Helene Jelenc:
If you're not thinking about how do I appear as a brand inside AI, you're losing money. But I think complicating and trying to be everywhere at once as a new brand, no, it's just not going to work that way. Locking in on what you do, your messaging, your positioning, your content, these are things that SEOs have been talking about forever. And so now people are turning and looking back at us like, "Wait, what were you saying two years ago?"
Steve Smith:
Hello everyone. Welcome back to another episode of Work Tech Weekly. I'm Steve Smith, Managing Director for growth at Rep Cap. There's a lot of noise right now about search. "SEO is dead." "AI has taken over." "The old playbook doesn't work anymore." And if you spent any time on LinkedIn, you'd think the entire go to-market world flipped overnight. But what if that's not entirely true? Today I'm joined by Helene Jelenc, Director of SEO at Flow Agency. Flow is a B2B search and visibility agency that works closely with software and professional services brands, especially in the work tech and HR tech space. Helene and her team have been digging into real data behind AI visibility, AI mode, answer engines, and how workplace technology brands are actually showing up.
In this conversation, we get into what the numbers really say about AI traffic, how discoverability is changing across search surfaces like Google, YouTube, and ChatGPT, and why brand authority is becoming the deciding factor in whether you appear or disappear. If you're wondering whether this is a complete reset or simply a shift back to fundamentals, this episode will help you recalibrate. Let's get started.
Helene, welcome to the podcast. Glad to have you here today.
Helene Jelenc:
Thank you. I'm so happy to be here, Steve. Thanks for having me.
Steve Smith:
Well, you're very welcome. I think just to start things off, it's worth mentioning that you're from our partner agency, Flow Agency. We partner a lot of work together. And I wanted to give you a chance just to kind of give the quick 30-second overview of why we think you're so awesome and brilliant, and what do you guys specialize and focus on?
Helene Jelenc:
Oh, well, it's very generous. Thank you for the shout out. Yeah, Flow Agency is a B2B search strategy agency. We work with organic, AI, and paid for B2B SaaS brands, B2B professional services. Definitely focus a lot on the work tech, HR space. And yeah, we love to be a strategic partner for our clients and bringing them into the new world of what search is now. It's a whole new space than it was a year ago.
Steve Smith:
Oh, yes, it is. And why don't we just kind of dive on in there. I think that one of the conversations that it seems like I'm having several times a day with our clients or with sales prospects is basically what the hell is going on with go to-market specifically with search. Because I think that a lot of the people we work with, they had the SEO playbook down. It really hadn't changed that much over the past 15 years other than when Google would update their algorithms and things like that. But everyone kind of knew how the game was played. And since the launch of ChatGPT a couple of years ago, what the experience has been is, "Oh, all of that magic and that voodoo doesn't work anymore." Cost per lead is going up, conversion rates are going down. Everything is, it feels like the bottom's just kind of fallen out of a lot of go-to-market and specifically demand gen. And now everyone's just like, "Okay, so how do I surface in an answer engine search?"
So based on kind of that context, I know that you've done a lot of research on this and that you can kind of cloud the situation with some facts. As we're entering 2026 and we're dealing with this, with companies that are in the work tech space, what is your top of mind observation? What is the data telling you and what are you talking about most in your conversations today?
Helene Jelenc:
There's a lot of things happening. I would say that if you were to listen to a lot of the posts you see on LinkedIn, or the SEOs, or other marketers crying, "Search is dead." "SEO's dead." "My traffic is gone." "Organic traffic is gone." That traffic wasn't really serving you anyway. It was unqualified traffic in my book. And we got away with basically kind of replicating Wikipedia for a long time. I mean, if Wikipedia can answer it, why are you creating the content? You don't need to be. The point of creating content is to connect with your audience. Like I said, they can trust Wikipedia. They know that's been verified. They don't know who wrote your blog post or when's the last time it was updated. And so I think we got away with doing very little for a long time. So a lot of relying on top of funnel, what is content.
You could easily just bash out a bunch of content each week, each month, and you were going to see your traffic grow. It would help you probably build your brand awareness, but it didn't necessarily go into your pipeline. You didn't see an increase in leads or high intent traffic unless you were focusing your strategies further down the funnel. So I think SEO definitely was more at the top of the funnel and now we see it throughout the entire journey. So where more value is being held, where those high intent spaces. And I think if you're not thinking about how to optimize your high intent pages and you're not being surfaced in search, you're not being surfaced in AI, you're not listed in a listicle, you're not on review sites. You just don't exist, unfortunately, because the spaces where people search are different.
Steve Smith:
And so how do you think maybe being discoverable has actually changed in the AI era, especially related to how work tech brands go to market?
Helene Jelenc:
It changed the way people think about search and I think the number of surfaces. If you think about search surfaces as there's a search surface of YouTube, Google, even ChatGPT, these are places where people can search and get information. And so this has gotten very broad and diversified. And not only that, but we have these new systems with AI. So they're retrieving their process of retrieving knowledge or retrieving our content or serving our content, whether they choose to site us or not. All of this is a whole different system.
So for me, I would tell everybody, just lock in, where is your audience? You don't need to be everywhere, especially for the work tech. You know exactly where those people are. You know the communities they're hanging out in, you know the conferences they're going to, the newsletters they read, the topics that they care about. So I would say locking into that and the spaces where they are and not just trying to throw as much spaghetti at the wall, hoping something sticks. So random acts of marketing are out.
Steve Smith:
And certainly, we subscribe to that point of view here at Rep Cap that there needs to be more structure. I think that... Well, I guess, I mean, here's a question. Do you think that maybe go-to-market teams are making this more complicated than it needs to be? That this is actually, it's a leap, but it isn't a completely different world from what they were used to with SEO.
Helene Jelenc:
Exactly. Yes. I agree with this. I think of course AI is one component. When we look, say across our clients, I think AI traffic accounts for 1% to 3% of their traffic. That's very low. It is high intent. I will say it converts very well. So it is money on the table right now. If you're not thinking about how do I appear as a brand inside AI, you're losing money. But I think complicating and trying to be everywhere at once as a new brand, no, it's just not going to work that way. Locking in on what you do, your messaging, your positioning, your content. These are things that SEOs have been talking about forever. And so now people are turning and looking back at us like, "Wait, what were you saying two years ago? How do we set this up? How do I make sure something can crawl my website?" And yeah, so there's been education I think among that, but then poor marketers, they're like, "What do I do? There's so many channels to work with."
Steve Smith:
Well, yeah. And you brought up an interesting point because there was an article that I read the other day in the Wall Street Journal about AEO and they were talking to someone who had done some brand research and they were mentioning something like, "Well, typically we're seeing brands are getting about 40% of their traffic from LLMs and answer engines." And I was like, "40%, what brands are those?" Most of the brands I'm seeing top out at about 10% or 11% and that's pretty good. And so when you mentioned kind of 1% to 3%, I think it's helpful because I think it really does kind of set the baseline in the proper place. When you mentioned numbers like that, is that a revelation for your customers and the people you're talking to?
Helene Jelenc:
Yes. Yeah. I think you're hearing so much about AI search and AI visibility, how important it is, but when you look at the numbers, you see it's just the current hype. Of course, everyone's talking about it. Everyone's excited. We're all trying to figure it out, but I'm not saying it's not bad. This is highly qualified traffic. I actually think this is a very high intent source of traffic and that we should be very mindful what pages are they visiting and making sure that those pages are good. Because the other side of that is that AI can pull pages you maybe published 10 years ago and forgot. So it's definitely sort of a house cleaning moment, I would also say, for people who have an old blog.
Steve Smith:
Well said. I think another thing that's been coming up a lot is how is AEO changing how different sort of validation signals or citation signals are weighted. So we went through this period, it seemed like about six months where everything's moving to Reddit, you need to go and optimize around Reddit. And then it seemed like that blew up. We have a lot of conversations around YouTube, which probably could be 30 minutes of its own conversation where it does seem like YouTube is having a moment. And then sort of getting into PR and public relations having a renaissance because all of a sudden, earned media seems to be even higher value as far as answer engines are concerned. What's sort of your perspective? What are you hearing? What are you talking about? What is the research showing around these topics?
Helene Jelenc:
Yeah. If you read most research, you would see Reddit as a top source. Everyone says, "Let's go to a Reddit strategy." I'm the first one who will say, "Let's pause the breaks here. Let's pause and see, does that make sense?" And it actually is a reason that inspired me to start researching keywords in the work tech space. I wanted to see in when they appear for AI overview or other AI systems, do they react the same way as other studies are saying? So other studies are doing like, "Oh, top 10 sides, top 10,000 keywords." That's generic. That list of keywords is a joke. I'm sorry to everyone. But we have a highly curated list of keywords that we've done through multiple keyword researches and then we mark them with their search intent so that not only were we looking at all of those keywords, but we were looking at the intent and how search would act differently.
And what we found actually was that Reddit wasn't that big of a source in the work tech space, and which made sense to me. That's just not a community where they're hanging out and talking. And so while maybe field service management or our clients doing that, yes, their clientele is on Facebook, they are on Reddit. But when you're talking about the work tech and HR tech, did not see Reddit as a top source, ran the study twice and it wasn't there, but YouTube was the biggest one. YouTube is the unsung hero of AI and of search. We did a small study last year and saw that more than AI overview video appears. And then Google started integrating video into AI overview. ChatGPT is giving videos now as well. Sometimes I'll get three videos in a single AI overview, especially if you're asking a bottom of funnel, high intent software, you should be there, create your product demos, get them uploaded, keyword them.
Steve Smith:
So Helene, one thing I wanted to ask you about is I know that you've done a lot of researcher and keywords. What has your research told you about maybe how AEO is different for workplace technology companies and what, if anything, surprised you about your research?
Helene Jelenc:
Sure. Yeah. Our latest research was looking at the visibility and consistency of AI mode. So AI mode is sort of an extension to the AI overview. If you've even clicked like view more, now it sort of pushes you into AI mode. So Google's pushing this as their sort of future search space. And we wanted to understand, okay, if we can rank in Google, does that mean that we also appear in AI overview and AI mode? That's the question. That's what everyone wants to know. Is my SEO going to give me AEO, GEO? The answer was not really. We were not expecting this answer at all because I was the first one to say, and I still agree with it, that if you have good SEO, you will appear in AI. That isn't really entirely the story, I will say.
If your content is doing well inorganic, it does increase the chances that you will appear in AI systems within Google, but specifically AI overview has a heavier lenient... It leans heavier on your domain authority and your brand authority. And so it's selecting domains based on where it thinks the authority should lie. So when we looked at work tech, for example, looking at the sources that were chosen for bottom of funnel versus top of funnel were very different domains. What that told me was that, well, it understands this intent. It understands that people are looking for a solution. They're looking for a product, a feature, whatever it may be that they're looking for, and it gets it. It also gives a lot more citations. So the higher the intent, the more citations, the more competition you have. Usually, unfortunately, there are also aggregators like G2, Capterra. Maybe we will see a consolidation of all of that. Our fingers are crossed, but for now we are competing against those sort of domains when we're looking at the bottom of funnel.
Whereas on the top of funnel, it can be just any sort of product brand. And if you have blog posts on different topics, just like I was saying, top of funnel, maybe then not as what is field management service, but if you put a little twist to it, then you still have the ability to appear there for top of funnel. But I will say the high intent is very competitive. It's a competitive space. It's transactional. The ads are going up. AI overview is taking up more and more space in real estate above the fold. Sometimes you have to scroll forever to get an organic result. And so these systems are so interconnected. I think this is the big giveaway. It's not going to replicate between organic results and AI mode. It's just not. I even have a nice number here. The overlap between organic and AI mode was only 25%.
Steve Smith:
Wow.
Helene Jelenc:
And between AI mode and AI overview was 16%.
Steve Smith:
Wow.
Helene Jelenc:
And so there's very low overlap there. And then another part of the study, we did multiple parsings just to see what is the consistency of the citation choices. It was about 18% of overlap.
Steve Smith:
Wow. To me, I hear that, and that's a lot lower than I thought it was going to be. Did you have the same take?
Helene Jelenc:
Yeah, I was shocked. I was absolutely shocked at how low the overlap was between the three systems. And so that tells me that they are acting differently. As Google stated, the tech behind them, the way that they're generating answers and the way that they're taking your search query and processing that are different. So they're different search surfaces, as I was mentioning before, and they require different things to optimize for them. So I kind of take back my original saying of good SEO is GEO, but you still need the foundation. You need a good site. You need to be able to crawl it, and you have to have content.
Steve Smith:
So what are sort of the actionable takeaways from those conclusions? Because what I heard, and maybe I heard it wrong, is you really need different types of content for... And it should be not a surprise for marketers, you need different content for different stages of the funnel. But having that be discoverable is something that I think is new because traditionally you would see that kind of mid and late funnel content that is, as you know, high intent that is not always front facing or visible or discoverable by design. So are you seeing that same... Did I get that right?
Helene Jelenc:
Yeah. Yeah. I think it's really important to be more transparent with your content, your brand, your product. If you're not talking about it, someone else is, or AI is going to make it up. People don't know the difference.
Steve Smith:
Well, that's the thing is that I think that people worry about the hallucination factor. It's just like-
Helene Jelenc:
That's true.
Steve Smith:
... are you really... Does that change go to market at all, just that one variable? Or is that maybe... Does that influence the recommendations that you would make?
Helene Jelenc:
Yeah. I guess it ties into one thing that stood out to me, and then you'll hear a lot of SEOs talking about is your brand. What does that mean as an SEO? Well, that means that your brand is consistent across platforms. Everything, your owned, your borrowed, your paid platforms. The way you talk about your product, the way you talk to your audience, is it the same on G2 versus Gartner versus your website versus a listicle? Or do you have your old category or old messaging out there? That actually will help AI to hallucinate. When I said about purging old content, that's going to help AI hallucinate. And so I think getting your brand is the main layer for AI. If you got content moving, that's great, but you need to lock in how your brand is perceived.
Steve Smith:
And what's funny is that brand perception could be shaped by webpages that you forgot that you had from five, six, seven years ago that have just were thought to be lost to the mist of time, but the internet never forgets.
Helene Jelenc:
Yeah.
Steve Smith:
Well, I'm curious, do you see a change in how brands are building content away from writing an interesting blog post to writing what they think will be the answer that gets surfaced in an answer engine?
Helene Jelenc:
Yeah. I will say there's definitely a shiny thing syndrome going on with AI and people just want to appear. Executive teams just want to see the sessions go up. They want to see the visibility go up. They don't care really what the results are of that, what is the impact of that. Nobody really knows at this moment, but they just know they want to be seen. And so yeah, there is definitely, in some cases, content that we're just like, "No, you should not be doing that."
Steve Smith:
So what's a typical recommendation that you find yourself making over and over when you're talking to clients? What is something that you see that is either low hanging fruit, simple to fix, or a consistent mistake that's happening across everyone you're talking to?
Helene Jelenc:
I would say, especially if you have a small team, and just starting off, you're going to feel overwhelmed. There's a lot of players out there who have a lot of years ahead of you. Lock into what you're good at. Lock into that one thing that's working. Create your content around that. Create it and build it out richly. Don't just start publishing things randomly and hoping maybe one of them will stick. But I think committing to a very specific strategy, committing to being aligned with your product and showing what it does, showing how it helps people. I see these sites thriving, even with organic traffic. The clients who are publishing regularly original content. Sometimes we don't find the keywords in a database. They don't exist yet. We're pulling these ideas because they're fresh emerging ideas. We see them on forums, we see them in the news, we see other things. We're trying to catch the Google trends before it goes up. And so we're definitely trying to do a bit more of this and really leaning into your expertise. And I'd say, if you do this, you're going to be okay.
Steve Smith:
Well, that's reassuring. I think that there certainly is a lot of worry going on out there, at least among the go to-market people I'm talking to, about is this something that we can get ahold of and that if we can turn it to something that we can work with as opposed to it all feels very, very uncertain and kind of squishy. Because even you talk to the people who are building these algorithms and you ask them, "How does it work?" And they're like, "We don't know." And so it's just like, well, if they don't know, it's just like, how are we supposed to get a handle on this as people who are expected to produce content that will surface in a response to a prompt?
Helene Jelenc:
You brought up a really good point there. The tracking of how often your brand actually appears inside AI. How often are people talking about you? We will never get that data. Due to data privacy, it's just not going to happen. And so also when you're looking at data from AI, you could see a session came through, but that's it. You could just see a session landed on a page. We have nothing else to work with.
Steve Smith:
Wow. So there's a lot that's been made recently of ChatGPT saying, "Oh, we're going to have ads." Do you have any sort of immediate knee-jerk reaction to that? Of course, we don't know what it's going to look like. We don't have any data. What's sort of your immediate thought on that? Is it just like, oh, they're going to be like... In two years, they'll be like Google was five or six years ago, or is it more complex than that?
Helene Jelenc:
If they do it smart, maybe. I'm curious because of their market share loss over the last few months, but yes, ads are 100% coming this year. They will be a little bit different than Google's. They're going to be paid per view rather than click and they're going to be quite expensive and the data will be limited. That's all we know so far. But I think they're having an exclusive sort of rollout with brands that spend significant money to start as testers. It is coming. I just feel like you're never going to escape ads and we should have known it was coming. The product is way too cheap for what you could get out of it. So yeah, it will become a Google. I think whatever AI system everyone decides to adopt, it's going to have ads.
Steve Smith:
Wow. Do you see a future potentially where AI and answer engines become the industry analyst? We talk about people being put out of jobs. It seems like the people who supposedly are the experts in the software who have a good revenue channel and certainly a good livelihood, is that a position that potentially you think could be disintermediated or go the way of the dodo?
Helene Jelenc:
I don't know. I don't think a lot of marketers' roles will be taken over by AI just yet, especially because of everything we just talked about. There's a human cultural nuance to everything we do that AI does not understand. It's just not going to get it. It's not going to get the daily lived experiences of people, relationships. It's just not so... And I think with marketing, you need the human touch. And so for now, I hope so. I hope we all have our jobs, but I think being able to use your brain and come up with creative solutions is something that AI can't always do.
Steve Smith:
Well, I think that's an interesting point because I heard someone talking about this exact question the other day and this person made a point that when it comes to pulling discreet facts, yeah, an answer engine can pull all kinds of things. But if you think about true creativity, you can grab... Well, don't grab, but you can find any three-year-old out there who is just inherently more creative than the most sophisticated answer engine. And that really kind of gave me pause to think, wow, there's something that is hopeful and beautiful about just the mysteries of the human imagination that we are right now not close to even being able to replicate in answer engines. Do you think that is just... Do you agree with that or is that just things we tell ourselves in order to not lose our minds?
Helene Jelenc:
No, I agree. I will preface it by saying I do have a master's in anthropology, so studying culture and humans is one of my personal passions, and I think which gives me an advantage in SEO. And yes, I work with AI every day and it just doesn't get it. It is very good at being like a calculator for words. I don't know. It can do things and give you some stuff, but yeah, it's missing the human touch.
Steve Smith:
I want to dive into that anthropology thing really quick because I think that anthropologists make the best marketers. What do you think you've taken away from your background as an anthropologist that has been highly useful either in marketing generally or working with the mysteries of SEO and AEO specifically?
Helene Jelenc:
Two things. One, learning to interview people, to listen, and to kind of leave yourself at the door, just really immersing yourself into a new space and really just understanding what's going on, seeing who's speaking, who's not speaking, understanding the whole picture. It seems like it doesn't make a difference, but I promise that being able to read a room in that way, especially if you're not someone who's always in there and being able to make observations is a skill that you need to train. And also mixed methods research. One thing I did focus on was qualitative, doing interviews, but I also did a lot of mixed methods. So taking that qualitative ideas, how do I also mix quantitative data and then how can I get a story out of that? How can I take that qualitative, mix it? Where's the story? What's going on? What's happening? And yeah, I think that's really valuable for SEO, being able to see the whole picture and yeah.
Steve Smith:
I have a philosophical question to close us out today.
Helene Jelenc:
Okay.
Steve Smith:
So in two years, what do you think we're going to laugh about from all of this, kind of this hype cycle of AEO, GEO, answer engine kind of freak out? And what do you think is actually going to stick from what we're talking about today?
Helene Jelenc:
What are we going to laugh at? I would say the LLM TXT file, but I've already been laughing at that already, so that's a bit of a cheating answer.
Steve Smith:
Okay. So explain that to people who may not understand it.
Helene Jelenc:
Some SEOs, I guess, have decided that there was like a robot TXT file, but there's an LLM TXT file. So you just explain everything about your company on this TXT file, upload it to your website, and AI will read it. I mean, cool. It reads every page on your website. You just discovered how Google works. Why don't you put that content on a page you want a human to visit? If I was a human and I landed on an LLM TXT file, what am I going to do there? You can't navigate to the site.
So yeah, I think trying to do a quick fix like this, everyone said schema was going to work, AI engines cannot read schema. So I think we're going to be laughing at a lot of these little tricks that we tried to pull. What's going to stick? I think brand. In a sea of millions of software options out there, how do you differentiate yourself? It's always been the same, through your brand. And right now brand is the way that AI understands you as someone different. How do I know that you're not BambooHR versus Betterworks? I don't know. AI does that. So your brand determines yet the differentiation of your competitors. So yeah, I say brand, it's here to stay and I think don't sleep on it.
Steve Smith:
Well, fantastic. I think that's a good word to end on. Helene, thank you so much for joining us. It's always a pleasure to have someone so thoughtful and knowledgeable who actually has the data behind them to inform their perspective. So thanks for sharing that with our audience. And this has been a delightful conversation. I really appreciate it.
Helene Jelenc:
Thank you so much for having me.
Steve Smith:
As the dust starts to settle around AI and answer engines, one thing becomes clear. The fundamentals did not disappear, they just got exposed. It's easy to chase visibility for the sake of visibility. It's easy to look at a surface like AI mode or an answer engine and assume the old rules no longer apply. But what Helene's research makes clear is that these systems behave differently, reward authority differently, and surface brands differently. That means shortcuts are unlikely to last.
If you want to be discovered, cited, and trusted, you need a site that can be crawled, content that reflects real expertise, and messaging that's consistent everywhere your brand shows up. Because if you're not clearly defining who you are and what you do, these systems will do it for you. And in a market crowded with software options, that clarity is the advantage. If you enjoyed this episode, make sure to subscribe to Work Tech Weekly on Apple Podcasts, Spotify, or YouTube Music. And I'll see you next time.
