Attribution Solution Options and a Quick Way to Get Started

May 14, 2020

Korey Thurber, SVP - Data Science & Marketing Service Solutions

As a follow-up to my Attribution related post on November 13 of last year – “Don’t Waste Your Marketing $$!  Why Good Attribution is Critical” – I will briefly describe a few attribution options to help get past the dreaded last-touch method that assigns 100% of a purchase (or some other desired behavior) to the last interaction that a prospect or customer had with your company – whether that be a direct response to a direct mail piece, a click on a digital ad or an organic visit to a website.  


The obvious problem is all other touches and interactions that happened before that last touch are completely ignored, ultimately resulting in a very incomplete and inaccurate assessment of the contribution each of those touches made toward incenting the prospect or customer to do what you wanted them to do.  So, what is a marketer to do?  What are some of the options?


1)      Linear Attribution – distributes credit evenly across the various identifiable touches.  For example, if you can identify 5 touches a customer received within a defined tracking window that ultimately led to a purchase, then each touch receives 20% of the credit for that purchase.


2)      Lead Conversion Attribution - gives the credit to the touch that directly resulted in the generation of the lead.


3)      Time-Decay Attribution – gives the most credit to the touch that is closest in time to the actual purchase.  The further away in time a touch is, relative to a purchase, the less credit that touch is assigned.

4)      Modeled Attribution – through advanced (and proven) analytic techniques, a weighting calculation is developed and applied to the various marketing touches during the customer’s buying journey.  In short, using the resulting weighting algorithm, a fraction of the customer’s purchase is attributed to each of the marketing touches that impacted the customer’s decision to buy.  Though this approach is often seen as the most accurate, it’s also the most technical, time consuming approach….and often the least practical.


So how does a marketer choose the right solution?  


Each of the options briefly described above have their positives and negatives, but there is no simple, one-size-fits all answer as each business situation is unique with its own unique set of needs and challenges.


A great way to get started is a project I recently proposed for one of my clients that is currently leveraging direct mail, email, paid search, paid social and organic search to acquire new customers, but my client does not have a clear view of the value being driven by each channel.  Critical to note here is the most important metric to consider in any attribution solution….is Customer Value!


We decided to first start simple by creating a table of newly acquired customers and identifying their “Primary” and “Secondary” response channels.  We defined these response channels as:




Utilizing the table described above, we will then create Acquisition Channel Segments based on Primary and Secondary channel usage and offer.  Next, we will examine 12-months of purchase history by the above-mentioned segments to create CPA, ROI, breakeven and loyalty metrics.  By doing so, we will generate a rich set of initial attribution insights regarding what channel or combination of channels is driving the most value!


Finally, we will construct deep customer profiles for the Acquisition Channel segments using transactional data as well as 3rd party data elements.


That’s it for now!  Please keep an eye out for my next blog post on the evolution of the project described above!