Episode 10: Machine Learning 101 - The How, the Why, and What ML Means to Us as Marketers
Episode 10: Machine Learning 101 - The How, the Why, and What ML Means to Us as Marketers
Machine learning and artificial intelligence were being used interchangeably in 2018 in most marketing conversations, and imprecisely in most others. The conflation mattered because it produced muddled strategy: if you don’t understand what ML actually does, you can’t make sound decisions about how to optimise for systems that use it.
Episode 10 of the #SEOisAEO series was built to fix that. Purna Virji, Stephen Kenwright, and Lexi Mills joined Jason Barnard for a practitioner-focused ML primer - not academic theory, but the working knowledge a marketer needs to understand why answer engines behave the way they do and what they can be trained to do.
The distinctions the episode established are still the right ones. Artificial intelligence is the broad field - systems that perform tasks that would otherwise require human intelligence. Machine learning is a specific approach within that field - systems that improve their performance on a task by learning from data, without being explicitly programmed for each new situation. The difference matters for marketers because it explains why Google’s algorithms don’t have static rules you can optimise against. They have training objectives and feedback loops. Optimise for the objective, not the current output.
The episode also examined where ML was already active in the marketing context in 2018: content recommendation systems, ad targeting, spam detection, query interpretation, and - directly relevant to the series - the construction and maintenance of Knowledge Graphs. Each of those applications has deepened significantly since. The content recommendation ML of 2018 is the generative AI content engine of 2026. The query interpretation ML of 2018 is the natural language understanding layer of every AI assistant on the market today.
What the episode made clear for a 2018 audience, and what remains true, is that the direction of travel was set. Machines were getting better at understanding meaning, not just matching patterns. The marketers who understood that early built strategies that aged well. The ones who didn’t kept optimising for the pattern-matching layer that was already being replaced.
The presentation deck used during Episode 10 of the #SEOisAEO series is preserved on SlideShare.
Published by: Semrush. Host: Jason Barnard. Guests: Purna Virji, Stephen Kenwright, Lexi Mills. November 06, 2018
