Traditional marketers were worried about a product’s marketing mix (the popular 4Ps). Modern marketers are even more worried as there are more Ps added to the mix and also the option within a single parameter is widening, making sub mix. If you consider the ways of promoting a product, there are different channels through which you can accomplish it. In this digital age, there are many online touch points (e-mails, organic search, SEO, blogs, newsletters, website, affiliate marketing, online ads, etc.) through which you reach your customer. When a conversion happens, to which promotion channel do you attribute the conversion?
Digital attribution modeling helps one to give due credit to the digital media/channel for bringing about a conversion. The business objectives behind digital attribution modeling can be summarized by the following questions:
1. Which digital channels are to be selected in a Promotion mix?
2. What is the proportion in which investments are to be made to the selected channels?
3. How to optimize the conversion rate, given a budget?
Google Analytics and a few others offer simple models to work upon digital attribution. Last –click attribution gives full credit to the last channel where the customer has made a visit. There is also a linear model which gives equal weight to all the channels. Still in its fundamental research stage, statisticians and computer scientists are yet to come up with a model that is more realistic in capturing the right attributability of channels in bringing about conversion.
The challenge to modeling digital attribution is that it requires a great capacity to precisely track customers’ exposure to online campaigns. This is usually achieved through cookies; whose usage is being restricted owing to privacy issues. In such a situation, the other method is to observe the online behavior of select sample of willing customers. The limitation of this method is that the modeling becomes a one-time process. Digital attribution shall be modeled frequently, as there will be changes in the attributability of a channel over time.
(Alby is business analyst at Software Associates. Get in touch with him at alby [at] softwareassociates.in)