Every year, new buzzwords enter the B2B marketing lexicon and 2017 is no different. Inspired by new advances in machine learning and artificial intelligence, a new buzzword that’s making the rounds now is “algorithmic attribution.” So what exactly does that term mean?
“Algorithmic attribution” is a new, more sophisticated attempt to quantify which parts of the customer experience are most responsible for new leads, new sales opportunities and new revenue. As such, it represents a significant improvement to current attribution models, such as the “First-Touch” model, which assigns 100% of the value of any new sale to the first step of any customer experience, or the “Last-Touch” model, which assigns 100% of the value of any new sale to the final step of the customer experience. That first step is typically the process of generating the lead in the first place, while the last step is the final sales meeting.
But what about all the little steps that happen along the way? Determining a value for all the “middle” touches is the primary goal of algorithmic attribution. These “middle” touches might include things like scheduling a live demo for a customer, or having the customer download a free e-book or whitepaper. Arguably, these steps are just as important as the first touch or the last touch. But how to assign a value to them?
That’s the goal of a new generation of machine learning software, which is attempting to refine and improve existing models. The idea is that companies can create very customized models based on historical customer data. Then, the software will “learn” over time which factors should be given the greatest weight in a sales model.
For example, it may turn out that the proper “attribution” model is 50% for the first touch and 40% for the e-book download, and just 10% for the final sales meeting.
At the end of the day, the goal is to get more visibility into the overall B2B buyer journey. How does a prospect move along this journey, and which stimuli are the most important to get that prospect to commit to a final sale?
While “algorithmic attribution” sounds a bit intimidating, it’s essentially just a more sophisticated version of what marketers are already doing today. There is an insatiable need to know what’s working, and what’s not. So you can think of algorithmic attribution as a way to peer inside and see what’s really happening inside that black box.
Once the proper attribution ratings have been determined, then it’s a much easier task to decide how to allocate marketing dollars based on platform, channel or type of content. Maybe a company should be allocating more of its B2B marketing budget to social media, for example. Or maybe the company should be investing even more heavily into the lead generation process if it turns out that “first touches” are inordinately important in being able to close the final deal.
With new advances in machine learning, it’s becoming easier and easier to arrive at much more sophisticated models of what’s actually happening as part of the customer experience at every step of the journey.
IMAGE: Designed by Freepik
Dane
Dane Cobain is a published author, freelance writer, book blogger, poet and (occasional) musician from High Wycombe, UK. Before going freelance, he worked as a social media marker, now specialising in writing about publishing, technology and marketing.