Introduction:
Welcome to the first episode of In Practice, where Jason Pistulka and Deven Bhagwandin explore how AI-driven search is reshaping visibility in healthcare recruiting, why traditional SEO and paid strategies fall short, and what organizations need to do to ensure they show up when it matters most.
Jason Pistulka
Welcome everyone and thank you for joining us today. We have an exciting topic to talk about. AI Search in Healthcare Recruiting and Why Credibility Now Determines Visibility. This podcast is sponsored by Joveo. Joveo manages and optimizes the digital recruiting funnel for employers from job distribution, media spend, career sites, and apply flows to improve the applicant quality, efficiency and hiring outcomes.
So this is an entirely new series that we’re doing dedicated to healthcare. Each episode’s gonna explore some of the real challenges that health systems face when trying to attract and retain talent. From workforce shortages to fragmented brand architecture and to rising talent acquisition costs. Some episodes will focus on technology, others on operational excellence, and some on strategy. But every conversation’s gonna center on one big thing: how healthcare organizations can build sustainable recruiting advantages. So in today’s episode, I’m particularly excited because we’re talking about how AI-driven search systems are changing the way healthcare organizations are evaluated and surface to candidates. Because, increasingly, candidates are not just searching Google or job boards, they’re starting to ask AI the questions, and they’re taking their answers that they receive as being facts.
So before we get into that discussion, let me do a few introductions. First, I’m Jason Pistolica. I run a strategic consulting firm called StratTech Talent Consulting and Advisory. And Joveo happens to be one of my clients. My background is in enterprise recruiting, HR technology and architecture, and advising executive teams how to align recruiting operations with business performance.
I’ve spent a number of years inside of large health systems, working alongside leaders, navigating labor shortages, and dealing with operational fragmentation and rising costs. So today we’re gonna unpack one of the newest variables influencing all of that. And that’s where my friend Deven comes in. Deven, would like to introduce yourself?
Deven Bhagwandin
Yeah, hey, thanks everyone. I’m Devin, the founder of PenPixel Creative and we focus on how organizations show up in AI search. Right now, I’d say 99 % of the companies on the internet still think of search as traditional SEO, but AI systems don’t rank pages. They actually look for entities they can trust.
So I work with organizations to ensure that their entity is structured correctly, that their data is structured correctly, and that all their brand signals are structured so that AI can clearly identify them as an authority that they can trust when they’re citing or when they’re looking for information to give in generated answers. So unfortunately, pretty much every business has this challenge so I’m really happy today to explore through the eyes of the healthcare system.
Jason Pistulka
Yeah, today, you know, more than ever, search is not just driven by search engines, but, you know, AI platforms and AI systems are acting as those reputational filters. I know I primarily seem to be searching and finding information now through AI. And they’re deciding a lot of really important things for us. They’re deciding what’s credible, what things are actually connected and verifiable, and what gets surfaced.
Ultimately to your candidates in the world of recruiting. And in healthcare organizations, especially those with multiple facilities and complex brand structures, which is so common, they’re really uniquely exposed in this. So it’s not just a marketing conversation. It’s really about reputational data infrastructure conversations. With that, Deven, let’s dig in a little bit about how AI actually decides what is true out there in the world when giving recommendations to people.
Deven Bhagwandin
Okay, so before we do that, I have a quick question for you. I really want to know, when a health system needs more nurses or physicians, what do they do?
Jason Pistulka
Yeah, unfortunately, typically the number one thing they do is spend. We start spending more money on advertising. We start to increase our job distribution to different places. You we’re really buying visibility. We may engage our sourcing teams to start doing outreach to people and hopefully they know who we are. But in traditional search, you you just would pay your way towards the top if you need more candidates.
You turn up the budget, the problem with that is it’s a hockey stick effect. It gets exponentially more expensive as you turn it up.
Deven Bhagwandin
Right. And so if you’re in a healthcare system that has, I guess, different brands, how do you approach that?
Jason Pistulka
Yeah, it’s, it’s healthcare is so different. I didn’t know how really, how different it was until I entered it because most of them are not single brand entities. Most of the larger healthcare systems out there were brought together by acquisition. And so they have multiple hospitals, maybe even multiple subsystems within their hospitals. And that means lots of different facility names.
You have some legacy brands that may be really, really strong and other brands that aren’t as strong. And if you have a really strong legacy brand, you typically don’t want to wipe it away. You don’t want to become the Microsoft or the Amazon or the Walmart because these brands have been built locally. And so it gives us an amazing amount of brand complexity. And then sometimes you even have academic affiliations or specialty facilities you know, that you’re dealing with. So you may have a flagship academic center or medical center. Then you have rural hospitals tied to that, children’s hospitals, surgery centers, physicians’ offices, urgent cares that all have, many cases, their own brand complexity. They’re owned by one parent system, but to the public, they don’t necessarily know that. And to AI, it may not be able to just connect those dots easily just by looking at a bunch of websites.
The fragmentation becomes a real problem mostly when you’re dealing with AI search.
Deven Bhagwandin
Right, so I think it’s safe to say that in traditional search you could kind of spend your weight higher, but in AI search you can’t buy a better reputation, right? And that’s what’s most important to the candidates and to, I guess to patients also I would imagine. So.
Jason Pistulka
Absolutely.
Jason Pistulka
Mm-hmm.
Deven Bhagwandin
So let me dig into or let me detail what AI actually, what it’s looking for when it’s ready to provide true accurate answers. So when people think about search these days, we spent the last 20 years talking about SEO. So naturally they’re thinking about SEO.
But everyone that’s ever, anyone that’s Googled anything over the past year, you’ll notice that AI doesn’t rank pages. You just get that one answer at the top of the page now, right? And to get that answer, AI needs information that it can verify and trust. It doesn’t care about keywords. And that’s why you’re seeing those rankings being pushed farther and farther down the page. So how do you, where does it go to find this information?
Well, it goes to third party Wikidata entries. It looks through your JSON schema, which is markup that’s on your website. And it’ll look through the Google Knowledge Graph because that has some really good trustworthy verifiable information. And AI assigns the highest confidence to these.
Examples of untrustworthy sources are major news outlets or things that you can self-publish like press releases, blogs about pages, because AI knows that you control these pages, right? So it can’t corroborate that information unless it sees it somewhere else. And that knowledge graph, the Wikidata, all these other places is where it really wants to see that information.
Jason Pistulka
So if a hospital wins a magnet recognition, but it’s not structured in Wikidata properly or in your knowledge graph properly, AI doesn’t treat it the same?
Deven Bhagwandin
Yeah, that’s, you’re right. It does not treat it the same. Because what we want to remember here is that structure equals trust. And the structure comes from seeing it on your site, but also in other places, exactly as you have it on your site.
Jason Pistulka
So that consistency matters. So what you’re saying is AI doesn’t reward what you say about yourself. It rewards what the knowledge graph out on the internet it can verify about you.
So excellent, so before we move to a specific case study, I wanna address something critical. You were telling me that when you’re testing AI visibility, and this is one of the first things I started doing when you and I started talking, I started going out and searching for myself and my own organization out there, you said you need to be careful about how you test because the version you test with matters. Tell me more about what you actually meant by that.
Deven Bhagwandin
So this is something that a lot of, pretty much everyone forgets to do. Everyone in marketing forgets to do. You need to…
You need to test where your audience is, not where you’re working, right? So if you’re using a paid, if you have an LLM with a paid subscription and you ask it a question, you may get a really full detailed answer, but you’ve got to remember that I would imagine most of the, most candidates who are searching for a job, they’re probably using a free version.
And so the problem with that is the paid models, they use retrieval augmented generation, which is what we call reg. It means it can browse a fully updated version of the internet to retrieve its information. But the paid models, they are cut off at a certain point. So they may be using, they’re answering your question with data that could be a year, sometimes two years old. So the answer you get in a free model is not as accurate as it is in a paid model.
Jason Pistulka
Excellent. So, recruiting and recruiting that really matters because you need to optimize for what candidates actually use, not what their executives have access to.
Deven Bhagwandin
Right, so like I was saying earlier, you want to, you need to test your AI visibility using the tools that you think your candidates are gonna be using.
Jason Pistulka
So if your AI visibility changes between free and paid models, your credibility infrastructure just isn’t stable. So, let’s make a concrete example with this. And so I went back and I started doing some research with an HCA acquisition. And HCA had purchased Mission Health in Asheville, North Carolina a number of years ago. And on the surface, it’s very reputable, very well known hospital in its area. But I started examining the structured data and that’s where we start to see gaps emerge.
Can you tell me more about that?
Deven Bhagwandin
Right, so as we went through this together.
We could see that AI clearly understands it’s a hospital. It can tell it has a website. It knows where it’s located, what year the hospital was started. And it can tell that it’s got some other entities associated with it. For in this instance, it was Mission’s Children’s Hospital, but it was missing that link to HCA Healthcare. It missed any awards that Mission Hospital has received and it had complete modeling. So what ends up happening is that Mission isn’t associated with HCA, which is a big association that you want to make because candidates are going to want to know they’re working for this really reputable hospital. I would imagine patients again would want to know that they’re going to a really reputable hospital. It’s missing mentions that it received from the US News and World Report. I think it was on a best hospital rank and it’s missing that so again folks will they find it trustworthy or not we don’t know because it just has incorrect information in that AI generated answer and that as a brand is what you want to avoid. You want that answer to be as accurate as possible so that people have, it essentially leaves them no choice but to go over to your site and visit you or contact you, whatever your preference is.
Jason Pistulka
So this is really interesting. And this varies a lot from health organization to organization. I mean, at least Mission had a WCA data entry. I mean, we see a lot of them that don’t have a WCA data entry at all. And Mission was a strong brand before HCA bought them. And so not having the connection to HCA may not be as big of a deal as it is for an unbranded hospital or a lesser known hospital.
But you know, the fact that Mission’s a part of HCA and HCA is massive, the authority is not transferring structurally to Mission and it’s probably not transferring structurally to many of those other hospitals that an HCA would own that are lesser brands that could really benefit from that authority of HCA behind it.
Deven Bhagwandin
Right, and so what happens is because that association isn’t there, Mission or the smaller hospitals, they’re not connected to these bigger brands. And so they’re essentially unseen. They’re unseen to the AI, which means as the time goes, and as AI answers more and more of the questions and search rankings disappear, if you’re not appearing in that AI generated answer, your entire business or organization is invisible.
Jason Pistulka
So I think the takeaway here then is if your company awards and accolades are not a part of what AI can find and verify, and if your association between your different entities are not part of what it can verify, then they just don’t exist to AI. So let’s connect this to recruiting, because candidates increasingly ask AI questions.
Even more so when a candidate is a new college grad or they’re looking to relocate, where they really don’t know the lay of the land or what the hospitals are in this area, they may be asking, is this hospital a good place to work? You know, who is this hospital owned by? Is it highly ranked? Is it a system that is financially stable? And there’s a lot of financial instability that exists out there.
If AI is giving partial answers, candidates are assuming that that’s the full truth. And the sad part of it is, if your competitors have optimized in there, the partial answers that they’re getting are probably putting your competitors ahead of you. And so they don’t click any deeper. They’re just taking the answer that AI is telling them as being the fact. They’re assuming that the artificial intelligence is truly smart. It’s truly intelligent.
So those credibility gaps really are becoming recruiting gaps. And this is an issue in healthcare probably more than many other places because of the naming commonality. There’s approximately 20 hospitals that I could find in the US with Good Samaritan in their name. Mercy, there’s over 100 unique hospitals across multiple systems. Then we have the East Sides, the West Sides, the North Sides – all of these things that are causing entity ambiguity. And so if I don’t have the structured data that ties the location of my entity to my name of my entity, to my parent, to my awards, my award that I got at my hospital, Good Samaritan, may very much be attached to a competitor Good Samaritan in a whole different location. So…
Jason Pistulka
I really need to make sure that that structured data is there or it doesn’t really represent me correctly.
Deven Bhagwandin
You need to make sure it’s there and it’s verified. So I want to give a really quick stat too, to back that up, back up everything that you were just saying. Just today, I saw a stat that said for every 1000 Google searches in the US, only about 360 clicks ever make it to the open web. The rest are answered by AI before the candidate ever sees your apply button. So that means if you need applicants and you’re not appearing in that AI generated answer, you are not going to get those applicants.
Jason Pistulka
So that’s interesting. So what you’re saying is if I’m sitting here in Nashville and I’m looking to relocate to Nashville and I do a search for the best hospitals to work at in Nashville and Vanderbilt comes up, they’re likely just to go directly to Vanderbilt’s website and look up jobs. They’re not going to click further and dig down deeper in the framing.
Deven Bhagwandin
I mean, I hate to be blunt about this, but wouldn’t you? If the answer is yes, do you?
Jason Pistulka
You’re absolutely right. At least I give first shot to what it gave me, right? That’s where I go first and I test that premise before I move on and look for others. And a lot of people aren’t going to test it a second time. They’re going to go push that application out to that.
Deven Bhagwandin
Right.
Jason Pistulka
To that particular answer. So this becomes really critical. And this is where I think Joveo’s mission really intersects with this conversation because Joveo exists to make recruitment more efficient by connecting people in the right jobs and reducing unnecessary media spend, which is exactly what we’re talking about doing here is any time we can get people to come directly to our site and not go to an Indeed posting or anything else where we’re paying for clicks or we’re paying for applications is a good thing. And Joveo also improves that candidate journey quality. So when the candidate hits that website, they get a great experience and a great application. And it really helps the employers to build a sustainable visibility. But recruitment marketing efficiency can only scale when credibility is strong.
So if your AI presence improves, we’re gonna have more of those candidates arrive at our website organically. Now think what that means too. When they go and they apply on a Indeed website or anybody else’s website, they’re harvesting that data and they’re using it to funnel those candidates to other sites. We’re competing against other people in, or other companies in that ecosystem.
But when I drive them organically to my site from AI search, I’m not competing with everybody else’s ecosystem. I now have them in mine. I can build my brand trust and I can really get my message across to them as to why you’d want to work for us and improve the quality of that conversation and ultimately reduce paid dependency. And that’s the interesting thing with Joveo is that here the whole thing is, how can I get you better people, better qualified people, paying less money? So that AI credibility really becomes a force multiplier. So the efficiency in recruiting really starts with the credibility in AI.
Deven Bhagwandin
Right, right. So here’s what healthcare systems should do. Number one, should audit their, audit your technical SEO, your content, and your visual branding across the internet to make sure that it’s all clear and points back to your brand, your website.
Number two, audit your structured data and make sure that you, in that, you structure your parent-child organizational data correctly, right? So if you have, if you’re HCA, you need to let humans and AI know that mission is part of your healthcare system and vice versa. Mission should point back to HCA too so that that relationship is clear to AI. And you can do this in your structured data.
Another thing you can do is publish authoritative journals and cross-reference that across the internet. Again, that goes into your structured data. AI can see it. Humans start to see you in both places and associate you visually and mentally with each other. And all of that starts to bring your authority and trust up. It’s essentially branding if you want to boil it all the way down, but this is the new way of doing it for AI search. By the way, I actually offer these audits too if anyone’s interested.
Jason Pistulka
Well, and the other interesting thing that you’re talking about cross-referencing and you’re talking about auditing, the importance too of making sure that you have consistency in brand and message across sites really matters too, right? AI does not like when it sees differing information about a source, it starts to create distrust, correct?
Deven Bhagwandin
Right, so I think one of the best ways to give an example of this is to really think about your logo. If you have a logo here on your website, AI can read the file names and it can actually essentially see that logo and then say your mention on a best of best hospitals in a city list and your logo there doesn’t match the logo on your site. Just like it just like a human would AI will start to get confused and wonder whether this is one is actually the other right and so it may actually skip over that information where you’re on that best of site or best of list and decide you’re not trustworthy enough to add to the generated answer. It’s going to skip over you and go to someone else where it can verify the signals.
Jason Pistulka
And so this is not just about looking pretty marketing fluff, right? This is about infrastructure and understanding what all the pieces of infrastructure are across the web and which ones are the most prominently used, you know, in AI and in search as authorities or second check verifications, right? Because part of what AI is also trying to do is keep fraudulent stuff out and things like that. And when it starts seeing different citations or different, differing information about your organization, it starts to question is this thing really real? So, you know, companies are really reputation dependent, people dependent, and this is so critical, mostly when it involves things like life-saving care from health systems. I mean, people want, when they’re researching on getting away from just the recruiting, right? When they’re researching on best quality of healthcare delivery. Like when we think about if you have a loved one who is in need of cancer care or heart care or anything else, you want to find the absolute best. And so making sure that this stuff is right so your organization represents in the right way becomes great because if AI confidence weakens your reputation signal, then it impacts not only your recruiting, but also your patient attraction, making it more difficult and more expensive.
Deven Bhagwandin
I totally agree. While you were saying that, one thing that kept popping into my head for someone who’s not in healthcare, I kept imagining that say it’s a rainy night and I’m driving and I’m just looking for the closest hospital because my child is sick.
And then out of the blue, you see this bright sign with the hospital, the logo and name that you recognize. And you know in your head, I know those guys. I trust them. I’ve seen their commercials. I’ve read about them. I hear about them on these other, on these best of lists and so on. Can you imagine how that feels to a person where they’re like, finally, I’m going to get the help I need. And so you want that feeling to come across in those AI generated answers also.
Jason Pistulka
Well, and I think it’s increasingly important when you think of the elective things, right? Because there’s obviously, you know, hey, I’m sick, I need to go to urgent care, whatever that may be, that’s a little bit more transactional in nature. But when you’re talking about elective things, I need major back surgery, I need cancer care, et cetera, people have a little time to search, they have a little time to check things out, and they’re going to go deeper.
Conclusion
And so AI search becomes more and more important. It’s not about gaming rankings and it’s not about we’re just gonna buy our way to the top. It’s really about building something that’s verifiable, incredible. And that needs to be very intentional and you need to take the time for it. Like when you start talking about being in editorially reviewed journals and things like that, some of those things, you really have to think about as an organization building to understand what your AI presence is gonna look like. And in healthcare recruiting teams, if they have that structure, their authority will win. So those who rely solely on job advertising are gonna struggle, they’re gonna be left behind. So I find this whole conversation so interesting, it’s the latest game changer in the recruiting advertising world.
And I really wanna thank Joveo for sponsoring this discussion. And this is a constantly evolving area as the models evolve. So it’s not like once and done. This is a constant update that we have to be doing in our recruitment marketing worlds. And so if anyone’s in need of assistance with recruitment, operational efficiency, system design, anything in that space, feel free to reach out to one of us.
We’re happy to connect with you, both Deven and I. And Deven, this fascinating conversation, I really enjoyed. I look forward to our future conversations. So thank you so much for joining us.
Deven Bhagwandin
Thank you. Thanks for having me.
Jason Pistulka
All right, thank you everyone.
