How to Get AI to Recommend and Cite Your Brand

Stewart Gandolf (Healthcare Success): Hello everyone welcome to the Healthcare Success podcast today is going to be a fun one. First of all I want to introduce Duane Forrester. Hi Duane. Hey and also Brandon Schakola, who is head of our digital department in Healthcare Success. HI again, Brandon. Haven’t seen you for, what, two minutes before? So anyway, I’m happy to have you guys both here.

So today I’m super excited about. I know it’s going to be fun because we’ve been talking about guitars offline and SEO geek stuff. It should be a lot of fun as we get into this. This actual podcast was precipitated by a discussion internally where Duane’s book, The Machine Layer, there’s the plug. One plug.

Duane Forrester (duaneforrester.com): Thanks, sir.

Stewart Gandolf (Healthcare Success): …with something we were comparing notes about and we love this stuff. And we have a lot of SEO geeks at Healthcare Success. So we have a mutual admiration society, I think. And it turns out, like, a fair number of my team know Duane, either very well or at least tangentially. So we thought we’d invite him. And he said, yes, while we were on the call, by the way. We were talking about Duane, and he agreed to do it while we’re still talking about it. So super happy to have you. Duane, we’re going to ask you a bunch of questions, and then I’ll be adding some color commentary as well as Brandon.

First of all, I know that you’re—give a little bit of background on you, your book, what it’s for, maybe hint at the secret super uber secret without giving away the secret project you’re working on. Tell our listeners a little bit about you.

Duane Forrester (duaneforrester.com): Yeah. So for listeners who aren’t familiar, I’ve been in the industry a while. I originally built and launched Webmaster Tools. I was one of the people who launched schema.org, spent a decade at Microsoft, ran Bruce Clay for a while after that, so I have some direct agency experience, ended up at Yext prior to the IPO, worked with them for seven and a half years, and for about a year and a half now, I’ve just been out on my own.

And during that time, obviously, AI has been a big part of everything we talk about. And answer platform, answer engine, all of these things. I’m not worried too much on what we’re calling these things, but it did occur to me that we’re probably not leaning in quite enough as an industry. And at least projecting my own self two years ago at that time I knew I wasn’t.

And so then I thought, okay, well, I have to lean in more. I have to embrace this. I have to learn from it. I have to find ways to exploit it because SEO, It’s hardwired, genetic, I guess.

And that kind of started the path to the book. Now, I’ve already written two books in my career with McGraw Hill as my publisher. And so the idea of writing a book is very straightforward to me. There’s a process to it. I know what that is. And you can go ahead and do it. And so I wasn’t really worried about the writing of the book. I was more worried about the positioning and the “why” behind the positioning.

Because you can take any topic and you can apply the process and come up with, in this case, it was like 65,000 words, and you’ve got a business book. There it is. Whether it’s a book that addresses a need or is utilitarian enough to be valued are very different things.

When you work with a publisher, they will give you guidance. They will tell you, “We want it at this time. This is what we’re looking for.” The envelope is defined for you.

So for this book, I decided I would self-publish. And it was a lot more work, but I’m grateful for the things I’ve learned, and you guys see the product that we now have. So this product was ultimately, as I learned about all of these things, this book was a way to go from fearing something to embracing it, essentially. And that arc is a core theme throughout the book, okay? The idea of, “I fear that AI will negatively impact my career” to “here are ways to use it to positively grow your career.”

So we go from that kind of like, “I’m alone; I’m in the dark; I’m really not sure here how this is going to impact me” to, “I have frameworks, I have ideas, I have a way to learn, I have processes, I have information,” so you, the SEO, who have been that expert that everyone in your company has come to because you know this dark art of SEO, you remain that same person within this new world of LLMs and AI and so on.

So that was really what the book kind of came down to for me. And then throughout all of that, you learn all these different things that matter. And it occurred to me, the metrics we were tracking weren’t the right ones. And that trust has a different level to play in this because in traditional search results, and this has been tried and tested in the courts as well, the engine is not responsible for your behavior as a consumer after you click on a result they gave you.

So, list of links. I, as the consumer, decide what I’m going to click with, what I’m going to interact with, and then my behavior after that is entirely up to me. Contrast that with something like an LLM that gathers the world’s information on a topic. You ask your question, and then the information that you get back is the synthesis of all of the world’s data applied against your question, essentially creating a unique answer in that instance.

And if that answer leads to a behavior that’s detrimental, does the business itself hold liability for your behavior? It’s a vastly different world, and that’s being tested right now in ways. Open AI has some open litigation. And that’s a very real thing today.

So this concept of trust and what it means is quite a bit more nuanced than what we were used to with E-E-A-T, right? That is foundational, but there’s a lot more at play here. And it starts to play into “Why is my content that ranks well not getting chosen for an answer?”

Stewart Gandolf (Healthcare Success): So that’s intriguing. And Brandon will probably throw a few comments a little bit later about, number one, that we for sure are embracing AI for our own processes, right? Not only in terms of the, because we’re able to do so much more now to benefit our clients. It’s not something we’re fearing we’re leaning into it.

And I don’t know, Brandon, do you have a quick comment on that because that’s just like where we live.

Brandon Schakola (Healthcare Success): Yeah, and I think, you know—picking up on on what Duane was just saying in the health care industry—we’re rendering a Your Money or Your Life industry, and the chances of hallucination are so incredibly high and it’s also  one of the areas that got impacted the most when AI Overviews started rolling out and we were told to put glue in our pizzas because that’s what you do.

I imagine if that’s some of like surgical procedure or a life crisis moment for someone. And I think you’ve probably seen this yourself, just navigating that risk layer, which is now much more present in our lives, along with this existential crisis us SEOs have been all going through over the past two years or more, it’s caused a major shift in our thinking as to how we’re approaching these problems, right?

How can we steer these large language models to have less hallucinations when we need to? And I think that’s a critical question for us to answer.

Duane Forrester (duaneforrester.com): I was going to say, I talk a little bit about that in the book, and I talk about it obviously on my Substack, but it is a critical thing, Brandon.

The companies themselves are actively working on this. You can dial an LLM down to virtually know of you hallucinations, right? Like you can do that. Unfortunately, what it means is you will immediately lose personality. So if that’s why you’re using one of these platforms, that would go away.

It’s not why I use the platforms, but it’s a part of that experience that gets dialed out when you dial down hallucinations. I think the most important way to dial back the hallucinations is to narrow the scope of the LLM, which you may be thinking, “Oh, okay, well, if you give me a healthcare LLM, that’s a narrowed scope. You really need to understand the concepts of scale here, right? Scale way beyond what you’re thinking of, so dealing with a billion active users a day is what OpenAI is facing right now.

And then the other version of scale, which is the tiny, minute version of scale, which is—it’s not healthcare. That is not small enough to limit hallucinations because you can give it a lot of information within health care and it will just pick and choose and say, “Here you go,” like you have a toenail on your forehead. And you’re just like, “No, I don’t.” And there it is.

But if you actually take it all the way down into the niche where we’re only talking about bone marrow and the only information that it can feed on is about bone marrow and specifically related to the answer to the question you’re asking, you will not get hallucinations. It will come back to you reliably.

But you can also understand how we’ve split the hair so finely that you can now no longer use that LLM in any other circumstance. So you’ve taken ChatGPT, essentially, and said “You can answer one question and one question only” and it will be perfect, but if you attempt anything else it’s just a mess. It’s like a moose on a skating rink; it’s just everywhere.

Stewart Gandolf (Healthcare Success): That visual on my mind that’s pretty good. I’ve never heard…

Duane Forrester (duaneforrester.com): You’re welcome.

Stewart Gandolf (Healthcare Success): I’ve never heard that metaphor before.

I think that before I jump into the next thing, I just want to add on our comment too. So you mentioned earlier, Duane, about, you know, it’s pulling from everywhere, right, and so that’s the challenge for us. And when we talk to clients and prospective clients a lot, I always talk about, “Okay well it used to be a technical SEO mattered—it matters now, but in a different way—and content mattered. It matters now in a different way, but then local and then reputation and then breaking it down we spend a lot of time talking about PR or digital PR and brand, so it’s kind of like you have to be everywhere and now do all this stuff differently and it’s like, it’s infinite.

Other than that, it’s an easy challenge to do.

Duane Forrester (duaneforrester.com): Look, I’ll tell you this. You guys have been around long enough to remember Y2K. For listeners unfamiliar with it, you should go do a little bit of dredging around on that particular horror that is in 1999, that turned out to be a large nothing burger.

I’m not going to say there was no work involved. There was, but it really wasn’t the hype that we thought it was going to be. And I suspect that a lot of the, “I fear AI is coming for my SEO job” is going to fall into that bucket over the next couple of years.

Checklist SEOs are going to have a hard time because they’re going to be given new checklists. Those checklists are going to require actual hands-on work based on knowledge and experience. And if you don’t have that, you’re going to fall into a hole.

However, I think for practitioners who have been paying attention and who long ago got the memo of quality and trustworthiness and you dot all the I’s, cross all the T’s, make sure you’re using commas where you’re supposed to, that kind of pedantic approach to everything, these people, these approaches are going to stand the test of time here because those systems need that information.

There’s all kinds of conversation out there when you get into it about the LLMs are now training on synthetic data because they’ve gone through humanity’s data and when you don’t produce information fast enough to fill their needs.

That kind of doesn’t matter. That’s the difference between a regular typewriter and I saw a video yesterday. Somebody created, they took an electric typewriter and created a Claude interface with it so they could type their question on the typewriter and then Claude would type the answer back on the typewriter to them. And I’m like, wow, and no.

Brandon Schakola (Healthcare Success): It’s about as ridiculous as this whole Claudebot, Moltbot thing.

Duane Forrester (duaneforrester.com): Right, exactly.

Brandon Schakola (Healthcare Success): It’s bonkers.

Duane Forrester (duaneforrester.com): Yes. Someone asked me about it, and I was like, “not worth the oxygen that we’re using to discuss it, because it doesn’t help anyone solve a business problem.” The problem with the typewriter is you’re still working with a single sheet of paper. So, you know, one question, and you’re done. Like, you run out of paper, and that’s it. And I’m like, “Okay, well, it’s cool that you did that. Like, I, you know, points to you, right? Like, I’d love to come over for dinner one night and look at this thing.”  But that doesn’t change anything.

Stewart Gandolf (Healthcare Success): I want to jump in and interrupt on that one because you just keyed into the next question that I wanted to ask. So thinking about all the hype in AI and all these different things we’re talking about—which are fun—but, in plain language, what is the change that the business leaders need to really understand? What is the big picture to help bring us to that?

Duane Forrester (duaneforrester.com): So first and foremost, there are new metrics that need to exist. I’ve outlined them a bunch of times. You guys have read them in the book. I’ve got a Substack about them as well. And then, frankly, if people want to use those metrics, that’s great. If they come up with their own, that’s fine.

The point here is you need to look beyond traditional search metrics in order to explain behavior and look at what’s happening. And even the highest performers, how you perform in an LLM answer environment is vastly different than what you’re used to.

So for one thing, I think if you’re a CEO, you’d better wrap your head around this idea that it’s not rational to go and ask ChatGPT and expect to see your brand as part of that answer. And right now, CEOs expect that when they go to Google or Bing and they type in a question or something related to their business.

And historically, over the last 20 plus years, it’s been stalk down the hall, yell at the SEO, “I just entered this query. Why don’t we rank number one? That’s your job. get it done,” and CEO leaves and SEO gets to work and so on. Some version of that.

Stewart Gandolf (Healthcare Success): I’ve never had that happen before.

Duane Forrester (duaneforrester.com): But that’s just not the way it operates today because if you do want to be a part of it, you look at the CEO in that situation you say—I would recommend calm—and “let’s walk you through how that builds its answer.” And you go through that cycle and explain it.

And then they realize, “Oh, hang on a second. I have to act as if my company was an individual. And if I am that individual, how do I appear to the world? Am I trustworthy? Do I answer all questions fully? Have I covered everything? Do I appear to be the expert that I’m telling everyone that I am?” All of this level of trust.

So if you thought E -E -A -T was cumbersome, I just published a Substack where I introduced SSIT. And it’s just an acronym. I made it up. People should take these as jumping off points and go use this to creatively figure out how things apply to your instance. But it’s a different layer of trust, and you really have to be able to prove it.

And then that takes us into, how does your proof get consumed by an LLM? And if people aren’t familiar with things like common crawl.org, they should probably wrap their heads around this and they should probably be looking to that it’s not a silver bullet but it is an indicator, and if you think of archive.org, you think of the Wayback Machine—this kind of idea common crawl is literally, that it’s a snapshot of the internet designed to be handed to services that need that snapshot that use that information.

And all the major LLMs use their data set as part of their training material for the LLM. So if you’re not in that data set, and you guys have all had this conversation with Claude or with ChatGPT, you ask a question and it’s like, boom, here’s your answer. And you’re looking at it. And you’re like, “it’s not quite right.” Yes, but I feel like that’s five years old. I thought it was new. I think I remember something fresher.

And then you go back and you ask him, goes, “Oh, let me do a live web search.” And then it comes up with completely different information related to the answer. And you’re like, “Oh, that’s better.” It’s immediate tuning, if you will, it’s designed to cut down on spinning CPUs.

So instead of doing the live work, which is more computationally intensive, it gets lazy by programming and definition and goes, “Oah, let me check the encyclopedia I’ve got here.” Well, the encyclopedia is from 1978. So it’s easy to see that there are new countries since then. There are different things that are happening.

It’s important to be in the training data, but you have to understand why you’re in the training data. I’m looking at data right now for a huge brand-name company, and I’ve got a pipeline plugged into Common Crawl, and it’s coming back with their information. And the information that I’m seeing is that they were last included in Common Crawl under this URL in 2014.

So the very beginning of Common Crawl was when this brand was last gone. And I know exactly what the problem is. A few years ago, they put a redirect on their trailing slash to ENUS. And now everything exists under the ENUS. So when the crawler shows up, it punted over. It ingests all of that. And if you go in and ask the system for the trailing slash brand domain, it says, “The only stuff I have is from 2014.”

So welcome to SEO in 2026, the stuff you didn’t know that could be impacting you, right? Like, it’s crazy, crazy complicated now.

Brandon Schakola (Healthcare Success): Yeah, I’ve been saying for a while, the shift in the mindset for technical SEO is, it still matters, but that also it matters in a much broader sense because you have to treat brands as a data set.

Duane Forrester (duaneforrester.com): Yeah, listen, and it’s this whole, like, Brandon, I saw your head nodded when I was saying, like, “if your business is a person, then what are you like as that person,” right? I mean, that’s not a new analogy, but it’s very apt in this situation because you’re dealing with a synthetic thinking system that to some people appears to act like a person and it’s judging you in a similar way.

And so that nuance on trust, like, these platforms, they will just skip you because, again, they have the world’s data at their fingertips. So coming up with the answer isn’t where they stumble, you know? Like, it’s out there.

If you’ve ever had that moment where you thought, “surely this product exists on the internet, I just have to go find it,” and you just know that it exists, even though you’re struggling to find it, that’s LLMs now. I guarantee you the answer is out there. They have everything they need for it. The struggle is how do you get them to say, “Oh, I like Stewart. I like Brandon. We’ll include them. They’re good guys.”

Stewart Gandolf (Healthcare Success): I’m glad somebody does, Duane. So, Duane, I’m going to ask you, like, you just led me into, I can’t. I’m on the edge of my seat. Like, I can’t wait to ask him about this.

So you mentioned the two gates in your book. I think you just led right into that. Like, explain the two gates. I think that’s relevant to what we’re doing here.

Duane Forrester (duaneforrester.com): So look, these two gates, and I’m going to refine this, okay? Because from the time that I wrote this to now has been measurable progress by the platforms, okay? So we’ll define the two gates as they are in the book right now, which is machine-validated authority.

And then the second one is called a universal verifier. So the machine-validated authority, I want you to think of it as kind of like all the work you’re doing in E-E-A-T, and it’s going through and it’s checking off the boxes. And it’s saying, “You’re a member of the local chamber of commerce, You’re a member of this. You’re a member of that. You’ve got this certificate. I see your educational component and blah, blah, blah, blah,” all the way down, whatever the list is that it has.

It’s looking for those overt things that it can cling to and say, “Ah, I trust Stewart. This looks like a footprint of trust. Okay? There it is.”

You know, Brandon and I might look at you and say, “He has a trustworthy face.” Whatever that means, the AI is doing a similar thing with your footprint and saying “This looks trustworthy; therefore things produced by him should be given a second look,” because if you failed that first machine vetting of you don’t talk about your background, you don’t talk about your education, you don’t have any certificates—all of those things, forget it You’re just not included in the answer.

You failed the first gate. You don’t move forward, right? Like, don’t pass go. Your game of Monopoly is ended for you. You’re out of cash.

That universal verifier, though, that’s something the systems are working on. We talked earlier about you can dial down hallucinations. The whole point behind universal verifier is to stop the answer before it hits any one of us and ask the question, “Hey, is this relevant? Is it safe? Is it accurate?”

So the universal verifier, this is kind of corny to put it this way, but it’s an LLM, fact checking in LLM. So, you know, everybody have a laugh there for a moment and insert your jokes here. But its only job is as a police officer. It has no other job. And so its focus is such that its ability to hallucinate is vastly mitigated, like extremely low.

And again, I have a Substack on this. There’s links in there to work that I know Google’s working on, and the major platforms are working on that talk about this concept of a universal verifier. Only one group uses that phrase was handy for the book and it’s easy for people to wrap their head around.

What we’re starting to see, though, is we’re starting to see portions of that concept get embedded into new models of Claude and of ChatGPT. And I’m pretty sure Google’s doing their part with GeminI and so on. So this concept of you show up and you’re tested at one level, it’s like taking a 101 level class, right?

Like, okay, well, you had to go through high school. Now you’re in university and you’re at 101. Great, you’re really smart; that’s awesome. You know a bunch of this stuff. That’s great. If I gave you the 201 test right now you’d fail, but you’re at 101 so go through all of that you pass that test now you can start learning 201 and then when you go through all of that you can take that test and then maybe graduate to the Ph.D. level. That’s really what we’re going through with all of this applied to trust and trust signals

So those two gates are a big deal because they are the, you’re gone and you just don’t matter. You’re dust on the floor or, “Hey, that sounded great, but we don’t need it for this take of the song. So we’re going to cut that out of the tape and you’re not in the final version of the song.”

Stewart Gandolf (Healthcare Success): That’s a great metaphor from a guitar maker for sure.

Duane Forrester (duaneforrester.com): Right?

Stewart Gandolf (Healthcare Success): Where do most businesses fail at this? How do they even know where they’re failing on the gates?

Duane Forrester (duaneforrester.com): Sadly, most of them fail on the first level. And this is where I think that it’s really important to understand the value of experienced practitioners. Because if you’re approaching SEO as like a checklist-style thing where you just sit down and you say, “you haven’t done this, go do this, go do this, go do this,” you collect your money every month from your client and off you go.

Or that’s how you approach things in your company where you go tell a bunch of people, “here are all the things you need to get done” and you wait for them to complete their work.

This is really not a good way forward today because it’s almost like what I need you to do is—if anyone has ever had a dog or a cat in their life or a pet, you’ll understand the concept of zoomies where they just go like crazy. They’re darting around and they’re all over the place, like a pinball in a pinball machine going nuts. And in my house, we call that exploding, where you see your dog and the dog wants to be everywhere all at once.

That is now your existence as an SEO. You have to be everywhere all at once. You have to touch everything. You have to know all of it.

There is literally a quote from Satya Nadella. I worked with him when I was at Bing and he was at Bing, and he used to say this thing. I actually have it written on the whiteboard behind me. It’s right at the top. You guys can’t see it because it’s cut off by the camera. But I have a little note written up there. And the note is, the quote from Satya is, “Don’t be a know-it-all, be a learn-it-all.”

And every SEO today, if this is your career and this is your focus, you have to be a learn-it-all.

My life has been taken over by all of this. Evenings, weekends, here, there on my phone, when I’m out for a walk. I’ve found ways to weave learning and education and information and thinking and creating throughout all of my life’s existence, simply because that’s what it takes from me at this point in my career to feel like I can still see the front edge.

I don’t think I’m keeping up with it. I can still see it because keeping up with it is a whole other level.

Stewart Gandolf (Healthcare Success): I love that. I love that concept. By way, I didn’t mention this really. Your credentials are so impeccable. Like anybody who understands the CEO, it’s like, wow, Bruce Clay Yeah, all these different things that we’re all, like, we look at and schema .org, all of it. And I love the fact that you have that humility to continue to learn.

And Brandon, I’d love to pass the ball to you for a sec about the learning part of it. Because Brandon and I are the same way. Brandon is working 12-hour days with us sometimes, and it’s still spending time learning more of it.

Any comment on that, Brandon?

Brandon Schakola (Healthcare Success): I think it’s part of that part of the existential crisis of understanding you still have to do your normal job which is being an SEO but the floor is sinking because the goal is going further up and it’s that anxiety of not knowing it’s driving us forward at this point. I don’t know if there’s a new term for it, or SEO nausea or something, but we probably should have one.

Duane Forrester (duaneforrester.com): I’ll tell you, Brandon, you kind of nailed it, right? And again, I’ve talked about this a little bit in the book and on my Substack. I did a survey of jobs, SEO jobs recently, where I looked at, I think it was close to 300 jobs, and I took a look at how many of those job descriptions, and these are all live jobs, how many of them mentioned AI in some way, whether it was ChatGPT by name or AI or the phrase LLM or geo or something related to this new emerging space. And it was like 2% of them that actually mentioned it.

And now here’s my takeaway, okay? Largely unknown to the audience. I am fortunate. My wife’s career has been in human resources, and she is a director with her company. She is a very senior HR practitioner. She actually teaches the certification class for people wanting to get the state certification in HR in California at California State University. So that’s one of the things she does.

So point being, I have expert-level HR in my kitchen. I make lunch for her. It’s that accessible to me. And I sat down with her and asked her, “Is this rational?” And she goes, “It’s not rational, but it’s real.” She sees it in the jobs that she’s hiring for and whatnot. And the craziest part of it is, we have companies right now that are hiring SEOs for skillsets that in two years, they are going to be asking different things of them.

And those same companies do not have any process internally for how to train the person they hired into those skills. So somehow all of these companies are hiring you, expecting that you will, through osmosis, I guess, learn all of this on your own time, and you will become an expert on it enough to impact their bottom line.

Save the company, grow the company, whatever your bottom line happens to be, and this is a fundamental problem. And what I see are, I see a lot of people sitting on the fence across our industry of, you know, “I don’t really like this. I’m not going to get behind it. I’m not going to focus on it.” They pick apart little things. And yet behind the scenes, I know everybody is working their backsides off to try “How do I utilize this? What do I do with this? You know, do I build something with it? How do I use it to do my work more efficiently?

I think it’s safe to say we all understand the efficiency gains that are to be had from this. But my biggest takeaway is, if you’re using all of these systems and you are talking about these things, you have to be an expert-level PM when you’re using these systems. I described it recently as it’s the difference between Ironman and Superman. And if you approach this and you say, “I want the LLM to be a Superman, I want to just tell it what I want, and I want it to go and figure it all out.” That’s agent-level stuff. And quite frankly, we’re not there yet.

Iron Man, on the other hand, is the suit that amplifies everything you do. It is what makes you stronger, faster, better. And that’s where we are. You have to be the pilot in an Iron Man suit. You don’t just get in it and say “Go fight the bad guy,” like, yeah there’s a bad guy; I’ll get you there, but throwing the punches is got a neat direction on that, right, and that’s our reality when it comes to where we stand.

Stewart Gandolf (Healthcare Success): I love these metaphors and I would just say as a quick aside, I marvel at how you know, going back to your sort of “SEO by numbers,” you know, like basically “do this and this and this and this” and that is the old school. A lot of companies build their scaling around that. They’re trying to find a replicable system.

And I get it. And I wish we could do that more with our agents.

Duane Forrester (duaneforrester.com): And you still need to do that work.

Stewart Gandolf (Healthcare Success): Yeah, yeah.

Duane Forrester (duaneforrester.com): You can’t just abandon that, right? You have to complete it.

Stewart Gandolf (Healthcare Success): But the insight and the experience—it’s such a driver. One of the questions I ask AI, which is always fun, and ask it about itself, “Is AI going to make us smarter or stupider?” And I think my opinion on this, and what the AI said is, it’s going to probably raise the tide, but the people that are smart that really are dedicated to learning are just going to get that much bigger of an advantage.

They already had an advantage, right? They have more education, they’re smarter because now it’s turbocharging them so now they’re able to move so much faster.

Duane Forrester (duaneforrester.com): That and I think it also keeps your brain sharper. That’s what’s going to happen, because if you spend any time using these platforms you see them make mistakes in real time answering you, and you have to correct that and that forces you to stay on top of things, so if you…

Brandon Schakola (Healthcare Success): Yeah, certainly, but it also kind of like—Nate B. Jones talks about this a lot as well and his Substack is how it’s changing and shaping us is that it’s requiring a certain level of clarity before even getting in there.

Duane Forrester (duaneforrester.com): PM skills.

Brandon Schakola (Healthcare Success): Yeah.

Duane Forrester (duaneforrester.com): You’ve got to plan this out, right? I don’t know about you guys or the listeners on this, right? But this is a frequent thing for me, like daily. I get so deep into a thread on Claude, and I have the full paid for like everything, right? So I’m totally unlocked. I get so deep into a conversation that the system runs out of capacity and can no longer properly function in that thread.

And what I have had to do is teach myself that when my slider gets this big on that monitor, that’s the indication to me to ask it to reframe everything into a new prompt to start a conversation with the maximum contextual awareness for continuing that thread.

And I mean, you can beat your head against the wall and get like basically what I think of as stupider and stupider answers as you go deeper, but there’s clearly an ebb and flow that happens with these systems as you start a conversation and then you’ve touched all the pieces where it gets smarter about everything it’s pulling in and then it truly is impressive with what it can do and then it starts to forget things or confuse things or kind of mash things together and this is where you come in to play.

This is why at like hour 12 or 13, I have to walk away because I’m getting punchy. I’m getting like I’ll miss things and I’ll say yes to something and it’ll generate some code for me and it wasn’t what I wanted. Watching for that is so critical. So does it make us smarter? I think possibly. But ultimately it’s like a spoon, right, it really comes down to how you’re using the tool.

Stewart Gandolf (Healthcare Success): I would say it satisfies my curiosity. I didn’t realize how curious I was until I came out because now there’s no question—I can learn anything about anything really fast and so I think that natural curiosity I’m glad I have because I’ve never been able to feed it so fast.

The other thing you just mentioned there is that does scare me there are times where I will catch it making a mistake or I will catch that there’s a lot of times where the context isn’t right so I’ll ask it a question about something that’s going on in politics and it’ll give the opposite point of view, like for like a journalist for balance, like “Wait I didn’t ask you for that. I’m not looking for factual back here, not, you know, political comments. I was just trying to find a fact.”

Go ahead, Brandon, so you have a comment on that.

Duane Forrester (duaneforrester.com): I was going to say, I see this all the time where you tell it not to fixate on something, and the only direction it takes is fixate now. And the next answer you get is exactly what you said you didn’t want. And then you’re like, “No, no, I said, don’t fixate on that.” “Ohhhhh. Mistake.”

Brandon Schakola (Healthcare Success): You know, I’m like, okay. It’s kind of like the, I’m glad you’re talking about this because I’ve had this in my mind for a while that really our role as SEO now is becoming context engineers. Mike King says “relevance engineer,” but I think it really is a contextual, like a broader problem.

And within that, it’s like anyone working with ceremonial magic or some other esoteric thing, that just by mentioning the thing, you invoked the thing that you didn’t want.

Duane Forrester (duaneforrester.com): Yeah, exactly. Right?

Brandon Schakola (Healthcare Success): So you really do have to think with these, like, massive guardrails and incredible clarity and accuracy in how we work with this.

Duane Forrester (duaneforrester.com): You know, Brandon, I think this is an incredibly overlooked important point, right? The accuracy and the clarity. So again, this is my Superman/Iron Man analogy, right? If you hit that context box and you give it a sentence and you convince yourself with your smooth human brain that there’s no way they can think anything other than what I’m thinking when I give it this sentence. You need to check yourself at the door because as someone who worked inside a search engine and I am intimately familiar with the concept of disambiguation and how important it is, these systems are not Superman. They are not Kresskin. They are not reading your mind when you write that sentence.

Like, you just watch. When you give Claude, you complete something and you say “Oh man that worked awesome,” when you watch when it’s thinking out loud what it’s saying it’s saying “parsing ambiguous statement from user” so its first touch on that is literally the admission “I have no context for what this means” and so when you’re thinking about asking it, giving direction, trying to use it to help you update something, define the envelope, take away as much possible offramp as possible. So there is only one road that it could possibly go down.

And if you said to 100 people, does this make sense? A hundred people end up at the exact same spot you think they should. That means you’ve got the clarity and you’ve gone to the depth. That’s how you have to write these things, which when you think about it is really problematic because we all turn to these things and we’re like, “write me an article on SEO.” And, you know, like 5 ,000 words later, you’ve got an article on SEO. I have to do very little. I got an outsized ROI for my time and input.

And that’s what people think these things are. It’s made worse by people running around saying, “agents,” and “you should sign up here.” And “Here’s a cryptic Super Bowl ad. Go claim your name for your agent at AI, blah, blah, blah.” And like this type of thing is problematic because they’re not capable of that yet. And so I get it.

We’re moving in that direction. I think it’s a logical direction. And with tens of billions of dollars being poured into it, we’re going to get there. There’s no question. But I’d like to build a camping bot where it answers all questions about camping, and just own that space. But even then, it’s not narrow enough. It’ll drift off and do stuff, and how do you fix it?

Stewart Gandolf (Healthcare Success): We would love to talk about all this stuff for the geek sides of us, Brandon and I do this forever offline, but I’d like to bring it back to the business leaders that are listening. And, Duane, I’d love to have some of your thoughts about, what are the biggest mistakes they’re making? And then we’ll pivot from there to what we’d like what they should be doing, you know, Monday morning. Frame that for us.

Duane Forrester (duaneforrester.com): Look, first thing, we’re measuring things incorrectly. We’re measuring the wrong things. If you’re walking in as the business leader and you’re still looking at all the SEO metrics, it’s time to have a conversation with the team and say, “Hey, what new metrics could we be looking at that tell us a story of what’s happening in an LLM interface?” Okay?

What I’m hoping is happening is most SEOs have already had that conversation with their executive. But, you know, you as an executive get handed information and you’re told here’s the story. It’s really important for you as the executive to look at that and say, is that the whole story? Is there more to this story?

Like, if I ask you the same question, why don’t I rank number one in Google? And I said, why am I not the answer to ChatGPT? Let’s have that conversation, you know? So that’s the first thing, right? Measuring the wrong things.

And it’s not even the wrong things because we still have to focus on things, like your keyword research and building up from that and then going through things. Do you rank well? And what’s your percentage of the footprint? And all of this stuff is still important.

The second thing that I see, though, is executives treat optimizing for AI answers as the SEO team’s problem. They have pretty much all turned to the SEO and said, “Fix it. You speak a weird language. This is another weird language. Go.” And off they go.

And listen, as an SEO, you feel good about that because that makes you feel safe in your job. If I learn a bit more and I know a bit more and I talk a bit more, then I am vital and that that feels good right that’s really important but the problem is it’s a cross-functional alignment that has to happen. You can’t have PR doing one thing. You can’t have marketing setting up campaigns on their own. You actually have to have IT tuned into this.

I’ll lose my mind when I see robot TXTs that are blocking the crawlers for the AI platforms. I’m like, so you just never want to be a part of the conversation? Got it. Okay. You know, welcome to Sears. Here we go.

And this is really a fundamental. And yet the debate rages on. Oh, should we? Shouldn’t we? Look, I can tell you right now. First off, they already have your content. Second, your content? Not that precious. It’s not that big a deal. Don’t worry. Either 1 ,000 monkeys with 1,000 typewriters are going to come up with that piece of content at some point because odds are—or they already have it. And so, you know, you’re fighting the wrong battle.

But PR, engineering, legal, content, everyone. You’ve got to be on the same page. You’ve got to be having cross-team meetings so everyone knows how they impact the entity that is the business and how it is seen by the platform, okay?

And the third, third number one, or third area, people are waiting. They’re trying to take their time. They’re approaching this, like, look, I went through and I researched all this and what I figured out was the update pattern that we saw from Google, there are major updates, the core updates that they do, right? We saw the pattern of what the frequency was on that. And then I went and I mapped the frequency of notifications about new features coming from Open AI and Anthropic, just those two companies are on a cadence that is 1.9 to 2.1 times faster than Google’s core updates.

So if your entire organization is based around this concept of every six months or so, there’s a big update and this is it, or every year we have this one and that one really affects us because it’s local or it’s YMYL, these types of concepts. And you think there’s cadence to this? Hell no, there is not.

Brandon Schakola (Healthcare Success): It blew up this past year.

Duane Forrester (duaneforrester.com): The only cadence that exists right now is someone gave someone else billions of dollars and said “Go make me trillions of dollars.” Okay, I’m gonna go Dr. Evil on it. And now they are breathing down the neck of that person saying, “Where are my quintillions of dollars?”

And so that person now, the only thing they’re focused on: push faster, push harder, do more, show more, engage more, a billion, two billion a day, three billion a day. Like this is their focus, not “Let’s put out an update once every six months to a core product.” That is a completely different mindset.

So those are the three areas, I think, right? Speed, unity within the organization, and what you’re actually using to tell you the story of what’s really happening.

Stewart Gandolf (Healthcare Success): I love it. Everything you just said resonates so much. We just had a conversation with our head of our internal marketing team about ungating all of our webinars and ungating all of our ebooks. It doesn’t make sense anymore. It made sense in a different time, but it doesn’t make sense now.

We have this incredible thought leadership and what, are we worried somebody might find our thought leadership? No, we want them to actually find it.

So it is a different era, but that’s a strategic decision, right? Because there is a downside to that.

Duane Forrester (duaneforrester.com): It’s a very strategic decision, Stewart, because we all lived in this world of, “I have content that I’ll put out that I will show that makes me an expert. Then I have content that’s behind a gate that proves I’m the expert to those who are paying me. And that’s how they know they’re getting value.”

And now we actually need to show all of that content to a third party so that they can determine if I am the expert that I say I am so that they can tell other people I’m the expert so that those people will come and pay me for my expertise.

It’s the same cycle, but we’ve kind of invited a third party to the conversation now and they demand full access. So I’ll tell you, they who open the kimono first will be more popular.

Stewart Gandolf (Healthcare Success): Yeah, for sure.

We have a good platform for—in our little world, we’ve been doing this content but—you don’t know this, but we’ve been doing content marketing before that word existed by a long time, but we want to always stay relevant. I talked to Brandon about the tiger, how far ahead of that tiger I am I? Sometimes I feel like a mile ahead sometimes I feel like about three feet. If I stumble, I’m eating, so you’ve got to stay ahead of the tiger.

So the last thing then is, before we talk about how to find out more from you is as a business leader listening right now, what are the first things they should be doing on Monday morning?

Duane Forrester (duaneforrester.com): Okay, very first things you guys want to be focused on is you want to go, like, let’s take a look at the three things that we identified that need focusing on. Number one thing you’re going to do is you’re going to sit down and you’re going to have a conversation with your team and ask your team, “Are you all talking about this and you’re all talking about it the same way on the same page?”

If not, you put a plan together, make that happen. Because right now, it’s really hard to get an SEO calling the IT team. IT just doesn’t submit a ticket and you’ve got to wait and there you go. It really needs to get tighter than that on the internal relationships.

You will see companies that will pull forward because they’ve done this and it will make a difference for them over time. You definitely want to be talking about “What stories should I be aware of that I am currently not being told? What metrics are we tracking? What’s our health look like?”

So that whole idea of “What does my health look like already as a company?” You now have to identify the new metrics and start normalizing those. And they don’t have to be metrics that an industry agrees on, but these can just be your own metrics in your company.

Frankly, I say track more than you need. And then over time, you’ll find that metrics just don’t really provide any leverage so you can stop tracking them and that’s it.

I’ve created 12 of those metrics, but keep in mind some of those metrics, we will never have access to the data. It’s platform-only data. But it’s important for a business to know that the data in the platform is used by the platform to make decisions and that these are the types of metrics they’re looking for. Because, again, how do you appear when you step outside as the company person? You can influence those other metrics.

I am a huge fan of continuing to learn. If I were you guys, and you hear this all the time, we hear this all the time, is everybody saying, “You have to use AI more” and this kind of idea.

Start learning about it. I have frameworks on my personal website at duaneforrester .com. Go, the frameworks are free. One of them is the 90-day learning path that I outline. It’s in the book as well. Like, here are things you can do over a 90-day period to increase your understanding and efficiency with AI, right?

Put something like that together in your company so that everyone feels like, “Oh, we’re all learning something. Here’s why we’re learning it.” It wasn’t just one group that was told, “Go do something with AI,” and then they do internal presentations. It means to be everyone. Like, these are the important things, right?

There’s a scoring system in there as well. If you want to be really, really geeky about this, come in on Monday morning and use that scoring system and start tracking yourself in a spreadsheet. See how you’re showing up for answers, whether you’re being cited. So there’s a link. Mentioned, you’re mentioned, but there’s no link. Paraphrased, you know you’re in the part of the answer, but it’s not your words per se. Or you’re not cited at all.

Those are the four states that you can exist in. And you can start tracking that tomorrow morning if you want.

Stewart Gandolf (Healthcare Success): Duane, I’m going to ask you about where to send people in just a moment. You’ve already started that. But Brandon, any of the final comments first before on this topic? And then we’ll give Duane…

Brandon Schakola (Healthcare Success): I would agree with Duane. Like, this is not a time to sit on the sidelines this year. If you haven’t started banging your head up against this wall yet, there are also other resources outside of the ones Duane mentioned.

There’s something called AI Cred. I think that was just released where there’s a whole system where like it literally tests how good you are at that sort of clarity moment that we’re talking about.

That’s another good one. This was fun. Good chat.

Stewart Gandolf (Healthcare Success): I would say that for sure the urgency part of it, you know, our business lives on its website, we do speaking, we do writing, we do exhibiting, we do lots of different things but the website is still really important for us and the urgency of updating various key sections, just like you’re talking about, creating content around things that weren’t even an issue a year ago, are really important to us.

I think that our clients as well, it’s like “We want to be in the lead position.” So working our own stuff as well as for our clients to pivot as fast as you can to get a leadership position or maintain the leadership position.

 So you had a couple of other, in addition to duaneforrester .com, what other places would you like to point them toward?

Duane Forrester (duaneforrester.com): Look, I’m going to keep it simple for folks, right? There’s three things I want to keep in mind. Duaneforrester.com will find me. I’ve got free frameworks and resources on there, some of which are from the book.

So if you want to get an idea of what the book’s all about, go there, get it, take the frameworks, use them.

Second is, you know, go hit Amazon and look up The Machine Layer. And if you have a copy of the book, leave me a review. Books live and die based on reviews.

Drop a review and help me along and obviously if folks want to connect on LinkedIn, I’m easy to find. If folks have questions and they know you guys, call Stewart  or call Brandon. They’ll get me on the phone and we can have a conversation. Just reach out if folks need anything.

Stewart Gandolf (Healthcare Success): That’s awesome Duane, I told you this would be fun. Hopefully you agree.

Duane Forrester (duaneforrester.com): This has been awesome.

Stewart Gandolf (Healthcare Success): So very good. Well, I want to thank you, Duane, for joining us. Brandon, thanks for being part of the call as well. I hope our listeners have enjoyed this. This is, you know, kind of a master class today. And Duane’s such a leader in the field. I love having you. Thanks for saying yes so quickly. And I hope to have you on again someday soon, especially when your new project announces. We’ll make sure that we’ll be happy to cover it for you. Thanks again.

Duane Forrester (duaneforrester.com): Brandon’s seen it. Probably another two or three weeks, I’m thinking.

Stewart Gandolf (Healthcare Success): All right, awesome. Thanks again.