SaaS Marketing Weekly

A handful of real data points surface (80% of LLM citations from third-party sites, YouTube prominence in AI Overviews, RAG mechanics) but they are diluted by extensive self-promotion, definitional throat-clearing, and recycled content-marketing platitudes. The ratio of novel claims to filler is low for a 23-minute episode.

“over 80% of appearing in LLMs is potentially third party websites”

“in AI overviews, YouTube has been found to be highly, highly cited source”

The core thesis - create genuine thought leadership, distribute widely, avoid AI slop, think top-of-funnel - is entirely conventional content-marketing advice simply relabelled for AI search. There are no contrarian or first-principles arguments, and the episode explicitly backs away from any tactical specificity, leaving only well-worn maxims.

“it's about creating really genuine thought leadership content that is different to the other kind of AI slop”

“It's consistency built up over time”

Daisy is introduced as a brand-new hire at the host's own agency, not an independent practitioner who has scaled AI-search programmes at recognised companies. One prior employer (Cognizant) is name-dropped briefly, but no scale, results, or seniority are established, making her effectively an internal team member rather than a high-calibre external expert.

“Daisy has recently joined RocketSAS as our head of SEO and AI search”

“at Cognizant we really focused on how we could create content that was more specific to the European market”

There are a few cited figures (95% of B2B buyers using LLMs, 80% of citations from third-party sites) and one named case study (Cognizant targeting European data-provider queries), but sources for the statistics are never identified, and the Cognizant example contains no outcome metrics, timelines, or dollar impact - just directional claims.

“95% of B2B buyers are now using LLMs as part of the journey”

“over 80% of appearing in LLMs is potentially third party websites. So that's off your domain”

The host routinely answers his own questions before the guest can, asks exclusively leading questions seeking validation, and uses significant airtime to promote his own agency and content output. There is no pushback, no probing of unsourced statistics, and no productive tension - the conversation functions more as a monologue with occasional guest affirmations.

“So is that a good way of looking at it? Like the kind of psychology of the difference between.”

“Um, you can correct me if there's things that I've missed there”