The Healthtech Marketing Show

There are several genuinely useful, non-obvious ideas packed in - LLMs weighting peer-reviewed evidence in vendor rankings, the preprint-to-chatbot pipeline demonstrated by Bixillomania, open access as an LLM crawlability issue - but the episode is diluted by extensive backstory, sponsor reads, summaries, and Mark's largely confirmatory commentary that adds little new substance.

“earlier this year some researchers submitted a preprint about a fake made up eye disease called Bixillomania...A week later, people with eye strain were being told by ChatGPT, ooh, maybe it's bixelomania”

“the fastest thing I've ever got now was 4 hours total time to get something in the Journal of the American Medical Association”

The core argument - that peer-reviewed evidence functions as an LLM discoverability asset, not merely a scientific one - is a genuinely fresh reframe for a marketing audience, and the Bixillomania example is striking and original; however, the surrounding advice (repurpose content, do open access, audit your AI presence) recycles fairly standard content marketing logic.

“what all of the chatbots say is that ada's got the most scientific evidence backing it up”

“large language models need to be regulated as medical devices. That now has been cited 200 times and has fed into conversations with regulators like the FDA and mhra”

Paul Wicks is a genuine practitioner with traceable, scaled output - neuropsychologist, led Patients Like Me's 110-paper publishing program in top journals, founded Proofstack which has worked with 50 companies and 200+ papers, and sits on the BMJ board; he's not a career podcast guest but a real operator, though the domain is niche and his company is small.

“patients like me published 110 peer reviewed publications in neurology and Nature and JAMA and Brain”

“we've done this for about 50 companies and published over 200 papers and I've sat on the boards of journals like the bmj”

The episode is notably concrete by podcast standards: named companies (ADA Health, Babylon, Patients Like Me), specific journals (BMJ Open, Nature Medicine, JAMA), cited counts (100+ for the BMJ study, 200 for the Nature Medicine paper), independent replication numbers (12 independent groups), and open-access cost ranges ($1,500 - $10,000); the Bixillomania illustration is also precisely described with verifiable details.

“we actually produced a study that we published in BMJ Open...comparing human doctors against ADA's AI based symptom checker and six of their competitors...that study has been cited over a hundred times”

“authors have to pay between $1,500 or a...the new Nature journals could charge you $10,000”

The host asks decent clarifying questions and occasionally invites specifics ('can you name examples?', 'talk about the process'), but there is no meaningful pushback - Paul's marketing claims about Proofstack go entirely unchallenged, Mark's contributions are mostly affirmatory, and the mid-episode sponsor read disrupts any momentum; the conversation feels collaborative rather than interrogative.

“Can you name any specific or give examples maybe without naming some specific companies that you've helping?”

“Talk about the process because I think for, you know, for most people listening to this sounds like a long, complex and potentially expensive process”