From Webcasts to Generative Engine Optimization: 30 Years of Digital Evolution with Steven Javor Head of Digital Customer Experience and Channel Digital Commerce Canada

Transcript

Our transcripts are generated by AI. Please excuse any typos and if you have any specific questions please email info@digitalshelfinstitute.org.

Jamie Clapper (00:00):
Welcome to Unpacking the Digital Shelf B2B Edition, the podcast for the practitioners navigating complex data, massive catalogs, and the shifting expectations of the modern business buyer. Every month we sit down with the leaders who are moving past the hype and building digital shelves that actually scale. Jamie Clapper here, and welcome to another episode of Unpacking the Digital Shelf B2B Edition. Today, we're joined by a true pioneer in the digital space, Stephen Javor, head of digital customer experience in Channel Digital Commerce Canada at Schneider Electric. Steven has been shaping the digital landscape for over 30 years. From the early days of webcasts in the 90s to pioneering B2B e-commerce across Europe, the US and Canada, he has seen every major shift in tech firsthand. In this episode, Stephen takes us to the absolute cutting edge of digital marketing, generative engine optimization, or GEO. We dive into what it actually takes to make your products visible to AI models, why 90% of your AI citations are coming from sources outside your website, and why the biggest challenge in AI isn't the technology, it's change management and human trust.

(01:20):
Whether you're trying to optimize for LLMs, lead an enterprise AI transformation, or understand how Gen Z and millennial buying habits are reshaping B2B, this episode has it all. So let's get to unpacking. Hey, Stephen, we are so excited to have you on the podcast. Welcome.

Steven Javor (01:37):
Happy to be here. Thank you.

Jamie Clapper (01:39):
Awesome. Well, we've had some shifting of our calendar, so I'm glad we're finally sitting down to do this. So I think the audience is going to love the information you have to share. I think I'm still laughing about you saying, "I don't know if I'm going to be a good fit for this podcast." And I think within the first couple minutes of chatting with you, I knew it was going to be a hit. So let's dive in. Excuse me. You've been in digital since 1995, so roughly 30 years experience in the industry, B2B since 2000. You've been across Europe, the US, and now most recently Toronto. What's the biggest shift in organizations of convincing organizations why digital matters from the early days to now?

Steven Javor (02:27):
Yeah, I guess you can say that I'm OG when it comes to digital. Yeah, it's been an interesting career. I would say that in the beginning, the argument has inverted now. Back in the 90s, 2000s, you had to prove that digital was real. You need to invest in a website, you need to participate online. Today, nobody disputes that. So the hard part is proving that their assumptions about digital are now out of date. Before it was like, should we invest in digital? Should we invest in a website? And you won that with doing pilots and getting a champion. Now it's like, well, we've already invested. Why isn't it enough? So you had to show them that the channel has moved underneath them, that a funded e-commerce site plus a mature SEO strategy can still be invisible where we're now discovering that discovery is now happening with these AI agents.

(03:37):
So it's always moving. Matter of fact, I teach a course at the University of Toronto, and every time I teach that course, I have to update it because I wish I was teaching something like accounting or something like that because you can stay within the thing. But digital, it's like, okay, now what's new? So it's always moving.

Jamie Clapper (03:58):
It's like the ever changing moving target that you have to constantly be updating to. A blessing and a curse, right?

Steven Javor (04:06):
Exactly. And I guess one of the constant things that's happened over my career is that you begin to realize that organizations don't really adopt a technology. They kind of adopt a change in customer behavior usually a couple of years after the customer has changed the way they do things. And I would say now with AI, it's happening even faster where everyone's going online and I'm still trying to explain to people what GEO is. And so back in the 90s, I was arguing to these leaders that customers were now going online. Now I'm telling them that the customers have already left your website.

Jamie Clapper (04:50):
Unfortunately, the truth of what's happening with AI. I do think though, thankfully, you're able to have those conversations and you're having them now because I've heard it from others that I've spoken with in the community, in the B2B space, there's those that are still just at that baseline of why e-commerce matters. And so I think the fact that organizations are paying attention and focusing on that is a huge win. It means that they're heading in that direction and they're creating a game plan for it. So in 2013, you joined Schneider Electric and took on e-commerce strategy for Canada, working with a colleague in Germany to prove out the case for digital before eventually training the entire organization, a process that ran through 2018, so roughly five years and no small feat for sure. I know that anyone who is listening is like, "Holy moly, that's an undertaking." What did it take to go from proving out a strategy with a small team to getting an entire company to adopt it, again, which is no small feat in this space?

Steven Javor (05:56):
Yeah. In 2015, after we did this, I was invited to join the global team and sort of break out and introduce e-commerce throughout all the European countries. So luckily by happenstance, my wife has a German passport, my daughter was wanting to want an adventure. So we moved to Germany in 2015. And that time, the Schneider Electric e-commerce team, you could fit the entire team in an Uber. And it was so small and we tried to figure out how we're going to break this out to all the different countries. And so while I was living in Germany, I didn't really work in Germany. I was working in all these other countries, the UK and Belgium and Netherlands and Poland and the Nordics and Italy and Spain, all these places. And so we had to first adopt a strategy. So my SVP and I, the way we would kick it off is that we would first go and visit the country president and say, "Hey, digital's a real thing now." We would come with a great 45-minute presentation talking about the country, about where they were as far as that country was concerned with digital and how things were changing, talking about Amazon, talking about the change in customer behavior, and try to convince them to give us one person from the distribution team and one person from the marketing team, maybe a younger person who's sort of been a digital native and start the conversation there.

(07:18):
And it was really kind of funny because you began to see the differences between Northern Europe where they were kind of skeptical but then bought into it and Southern Europe where they were like, "Oh yeah, really into it," but then wouldn't do anything. And I always remember the conversation we had with the Italian president, which was a story. So we went there, we did this hour long presentation, and he was OG. He was an older guy. For an hour, he lectured us about why no electrician in Italy would ever think about going online. It's not going to happen. They want to go and talk to a person. They know what they're doing. They're not going online. And we were like, "Oh my God, we just wasted an entire trip here and there's no way it's going to go." And after about 45 minutes of him lecturing us, oh, it's not going to work, he sat there and said, "You know what?

(08:06):
We'll give it a try."

Jamie Clapper (08:08):
What a change. What a quick change.

Steven Javor (08:11):
And it ended up being really successful. So the way we kicked it off, we would start with that, then we would just try to do a small team proof, no mandate, pick some sort of narrow, measurable slice, prove it with numbers that finance already trusts and borrow credibility across regions. Hey, look, it's worked in Germany. It should work here in Italy, which lands differently than just, trust me, I know what I'm doing. And then once we would run that pilot, we would then need to find a champion. We need to have a senior leader that would adopt it as their story, as their success. And that's where that executive cheerleader fits in because you need somebody on the executive team, especially when you're trying to introduce a new technology to champion it for you, to remove some of the roadblocks or some of the naysayers that don't really want to move forward with the pilot and then transition into the other aspects that you need to do.

(09:11):
And you also have to understand about how to talk to C-level people. And C-level people, they're really only concerned about what they're measured on. Is it going to improve my margin? Is it going to improve my customer satisfaction? Is it going to improve my share price? And you have to really couch your discussions and your arguments on showing them on how it was going to improve them to really get them to get on board. But once we got through that and we would do these pilots, usually we would go to a distributor and say, "Hey, let's try a campaign. Let's pick a product. We'll try a campaign. We'll create some content. We'll push it out there. We'll measure the impact over it," and it would move the needle. And then once we got past that, then you realize, "Oh, they want more." And now we got back into, "Well, we now have to move from campaigns to really updating their catalog, which means giving them all of our products with the right product information, with the right images and short descriptions and long descriptions and keywords, but even we weren't ready for that.

(10:17):
So part of that transformation with then going back into our internal organization, we actually created an entire division, which we called factory C, factory content, which was responsible for going through our entire catalog, which was 60,000 products, trying to figure out what the high runners were, and let's fix those first and then figure out a way to get them over to the distributor so they could implement it so that the content was consistent and the same and then build from there. So train the organization, that took a long time. That took the majority of our time to get everybody walking in the same direction, being able to drink the Kool-Aid, but then proving it with numbers. So really the strategy scales at a speed when you can make it someone else's job description. So that was really the key. It's not really the speed that you build it, but getting them to buy into it and then take it on as their own.

Jamie Clapper (11:19):
Absolutely. And I think that still reigns true for anything that you're doing in a digital side of the business. I mean, AI is obviously the hot topic and yes, we will get there because what podcast wouldn't be complete right now if we didn't talk about AI and what's happening in that space. But even I think the thought process that you had to go to the top, get that buy-in, build it, and then be able to bring it down from there, build out the teams. I mean, if you would go back and maybe you've had these conversations, how are these leaders reacting to, that was 2018, so that was roughly eight years ago that you felt accomplished and that you built it out. Have you had conversations with them since? Are they grateful for the push and the nudge to focus on digital? Because obviously I'm sure no one saw digital growing at the rate that it has and all the changes that have happened.

Steven Javor (12:19):
Yeah. And I mean, the pandemic really changed things as well because it came from us trying to call them through them calling us saying, "Oh my God, we have to change all so many things to be more relevant online because everyone's being forced to go online." So that really was a big accelerator for everyone buying into it. So they've bought into it now, but it's still a challenge for them to... I mean, the e-commerce team went from all of us into an Uber, now there's literally hundreds of people across the globe in the 130 countries or so that we work in that are all getting into it in one form or another. But it does start with governance. You need that champion at the top that believes in the story, that believes in the outlook of where things are going, and you can prove that with numbers.

(13:10):
Now, like I said, not so much. If I go back to them now, they all have teams, they all are working to go forward with it, but I think AI is that next level. So when it came to the AI discussion, I'm also one of the innovation leaders globally with Schneider on the AI and marketing topic. And about a year and a half ago, I said, "You know what? I'm using this great little app called ChatGPT." I was one of the first users to jump on that, and I'm finding it useful. "What happens when my customers start using it? What needs to be true for our brand to show up in the prompts that they are creating about our products and technology? "And so I kicked off a GEO pilot and we sat there and said," Okay, let's figure out how to bring this out into the organization.

(14:03):
"Just about three or four of us, I got the team together and we wrote 650 prompts across our top 10 product families, 10 or 12 product families across the customer journey, focused on different segments. And that took about a month and a half because we didn't write 650 prompts from scratch. We used ChatGPT to figure out what should the prompts be. But then I made sure I went to a 30-year vet who had spent 30 years working directly with customers and said," Okay, take a look at these prompts, and is this actually the way that a customer would phrase this thing? "Yeah, this is right, this is wrong. And we took a month and a half, and he went through every single one of those prompts and made sure it was written in the way that a customer would actually perhaps write it into a prompt.

(14:51):
And then we had to tag them because what we did learn as well is that while I write a prompt and I give you that same prompt, we both put it in, depending on our location, we may come up with different answers. So it wasn't really the prompt that was important. We had to tag it with what was that prompt about? Was it about pricing? Was it about installation? Was it about services or support? Because what you end up getting is you get basically what is that prompt basically telling you?

(15:26):
What is the feeling coming back from that prompt? So one of the things we found out was that one of our product sets was expensive, or at least the AI was coming back and saying that great product from Schneider, but a little bit more expensive than everybody else. And then we went there and said, well, where is that coming from? Well, it was coming from this one article from this one blog that wasn't really very accurate when describing our product because our product had a longer lifespan and had all these other sort of add-onset. So it may have been expensive at first, but you had a longer lifetime value. We were able to go back into that particular entity and update their content and notice that there was a change in the citation and recommendation of that particular product. So you begin to see that there are things that you can do when it comes to improving your citation recommendations when it comes to AI.

(16:25):
And then now having to convince the leadership about that wasn't that hard. We kicked this thing off and then within about two months it got in front of our global CEO as this is an important topic we need to tackle.

Jamie Clapper (16:41):
Yeah, I think that's great. I think it's showing instead of where a lot of organizations are said, "Go do AI," that's what everyone hears, and I think they get nervous and they don't have that plan in place or they don't have the roadmap. You went ahead, you built a team, you got buy-in at the top, you did things in the right order. And I think that's what a lot of our listeners need to hear is this is how you truly go do AI, and this is how you start to dip the toe in the water, be able to get that buy-in. Because I think there's this fear, maybe not fear, but misunderstanding of why aren't we participating more in AI, even if the organization is still struggling with data and content, which that's king and queen right there, we all know that, but this is a way that they're able to build out a reasonable way of participating in the AI space, and I think that's great that you were able to do that.

(17:41):
I love your stance on governance, and I know even in previous conversations, we've talked a ton about that, and I always give you praise for that because I think it's so valuable because if you don't have that buy-in and you don't have that governance, you don't have those legs to stand on and continue to test in that space. So you've said the first step in GEO is making sure your website is even visible to LLMs. This is where we continue to talk about AI and all the great things and pain in the butt things it's doing out there in our space, and that roughly half of most company websites aren't. What does invisible to an LLM actually look like, and what are the steps that companies should take to start fixing it to be able to get ahead of this curve as much as possible?

Steven Javor (18:25):
Yeah, it's really interesting. And so what we learned from our pilot was that there's, we call them the three gates that you need to review, and is your content fetchable? Is it chosen by the AI, and is it extractable? So fetchable means can they actually see your site? Can the AI engines actually go through and see the content that you actually have? And there's lots of reasons, and this is usually the biggest blocking point, is that you haven't enabled your site for ChatGPT bot or the Claude bot. You can start with your robots.txt file and making sure you're not blocking those bots from coming in. But then we learned that we had designed our site to be client side instead of server side. So with JavaScript, so to make it faster, you would create a JavaScript so that when someone clicked on the page, it would automatically build the page.

(19:23):
But guess what? ChatGPT doesn't read JavaScript. Claude doesn't read JavaScript. Gemini does for some strange reason, maybe because Gemini is the CEO goddess of the world, but they don't. So we found out that half of our site wasn't even visible. Is your content behind a form or a login?

(19:45):
PDFs also not readable by most of the engines. So the first thing you need to do is to figure out, is your site accessible to these agents? And you work through that. And when you've gone through that, at least that first gate, then the second gate is really, now that they can see it, is it going to choose your content? Is your content written in a way that it actually provides the answer that the prompt is looking for?

Jamie Clapper (20:13):
Hey, Steven, let me jump in here quickly. Sure. When you say once they know that their site is visible, how are they doing that? What's almost a quick, how do they jump in and do a check on that?

Steven Javor (20:29):
Ask your favorite answer engine. Write a prompt that sits there and says, "Review every single page of my entire site and tell me whether you can see it or not and give me a report or write me a prompt about how I can determine whether my site is being visible to you or not." I mean, that's kind of been the evolution of writing prompts. I hardly ever write prompts anymore. I basically write what is the outcome I'm looking for, and I ask the answer engine to write the prompt for me, and then I run that prompt and it'll tell you what it can see and what it can't see. That's a really good first step without actually having to go write 650 prompts and then get your feedback, which is the way we did in the first place. Just ask it if it can see your content or not, and it'll give you at least a first perspective of it.

Jamie Clapper (21:20):
Yeah, almost like a baseline. Okay, please continue. I love this.

Steven Javor (21:24):
Right. Then after you get through that point, I mean, probably some of your site will be reachable, maybe some of it won't be, but then you have to take a look at, okay, how is my content written? And what AI tends to like, it likes content that's written by people, but maybe influenced by AI. It does not like AI slap. AI will recognize AI. So it's looking for facts. It's looking for things that it can cite. It's looking that it's going to answer that question. So what we found really works well and what's really that the AI engines love are FAQ pages that talk about what is this product, who uses it, what are the examples of it being used successfully? And you need to give it real data and real examples for it to use. It's saying that it's the best or the most affordable or the typical marketing speak isn't going to be attractive to an AI LLM.

(22:26):
It wants to actually be able to grab things, which then leads to the third gate, which is it extractable content? So not only do you have to write content, but you have to write content in a way that the AI engine can pick up a paragraph or a few sentences and then use that to generate its answer back to the user who's written the prompt. And so we learned these three things. One, is your site reachable and can they actually see it? Two, is your content formatted in such a way that it answers the question and it looks like you follow the EAT principle. It's got to be something that is showing it's written by someone who has got expertise and experience and is authoritative and trustworthy. It loves all that stuff. It also loves its recency. If it's something new, that'll get picked up by AI very quickly.

(23:29):
But the third thing that was really interesting was that 80% of our citations we found out came from sites and domains beyond our control. So AI loves to get content from those entities that are not biased. So it likes to go to Reddit, it likes to go to review sites, it likes to go to trade magazines, people and places where it doesn't believe that is a biased answer because it wants to get that perspective that this is a real thing. So you look at that, it's got to be readable, it's got to be written the right way, and you've got to publish it in places where you don't typically have a lot of authority and you have to convince your marketing team to get there. And when you take all of this together, you begin to realize that GEO isn't a technology problem, it's a change management problem.

(24:28):
You have the people, you just have to change how they're programming the website, how they're writing the content, and where they're publishing the content. And it's those things that are actually a bigger change. I wish I would just say, "Just implement these MCPs and you're good." No, it's actually teaching your people that they have to do things differently to succeed in GEO.

Jamie Clapper (24:48):
Yeah, I think that's great. And I think that draws a good point because I think a lot of the content teams have always done things a certain way, and this is such a shift. It is forcing them to go against what they've been trained in a way. And I think there is a fine line, and maybe correct me if I'm wrong on this, of balancing how you're writing your content because while FAQ style information is great for LLMs, is it great for the everyday shopper? Is it exactly what they're looking for? It's like the balancing, I used to call it the balancing when I was in home improvement between the DIYer, the weekend warrior, and the true contractor. It was like you wanted to provide enough information for both of them, but you didn't want to offend the true trained professional that did this for a living, and it's how do you balance that?

(25:40):
And so I guess what recommendations would you give for a content team? Is that proper training? Is that testing and learning? Everyone assumes that you can just use AI to write your content, and I think there maybe is a level of that, but I think it's mainly because it's an efficiency thing, and that's where everyone always thinks about AI first is how is this going to create efficiency? But it might not always be the right answer. So any recommendations for our content listeners out there?

Steven Javor (26:13):
Yeah, I mean, you want to fix in this order, you want to audit your robot's TXT tag and your gated content, make sure that, like I said before, that your site is visible. Maybe put an HML gateway page in front of every critical PDF with your top specs in the front so it's easy to be grab, fix your titles and your schema, and then only then rewrite content. So you want to diagnose before you prescribe. Half the fixes are really just subtracting things. So AI can't really recommend what it can't read and it can't read a PDF behind a form. So start with making sure it's visible and then fix the basics before you jump into rewriting all of your content. Because I believe in the future, the future is maybe now, is that when you now design a webpage or you're designing a new section or you're introducing a new product, before it was what's the unique selling proposition?

(27:07):
How do I build the case and create the great images and put in the right tables? Now it's like you write prompts. What is the prompt I want this page to be the answer for? And then you see who's winning that prompt right now, what kind of content it's looking for. And then you work backwards from that and say, okay, what do I need to win the prompt that's really going to create this page? And I agree with you, it's sort of like now I'm beginning to lean towards the fact that you're going to be designing two different experiences. You're going to be designing the human experience for the people that are going to go to the page who maybe want to see some of the images and see the validation of your product. But then you're also going to have this agent version, which has got all of the facts, all of the FAQs, all of the support information.

(27:59):
Because what we are recognizing now is that the traffic to websites are going down. Your conversion rate may be rising, may be continuing, you're not losing business, but you're losing traffic. And so maybe measuring web traffic is not going to be as important in the future as it has been today as a key KPI.

Jamie Clapper (28:19):
Absolutely. No, that makes complete sense. And it's one of those things, it's funny, in my experience, in my role, it's like a love-hate relationship with AI. There's times I'm like, I don't want to hear about it anymore. I don't want to talk about it anymore. And then ironically, there I am at night Googling something, which essentially becomes me on AI or me on ChatGPT asking them Google lens, taking a picture, all the things that I'm doing that essentially are backed up by AI. So let's talk about governance because it's one of my favorite things that we talk about, and I think you do it so well that I think this is valuable to our listeners, but you've pointed to governance, an executive cheerleader, and one source of truth as prerequisites for AI success, plus obviously training the whole organization to use it. Kudos to you.

(29:13):
But drawing on the Schneider Playbook, what does that AI governance and enablement effort actually look like from a day-to-day perspective, just to give our listeners a thought on how often, because again, we talked about this at the beginning, it's shifting all the time. So what does that look like for them?

Steven Javor (29:32):
Yeah, it's a combination of things. I mean, really you have to find a leader that is willing to embrace the future. And I don't think as we're using it, our leaders are using it as well typically. So like I said, it's not hard to get their attention once you start doing some pilots and explaining to them that you've written some prompts and here's what's come back from it and we should figure out how to do something about it. So you want to find someone that's at a certain executive position that is going to back you up on that, that is willing to sponsor you going forward. And then when the governance, as you go through your pilot and you start updating them about what you're finding and what you're learning, you then set up a regular cadence to find out where we are. And this is where working with a partner works really well.

(30:28):
There is a bunch of them out there that can help you figure out where you stand versus your competition, where you're being cited, where you're being recommended, what is the sentiment that's actually happening? It's one thing to actually get cited, but are you being cited positively or negatively? You want to be able to figure that out. And then when you get through that, the governance will start building in. You understand that you're going to build a strategy from that, you're going to work with your leadership to understand that they are going to support that strategy or the strategy is in line with their strategy. And then you'll get into an operating rhythm because what we're finding is that this is kind of a loop. You're going to set up these prompts, you're going to tag them, you're going to then measure them, you're then going to learn from that measurement, you're going to then update content or work with third parties to get their content aligned with your voice and then see that citation rate go up.

(31:28):
But then the next week you're going to do it again or the next month or the next quarter because it's constantly changing. And this is the big change that we're seeing in marketing is that, especially in product marketing, is that you're going to need to write these prompts and measure them on a periodic basis. And we started off by writing these prompts for Canada and the United States, but depending on what the prompt's about, buying one of our products in Toronto is going to be different than buying that product in San Diego. So you need to have this localized as well, at least the version of this localized, depending what you're trying to do, especially trying to improve your conversion rates. So you set up your topics and prompts, you measure your visibility, you diagnose your gaps, you try to fix what you can, you then remeasure.

(32:19):
And this is owned by that cross-functional group. And you'll see that this is the way that I think marketing teams are going to start working in the future is that they're going to understand how relevant are we with these agents and these prompts because it's moving out of being... A lot of times what companies will do is, "Well, we already have an SEO guy. We'll just give the SEO guy the GEO." But this is much broader than that. This is involving your sales teams, your marketing teams, your service teams. All of them have a responsibility to getting content out there and having those reviews out there. There are certain things that you can fix and you'll find out, yeah, this is an issue because we're getting some negative reviews from this particular Reddit file or third party website. But if you start seeing negative things that are going across many sites, then maybe you have a reality problem.

(33:19):
Maybe your product does need to be redone because you can't just go ahead and just tell them to fix their reviews. Governance isn't really a policy document. It's who's allowed to describe the product and where that description lives and it becomes almost everyone's responsibility. In Schneider, we're not teaching GEO just to the marketers, we're teaching about GEO to everybody who's involved.

Jamie Clapper (33:47):
I love that. I feel like there's three good takeaways. The proof is in the pudding when you're doing this and sharing those metrics and the results. Two, you can't just throw this on one person's shoulders and you can't have them being the only one responsible. Three, it's everyone in the organization that's going to make it work. I agree with that wholeheartedly. So for any of our leaders that are listening here, take that into account because nine times out of 10, as we said at the start, those teams are small but mighty, they're lean, and so we want to make sure that if you want to really, I think, give it the manpower that it deserves, it's that cross-functional alignment that helps support. No one has ever done anything great in a silo, so it's incorporating those teams. Okay, one final question for our listeners because when we connected, I love the fact that you teach e-commerce at the University of Toronto because I know when I was in college, which was just not that long ago, I did not have the ability to learn anything in e-commerce.

(34:51):
I love that you are out there helping the next generations learn this.

(34:59):
I think you have done a bit of study that shows how Gen Z and millennials buy, including that they don't trust branded messages without third party validation. What are the most surprising findings from that research and what should B2B marketers be doing differently because of it?

Steven Javor (35:17):
Yeah, it's kind of funny because I believe that most of the people that are working in these frontier AI models actually come from the Gen Z millennial generation. So for some reason, their instinct is to distrust the brand and check their peers, read reviews and independent sources. It's literally encoded about how AI engines work. As I mentioned before, about 80% of our citations came from these third party sources and AI looks for consensus across all these independent domains before treating a claim as a fact. The young buyer does the same thing, so they've kind of converged with AI.

(35:58):
What B2B marketers should do differently is maybe treat the channel and review our ecosystem as citation infrastructure, not as a sales afterthought. You really have to give your partners factual content kits instead of marketing copy and make name customers with quantified outcomes, what are your most valuable assets because that's where they like going. They want to see third party unbiased reviews as a marker that this is a product that they can trust. My daughter's Gen Z, and I ask her how does she decide? She goes, "Yeah, I'll go to a brand just to see whether they have it, see what their selling proposition is." She doesn't use those words, but told me that it's going to do what I needed to do, but then I will go to and ask my friends, then I will go and ask third party. I won't make a decision just based on what the brand is saying.

(37:02):
AI is kind of the same thing. It's not going to trust your branded copy, it's going to trust what others are saying about your brand or about your product. The answer to they don't trust isn't really better brand messaging. It's really being boringly consistently factual everywhere that you can't control the message.

Jamie Clapper (37:26):
That's fair. No, I love that. I'm notorious for searching things on TikTok. It's gospel too, which I always said I would never do, but here we are.

Steven Javor (37:37):
Funny thing is that she also said, "I don't trust influencers because influencers are being paid sometimes to..." So they're being really selective about who they trust and who they don't trust.

Jamie Clapper (37:50):
I love that. I think it's interesting, and I think it will be, especially within your line of industry that you're in and different verticals, I'll be curious to see how this continues to evolve based on just what has typically been the shopping behavior, I would say, within the industrials and home improvement side of the business. And that's something that I'm kind of wildly watching because I'm curious to see how it plays out because I think to your point, there is this assumption that shopping behaviors won't change, but we know that demographically, younger generations are shopping differently, and so I think that's going to play tremendously into how organizations continue to evolve in the digital space.

Steven Javor (38:35):
And it's evolving so fast. Like I said before, companies tend to adopt the new customer journeys usually a couple years after. I mean, this is happening so quickly right now. I mean, I just started playing around with Muse, the new Meta program, which allows you to buy things online. It actually functions as you. If you're willing to give it access to your credit card, it can start actually buying things for you. I'm at that point where I'm not sure if I trust it or not, but I do remember the late '90s where I built my first e-commerce website and I couldn't convince customers to put their credit card online. Now you're going to be giving your credit card to an AI agent that may make the purchase. Then how do you measure that as a brand? Because you're not actually selling to a person, you're now selling to an agent.

(39:24):
That's the business model for all these frontier companies is that they want to be able to actually not only create the relationship with the prompter that this is the best product to buy, but then buy it for them.

Jamie Clapper (39:36):
Absolutely. Well, Steven, I think you just gave us the next topic for when I have you back on. We'll expect a full report on how Muse is working and if your credit card was stolen or not, as well as how you were able to track those purchases. Who knows? Always great connecting. Thank you so much for taking the time to sit down with me. I know I thoroughly enjoyed the conversation and I think there's a lot of great takeaways for our listeners.

Steven Javor (40:05):
Great. Glad to help.

Jamie Clapper (40:07):
Thanks for tuning in to Unpacking the Digital Shelf B2B edition. To keep learning from the best in B2B, subscribe now and join our community of leaders by becoming a member of the Digital Shelf Institute. Thanks for listening and we'll see you next time.