communications across multiple user touch points that lead to a conversion (like purchasing a new car). I also emphasized that without properly tracking and analyzing multiple touch points, marketers might not be fully optimizing their cross-channel marketing programs, thus leading to low ROIs. In this article, I present you an overview of different attribution models and strategies to champion marketing attribution in your organization.
Forrester provides a crisp definition for cross-channel attribution as “the practice of allocating proportional credit to all marketing communications, across all channels, that ultimately lead to the desired customer action.” The marketing communications interact with the end user at various touch points, like email, display advertisement, SMS message, TV ad, flyer in the mail, etc. throughout the paths that lead to a customer action, or a conversion. Google in turn defines an attribution model as “the rule, or set of rules, that determines how credit for sales and conversions is assigned to touch points in conversion paths.”
The main objective of an attribution model is to offer a simple mathematical formalism to measure the impact of each marketing communication at different touch points throughout the conversion path. For example, if a user received an email, saw a TV ad, clicked on a sponsored listing, and used a coupon code within five days before purchasing a product, the attribution model is used to assign credit to each of these marketing touch points. A smart marketer can then gain insights from the results of the analysis of an attribution model to more effectively orchestrate a cross-channel, multi-touch marketing program to increase conversions that in turn lead to higher ROIs.
Common Attribution Models
An attribution model consists of a set of sequential touch points that start with the first marketing communication that initiates the interactions with the user and ends with the final conversion event, like a product purchase. Different attribution models assign credit to each touch point differently depending on when the corresponding marketing communication occurred. One model may give more credit to the last touch point before the final conversion while another one may be more democratic and give equal credit to all touch points in a conversion path.
Here are the most common attribution models that you can consider for analysis of your cross-channel marketing programs.
Last Interaction. All of the credit goes to the very last touch point right before the conversion.
First Interaction. All of the credit goes to the very first touch point that initiated the sequence of marketing communications.
Linear. Every touch point gets the exact same amount of credit regardless of when it occurred during the conversion path.
Time Decay. The touch point right before the conversion gets the highest credit while the first touch point gets the lowest credit, and the touch points in between get lower credit as they move away from the final conversion toward the initial touch point.
Position Based. The very first and the very last touch points get equal credit and the touch point right in the middle get the lowest credit. Other touch points gradually get more credit as they get closer to the first and last touch points.
Of course there is always the option of creating your own custom attribution model by assigning explicit percentages of credit to different touch points throughout the conversion path of your cross-channel marketing program. This option may be most suitable if you plan to experiment for a while to come up with the model that fits your specific analysis needs that can generate higher ROIs for your programs.
(image source: Kensoo Local)
The Dilemma of Choosing the Right Attribution Model
In order to choose the right attribution model, you may need to experiment with multiple attribution models and consider factors like time of year, product category, amount spent, customer loyalty, initial and final touch-point channels, length of conversion path, etc. For example, if the conversion was around the holiday season for a consumer electronic item less than $200 promoted for a very limited time period through an online coupon code, perhaps the “linear” or “position based” model would be more suitable than a “time decay” model.
Because marketing attribution analysis is still more of an art than a science, experience and judgment are very important. Therefore, initially you may want to experiment with a few different attribution models for a rather small and simple set of marketing programs – maybe with only a couple of channels. You can then gain a wide range of insights and see if they make sense by discussing them with your expert colleagues. If you can afford it, a consultant with extensive experience in this field could always be of great help. Over time, you will gain enough experience to be able to choose the right attribution model for a specific type of cross-channel marketing program.
Finally if you are quite serious about attribution analysis and have a decent budget for third-party software, you may want to consider these products – but always remember that a technology is only as good as the people who build and use it!
There is plenty of information available online on these and other software products for attribution modeling and analysis. You may also want to read Tina Moffett’s 2012 Forrester Wave report titled: Cross-Channel Attribution Providers.
How to Master Marketing Attribution
Recently Adobe and Econsultancy teamed up to create a detailed report about Making Sense of Marketing Attribution. Below are five key steps they identified for mastering marketing attribution:
Therefore to become a marketing attribution champion, you need to focus not only on the technical challenges, e.g., collecting data from multiple sources and picking the right attribution model and analysis software, etc., but also on organizational opportunities, especially educating and convincing the business leaders to set the right priorities and get you the necessary resources.