It was a real pleasure to catch up with Samir Patel, VP/CMO at GrowthMachines, a Berkeley, California-based consultancy powering growth marketing campaigns. Samir and I have a lot in common: we’ve both been in paid search and SEO for many years, and we both watch the latest developments in search algorithms and AI very closely.
We cover lots of topics in our interview, including the frustration that many marketers are feeling right now with the forced “upgrade” to GA4. As Samir notes, while the move to GA4 might have great payoffs in the long term, in the short term it’s a “hot mess” that is imposing major costs on marketers.
Samir and I trade notes on the potential impact of LLMs (Large Language Models, a type of artificial intelligence program that uses deep learning techniques and large data sets to perform natural language processing (NLP) tasks) on SEO, how they are useful from a content creation perspective, the likelihood that they will be abused by spammers seeking to flood the web with robotic, low-authority content, and the countermeasures that Google might employ to downrank AI-generated spam.
Samir notes that many SEO practitioners he works with don’t seem to be making any real effort to get up to speed with LLMs, despite the fact that AI-enabled search is likely to have a profound impact on SEO tactics and strategies that have remained relatively stable over the past decade. “Surprisingly,” he notes, “they’re not even trying the latest models out — it almost feels like ‘wow, OK, this is, like, in your face, this might eat your lunch, might eat your breakfast, might eat everything out of your fridge, and you’re still not even trying it?’ Some of them, shockingly, haven’t even tried ChatGTP.”
Samir and I agree that voice search might be the first area where AI will make a major impact, given that those making voice queries are accustomed to single, (hopefully) authoritative results. Our conversation then progresses to a discussion of how LLMs might improve their utility should they be given access to more personal, historical data relating to the user and chat about the trade-offs between user privacy and user convenience that might be necessary for LLMs to further enhance the online user experience.