>> Listen to the episode on Spotify, Apple Podcasts, and others.

TRANSCRIPT:

 

DOUG: We’re going be talking about the intersection of media relations, and healthcare, and surprise, AI and the effect it’s having. So, Campbell, when you look at the landscape and how AI is affecting it, what’s your headline?

 

CAMPBELL: That’s a great question. I think the headline here is that AI is changing how stories surface, but it has not changed what makes stories credible. AI is reshaping discovery in the sense of how people find information, how they encounter a story, and form that first impression. That’s what’s rapidly evolving. And in fact, a new poll from KFF found that one in three American adults say they turned AI chatbots for physician or mental health info in the last year, but what ultimately makes health related content and communications credible is trusted, accurate, and scientifically backed info. That hasn’t changed. And that’s great news for communicators who have always been invested in building credibility.

 

DOUG: Yeah. And I actually call it, like, doctor AI even if I’m talking to one of my doctors. I said I checked with doctor AI, and I’ll get, like, a little bit of a smile from them but, you know, then they’ll tell me what they think, which is pretty cool. Is it changing the strategy at for all communications? Obviously, the core of communicating good stories and important information is there, is what you’re saying, but our strategy is now being adjusted to sort of compensate for all these changes that are happening so rapidly.

 

CAMPBELL: There are really three big things that come to mind for me here. I mean, I think first, it really raises the stakes on accuracy and specificity. AI systems are summarizing our stories at scale. So, if the underlying source material is vague or thin, the AI output will reflect that. Second, it changes where we invest. The publications carrying the most weight with AI systems are not always the ones with the biggest readership, you know? Deep, authoritative, and frequently site outlets carry a lot of weight too. And I think the third big thing is it really just changed how we think about message consistency and how important that is because in a world where AI is aggregating your narrative across dozens of sources and places, each and every piece of earned coverage needs to be pulling in the same direction. That’s really critical. So, I mean, again, I think that, the bottom line here is that when AI is summarizing your story before any human being sees it or has a chance to read it, that consistency and credibility are not nice-to-haves, those are strategic imperatives.

 

DOUG: Things are changing so rapidly. So, what you thought was a source three months ago might not be the one driving. In fact, we saw some new data that for healthcare specifically, YouTube is actually driving the most discovery.

 

CAMPBELL: As I think the primary social media site that’s being cited. Yeah, that’s really interesting.

 

DOUG: Yeah. And of course, you know, Google owns them, so that’s no surprise. They’re supporting each other, but it’s really made it an imperative that we’ve got to make sure the content we do is also AI optimized. And you saw our “AI and the Newsroom” survey. Did you have any takes on that from some of the information that you saw there?

 

CAMPBELL: Yeah. I mean, to your point about the social sites that are driving traffic, that’s really interesting to me. I’m also finding that we’ve seen a lot of stats over the last year too about what types of outlets are cited most. And I think the big through line that I’ve observed and some of the data showing is that AI tends to like stuff that is that is easily digestible. So, Axios being an outlet is cited a lot because of that smart brevity model. Other outlets like Forbes, Reuters that really kinda just, like, get to the point upfront, that bottom line upfront kind of approach, but still with deeply cited and sourced reporting. So, kind of that, you know, funnel down approach of here’s the bit what you need to know, the key thing you need to know. So, another example being the Washington Post with their Gemini AI summaries integrated. Outlets that are doing stuff like that is helping, I think, drive traffic to AI systems to kind of gravitate towards using them as sources in terms of the kind of media outlets we’re seeing.

 

DOUG: Yeah. And the media is getting it now, which is because they’re focused on what can they do when their content that’s optimized. More optimized content for them actually becomes a selling point weirdly enough. And, you know, as we focused on generative engine optimization and discoverability with our rebranded satellite media tours and a new approach, Real Chemistry has been doing that in the health space, and you launched HealthGEO. Can you tell us about that?

 

CAMPBELL: So, for those unfamiliar, GEO standing for generative engine optimization, the emerging discipline of understanding and influencing how an organization’s info is surfacing in AI-driven environments. You know, for health communicators and the health sector, the stakes are a lot higher than in almost any other sector because, let’s say, when a patient asks an LLM about their symptoms or diagnosis or a health provider is double checking some treatment options for a patient or an investor is querying AI about a company’s product pipeline, the accuracy of those outputs can matter enormously. Back to, like, you know, patient safety issues. If patients are being driven to incorrect information, that is literally putting lives at risk, right? You know, GEO is how you stop being a passive subject of AI outputs to start actively managing your narrative in this new environment. And that’s the premise behind Real Chemistry’s proprietary AI search intelligence platform, HealthGEO. We purposely built HealthGEO for the health and life science sectors to measure how AI is representing a brand across all the major LLMs as well as the sources that are driving those representations. It solves a big problem for, you know, healthcare and life science companies who, you know, have some awareness but are not fully invested in how much visibility they have into how AI is representing their science, you know, across LLMs and to their audiences. And, you know, its purpose built to reflect and address the nuances of the health sector, including the inherent regulatory complexity, those patient safety considerations, and the multistakeholder environment that is healthcare in a nutshell.

 

DOUG: That is amazing. You touched a little bit on the misinformation concerns with that. You know, can AI help address that? Because what it seems is it’s really been a force multiplier to stuff. Good stuff. If you’re doing it right, your good stuff is getting out more. Now it’s about getting the good stuff out more. Try and cut back on the bad stuff. To your point, it’s very challenging to control everything. How are you going about trying to increase sort of your percentage of control?

 

CAMPBELL: It’s funny you mentioned the force multiplier term. It’s exactly how I describe it, and how I look at it, and how I use it in the day to day, but in terms of that misinformation element, I mean, I think it’s important for all health organizations to understand that they already have a presence in AI content whether they are actively and directly managing and shaping it or not. Back to that age old adage of if you’re not telling your story, someone will tell it for you, right? You know, LLMs are generating answers about their work, their impacts, their operations, and their leadership as we speak right now. And knowing that, the question we’d be answering is, are these answers accurate? You know, as I mentioned, we’ve seen cases where AI is hallucinating medication dosing information in these summaries, confusing competitor products or resurfacing outdated clinical data, which is again, that’s not just a reputational problem, but a critical patient safety issue too. And, you know, the approach that we take to help our client’s kind of address that at Real Chemistry is built around three big steps. You know, the first being to monitor what AI is saying about your brand across the major LLMs. Two, understand what sources are driving these outputs, and finally, optimize your content and comm strategies to improve what is being serviced to these AI systems because in healthcare, credibility is not optional, and any gaps in that credibility become more visible, not less.

 

DOUG: So many of the folks are trying to approach it like, hey, what are the answers out there? An area that we’re focusing on is what are the questions people are asking? And, you know, we found within our platform, it’s been great to really be able to scrape all these large language models and see what the heck people are asking. And I know one thing they’re not asking is jargon filled stuff. You touched on the simplicity. You know, how are you utilizing and what’s your thought process and the importance of really getting to what the questions are? Because you were relating earlier to things that were done in the past where it’s not changed. I always thought it was about trying to figure out, hey, what the heck are people asking or don’t even know that they’re asking about your product or service. They might not know your product, but if you know what they’re asking, you can then answer it and sort of raise your level. How important is that to what you’re seeing?

 

CAMPBELL: It’s critical. And you know, even beyond HealthGEO specifically, you know, data analytics is kind of that is the focal, you know, centerpiece of RC. It underpins everything we do. And we’re looking really carefully about not just, you know, what topics people are asking about, but how they’re framing those questions? And that’s something with our in-house tools and other resources that we’re able to look at, and that informs what we feed into HealthGEO and how our clients can actually learn and adjust in real time day to day their approach, seeing, okay, we’ve made these, you know, optimizations based on, what HealthGEO is telling us. Is it working? Do we need to tweak it? And that’s the data and analytics element there. That really does kind of inform it on an ongoing basis. So, that way, it’s not like we need this one round of updates and just hope for the best. You want to be constant monitoring, constantly adjusting because if, let’s say, there’s a big story that breaks on a news clinical study that is changing the way we think about, I don’t know, liver disease or a breast cancer, that’s going to change, obviously, what we’re seeing in SEO, but also in how people are asking AI models too, and that’s what helps HealthGEO look at what are people asking Claude, what are people asking ChatGPT, and more importantly, how are they asking about it?

 

DOUG: Yeah. And, Campbell, you’re not just driving AI literacy among your clients and, of course, within your own team, you’re a big proponent of it for communication leaders. How do you see sort of the level of AI literacy for communication leaders? And I’ll ask it in two parts. First part today, and then because it’s so chain changing so fast, what do you think of next week? How would it look different?

 

CAMPBELL: You know, it’s funny. I mean, I think that it’s easy to get caught up, particularly with social media, with these new posts and stories about all these new prompts and things you need to know about how to how to maximize Claude, how to maximize ChatGPT. And I think that for communicators or just the general public can give us an impression that everyone needs to be a bona fide expert in each and every one of these AI tools, which, worth knowing, are growing in number by the day. I think it’s less so about understanding, how to be a bona fide expert, one percent of one hundred in each and every model so much as it’s understanding for communicators how AI shapes visibility, influence, and reputation in the digital environments where the public and their audiences are increasingly living and choosing spend their time. I think there’s three questions I would encourage. I try to encourage my team, my colleagues, as well as just any communicators, you know, to be asking themselves and their teams on an ongoing basis in terms of this. You know, how is AI representing our brand today? Are our comm strategies accounting for AI mediated discovery? And, you know, are we using AI in ways that sharpen our judgment and leverage our team’s unique expertise or are replacing it? That’s not an exercise one time. I try and ask myself anytime I’m using AI and counseling my clients on AI. And I think the most effective comms leaders are not asking where we can deploy AI, but where does AI improve our decisions, and where is it actually and potentially introducing risk.

 

DOUG: Thanks so much for participating, Campbell.

 

CAMPBELL: Thank you so much for having me.