The future of digital marketing is already blueprinted by our client industries, most notably health.
In November 2019, a paper called “Integrating overlapping datasets using bivariate causal discovery” was published by Anish Dhir and Ciaran M. Lee (Babylon Health, University College London). The research team built on top of existing research in the field of bivariate causal discovery algorithms and developed an algorithm that works across multiple datasets with overlapping variables that’s able to spot cause and effect to assist doctors with patient diagnostics.
Now lets reflect on digital marketing. My clients typically come to me for any of the following reasons:
No matter which of the above I’m looking at, my starting point is always to acquire all available data. In the end I end up with multiple, often overlapping datasets and a whole lot of mystery involving rank and traffic fluctuations. As I described here, I do use algorithm tracking, but just to see whether it was a local event (on-page, redirects, speed), an external event (links, ads, branding) or a global event (algorithm change). Other than that I hold no hope in understanding what actually happened behind Google’s algorithm update.
Well until now.
Researchers at Babylon have tackled an infinitely more complex field and made impressive progress and even commercialised their work. Interestingly, they haven’t employed traditional machine learning, but instead they utilise the mathematics of quantum cryptography. While their work isn’t perfect (peer review reveals situations where their AI may be very wrong) their example offers hope for solving the complexities of forensic SEO and digital marketing in general.
I see the role of AI in digital marketing as something that surfaces insights, connections and information that would otherwise be inaccessible by human review.
This will free up an enormous amount of time and allow modern marketers to focus on creativity and decision making.