Earley AI Podcast
In this podcast hosts Seth Earley invites a broad array of thought leaders and practitioners to talk about what's possible in artificial intelligence as well as what is practical in the space as we move toward a world where AI is embedded in all aspects of our personal and professional lives. They explore what's emerging in technology, data science, and enterprise applications for artificial intelligence and machine learning and how to get from early-stage AI projects to fully mature applications. Seth is founder & CEO of Earley Information Science and the award-winning author of "The AI Powered Enterprise."
Earley AI Podcast
Earley AI Podcast - Ep 75 Why Data Quality Matters for AI and Digital Maturity in B2B Enterprises
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
In this episode of the Earley AI Podcast, host Seth Earley welcomes Eric Rehl, Vice President of Digital Customer Experience in North America at Schneider Electric. With over 25 years of expertise in digital strategy and customer experience, Eric has guided global organizations through complex digital transformations, always keeping business outcomes and customer needs at the core. Drawing on his deep industry knowledge, Eric shares how large enterprises can move beyond buzzwords like “digital transformation” and “AI,” instead choosing a pragmatic, data-driven approach to drive real business value.
Join Seth and Eric as they discuss the evolving role of digital capabilities in business strategy, the foundational importance of high-quality data, the unique challenges faced by B2B organizations, and how AI can power truly personalized customer experiences—from the ground up.
Key Takeaways:
Digital transformation should be rooted in business outcomes, not technology hype; focus on the “so what” for your customer and organization.
Strong, clean, accessible data is critical for scaling digital experiences and enabling AI-driven personalization—without it, even the best tools will fail.
B2B companies often lag in digital maturity due to legacy data architectures and complex customer relationships, but can catch up by investing strategically in foundational capabilities.
A robust digital journey relies on operationalizing and continually improving product and customer data, rather than one-off fixes.
Maturity in B2B digital experiences evolves from simply “doing no harm,” to enabling ease of business, and ultimately leveraging digital platforms for growth and commercial impact.
AI’s promise lies in moving from segmented personalization to real-time, dynamic customer engagement powered by integrated data and knowledge.
Preparing for AI-driven customer discovery means syndicating high-quality, semantically-structured content across channels—both on and off your own domain.
The next frontier is operationalizing knowledge (not just product or customer data) to fuel AI tools for differentiation and problem-solving.
Continuous experimentation and responsible opportunism allow organizations to discover new outcomes and business value.
Insightful Quotes:
"I think as you start building maturity, you're learning how to orchestrate those pieces. You're getting more of that harmonization of organizing principles across those disparate departments, across knowledge and content and customer experience and product information. And so that becomes kind of the holistic journey that you're thinking about." - Seth Earley
“We always start with the outcome. Like, why are we talking about capabilities here? Why are we talking about AI? What are we actually going to do with it to get to what the business outcome we’re trying to drive or the experience outcome we’re trying to drive?” - Eric Rehl
Don’t miss this in-depth conversation packed with practical advice and forward-looking insights for anyone leading or navigating digital transformation initiatives in the AI era.
Links
LinkedIn: https://www.linkedin.com/in/ericrehl/
Website: https://www.se.com/us/en/
Thanks to our sponsors:
This is the Early AI Podcast hosted by Zephyr. Join us as we delve deep into the passions, expertise, and experiences of thought leaders and practitioners to talk about what's possible with artificial intelligence. The Early AI Podcast is sponsored by Early Information Science, your digital transformation journey with design and deployment of innovative technology solutions as well as Better. Now enjoy the show.
SPEAKER_01Well, welcome to today's Early AI Podcast. I'm your host, Seth Early, and in this episode, we're going to be talking about how enterprise can move beyond the hype of digital transformation. Digital transformation has become a catch-all term. It's lost its meaning. And then you add AI in that, and you say, AI is so broad and ambiguous and entails so many different things. Of course, people think of AI as generative AI these days, but it's more than that. But people think of AI as generative AI, and then they think of digital transformation as this big, amorphous, ambiguous thing. What we're trying to do is get past that and really talk about the fundamentals of how we can drive results and not just focus on the broad, ambiguous catchphrases, right? We're going to take an ambiguous approach and we're going to add another ambiguous approach to that and then try to get specific results. So we're going to try to get beyond that. So real success comes from building on the foundation of data, on a process, on understanding of customer needs. And so today we're being joined by an Eric. I'm sorry, I should have asked you, is it real?
SPEAKER_02Is that the way you pronounce your Rail, like Railroad.
SPEAKER_01Eric Rail. I should have, that's one of the things we usually ask right before we start. Vice President of Digital Customer Experience for North America, Schneider Electric. He has extensive experience in digital strategy and customer experience. He's helped global organizations really transform the way that they engage with customers and leverage technology for value creation. So, Eric, welcome to the show.
SPEAKER_03Thank you, Seth. Thanks for having me here. I look forward to this discussion. It's certainly a target-rich topic. So thanks for having me.
SPEAKER_01It sure is. And there's so many things that are changing it so quickly. So you've been around this space for the past 25 years around digital strategy, around customer experience. How did that journey kind of shape your philosophy on how organizations really need to approach this writ large? What is your thinking and how has that kind of evolved? And of course, in the last few years, a lot of things have made that much more complex. But tell me a little bit about your philosophy and how you you think of transformation and experience, Josh.
SPEAKER_03Yeah, I think that, you know, over that 25-year period, I've watched the space, the sort of the digital space evolved in the place it occupies within these companies, especially big companies like Schneider. You know, it for years it was something we have to do in the sense that, well, we just need to be present there, right? And so it was kind of off to the side in terms of let's have a website and a few other things. But it really didn't, it really didn't, it wasn't kind of an organic part of the company's operating model, right? And so what you've seen happen over the years is it go from just something they did to something that, okay, now we can really see how these kinds of capabilities can enable some parts of our business, right? Where they can they can um kind of extend our presence in the market, right? We can get amplification if we're part of this digital ecosystem. We have some stuff out there that people can consume on their own. To where it is now, which is really um it's uh especially in companies like Schneider, it isn't the business strategy, but it is a key enabler of the business strategy. But only and you make you made a really good point in your opening comments, Seth, that digital transformation is a buzzword. It is business transformation, right? That is digitally enabled. But none of that matters unless it's results and customer driven, right? So it's not transformation for transformation's sake, it's transformation to realize some business or customer experience outcome, right? So the space and we turn to digital capabilities and now more increasingly to data to figure out how we can get to that result in a way that is going to not just meet our customer expectations, align to our business strategy and create impact, but also differentiate us. So how that shaped my philosophy is that we always start with the outcome. Like, why are we talking about some why are we talking about capabilities here? Why are we talking about AI? What are we actually going to do with it to get to what the business outcome we're trying to drive or the experience outcome we're trying to drive? So that's how we always approach it. And so I'd say that's kind of my philosophical approach to to what is now just business activation, you know.
SPEAKER_01And and it it's it's regardless of the technology, and it's regardless of what the digitization might be or the digital strategy, it's saying, what is it? It's the so what. I always say, so what? What if we have this information, so what? If we privacy capability, so what? What's it gonna how's it gonna impact people? How's it gonna impact how the customer does business? How's it gonna impact what they do, their efficiency, their effectiveness, internal, etc.? So I totally agree with you. It it has to be really brought down to that detail of what's gonna happen on the ground. People are interacting with our organization. And with that, you know, you talk about data. And I'll tell you a quick story. I was talking to, and and B2B has been lagging, right? B2B has kind of lagged behind. It's based on you know long relationships and kind of embedded products and product design wins and relationships with other larger enterprises. And so it's been slow to evolve, and and of course, the data has always been challenging. Uh, the other day I was talking to a B2B company and we're talking to their digital team about the the need to resolve their data. They've grown through acquisitions, they have lots of different catalogs, a broad range of products, and their data was a mess, right? And their their high equity was a mess, the product information was inconsistent, user experience is inconsistent. And so we talked a lot about that and talked about how to remediate that and how that would then enable things like conversational access and chat bots and and different ways of retrieving information. And then I got a note from them a couple of weeks later saying, Oh, we've decided to go in a different direction. They said, Oh, what are you doing? He said, We're gonna build a bot. It's like, okay, you're gonna build a bot. How would that be how likely is that going to be to solve their problem? It's not, right? Unless you're saying the bot is kind of the Trojan horse, and really we're gonna be fixing data, right? Because you cannot do that without the data. So let's talk about the data. You know, when you think about the role of data in that experience, it's all based on data. The whole digital experience is comprised of data. And so, how do companies think about you know, consuming and owning and orchestrating data to create that personalized experience, right? Because many times they're getting data from suppliers or they're getting data from upstream processes, they're getting data from uh their customer behaviors, they're getting data from lots of different systems and touch points. Uh, yet many times this is an area that's been underinvested in. And now, you know, to various degrees, uh, organizations are saying, hmm, maybe this is something that requires a little more investment, but it's usually they have to fail first before they they really come to that point. But what are your what are your thoughts there about how that's kind of evolved and what's happened over the last uh several years? So maybe you could just riff on that for a little bit.
SPEAKER_03Yeah, I mean, I I've been wrestling with kind of the notion of and the constraints around data in large B2B enterprises for some time now. And you know, I when I think about not a new problem. No, certainly not. I was gonna say, when I think about and I when I think about especially with like manufacturers, when I think about their progression through this very long journey that we've all been a part of, they started behind the eight ball when we started talking about leveraging data for customer experience and for digital engagement and things like that. Because if you think about, you know, 20, 30 years ago, what kind of the data architecture for a large for a manufacturer was it's I gotta build a product and I gotta ship it to a site, not to a person, to a doc, right, or to a door, and and I have to sell this through a person that is going to then enter this the sale, and you know, and the name on that on that sale is very likely not to be the name of the person I sold it to, right? It's the procurement manager or somebody who's responsible for processing the transaction and making payment. So that entire architecture was great when you are um you know kind of selling face to face and you didn't have to think about how am I going to extend this customer relationship behind beyond the sale or how am I going to digitally enable it to be more personalized. You know, the entire kind of back-end architecture of uh the back-end data architecture was misaligned to, and not on purpose, but it was it was not aligned to what we need today out of that kind of data. So, you know, that's why I think these companies tend to lag and they kind of had to completely turn their thinking upside down in terms of how they manage customer data, right? And what it really meant to what customer data really means to them. And so now to fast forward to today and your your example around the bot, I mean, I don't I discourage anybody from doing a bot unless it can really help. You know, I mean, how frustrating do we all find it when you're interacting with a bot and it only can take you one step into the conversation and it's as you know, was this helpful? And you know, it's like, I just want to talk to somebody, please. You're not helping me. So having very sound practices around acquiring data, structuring it and cleaning it in a way that it is gonna specifically support outcomes you need, and investing in the talent to be able to do that and scale that, I think is is the journey we all need to be on. Right? Because I'm a I'm a consumer of data. I own product data, I don't own customer data, but I have extreme dependencies on good customer data for me to be good at my job, right? And this is something that and deliver the outcomes I need to for for my company and for our customers. And this this is a journey that I don't know that we're ever gonna be done with, right? I mean, there's always gonna be a a new use for this stuff, a new opportunity. And and I don't know that anyone's totally cracked the nut on it, you know. So I think you cut you kind of have to take it, align it to your business strategy and your customer expectations, and that's where you start.
SPEAKER_01And and all of that is a moving target, right? All of that is absolutely evolving with the competition, it's evolving with changes in the product space and the solution space and technology and customer needs and the competition and routes to market. All of those things are you know, they're multiple cheap pieces that they're that are changing uh all the time. And and people used to say, well, when will the taxonomy be done? I said it's things are done with sales, manufacturing, and operations.
SPEAKER_03Right, exactly. It's not gonna be that that makes the product makes the products and gets our salespeople paid. That's always the priority.
SPEAKER_01Exactly. And and of course, it needs to be operationalized and it shouldn't be a big giant lift every six months. You know, we used to have customers saying, I want to solve this problem every six months, right? Now it's that's a there's a difference between solving the problem every six months and putting into place a solution that will continue throughout, you know, the next six months, 12 months, 18 months, right? Because it becomes woven into the processes as opposed to a big cleanup problem, right? Or a big mess. But the problem is when you solve it sometimes without that that view of this being operationalized and the long-term integration woven into the process of the organization, then it does become a big mess again, right? Or it does become a big mess with the next acquisition. So I think that's it's really important for organizations to realize that this becomes a core competency that has to be built and maturity has to be built over time. And you need the systems and tools and processes to to make sure that you can curate and manage and update and uh remediate and and constantly keep that data quality where it needs to be, but also have it be adaptable enough to changing circumstances and changing needs.
SPEAKER_03Absolutely. No question, no question. And I think there's the so I very much believe in the basics. We like to talk about the activation and the AI and all the cool things that we want to do in the front end of delivering these journeys to to the market and engaging our customers in this why this way, but all of that is is dependent on a basic foundational capability and competency here, right? And it's it's the velocity at which you can execute and the scale at which you can then deliver and sustain is completely dependent on the basics of data management, having good, clean, current, accessible data and processes in place that allow you to keep it that way, but then be agile enough as new use cases come on board to be able to activate and leverage that data, right? And so, but you never ever can get away from sort of that basic competency and foundation because it's the underpinning for anything you want to do that requires data to be effective and scalable, right? You can do anything in a vacuum and in a pilot. I can take I can take a slice of data out of my out of our customer database and out of my PIM, and I can run wild, but there is no way I'm gonna get to any point of scale or value creation if I don't have a solid foundation on which to to which to build that, a data foundation on which to build that.
SPEAKER_01Absolutely. You know, brings to mind there's a a story I want to tell, but I'm trying to think of how I can real time uh uh synthesize it and and and um uh and uh sanitize it, right? But a colleague of mine was in the um in the space where it was B2B, it actually had to do with, I don't think I'm I'm sharing anything, I shouldn't, um, domain names, right? Domain names where the the transfer of domains was very complex, right? There were no standards and there were no ways to do that. That was a very manual process. And building those standards allowed for those to build this enormous business, this $500 million a year business in selling premium domain names by automating many of those processes and by building those data standards, because before that it was very manual, right? And it took a lot of work to do that. So the soon, as soon as you have the ability to have those interoperability standards, you have the ability to optimize and to scale. Now, that said, that provides efficiency, but it doesn't provide differentiation. Differentiation is then what do you do with those standards, right? What do you build on that that is that is uniquely applying that data in a way that is meaningful? So in in the case of B2B, we do need some standards, right? Of interoperability. We need some standards around product data and you know doing any type of transactions. But then we need to think about well, what is the additional layer of differentiation that I can get on over that so that I can truly stand out from my competition? And that's part of a maturity curve, right? That's part of a maturity level because the basics is well, let's just get this stood up, right? Let's just get this so that we can, we're not spending inordinate amounts of time on the basic things that are wrote and repeatable. Once we do that, now we have to build on that. Do you want to talk a little bit about B2B maturity? And and you know, you mentioned that you know you're you're kind of in the middle of that maturity journey at Schneider, but you know, where is it going? What does that maturity look like? And then what are the stages as organizations kind of traverse that that um maturity curve?
SPEAKER_03Yeah, for sure. I you know, I you can I mean, all you need to go is do is go to Google or your preferred AI search platform and and look up a digital uh uh maturity curve for B2B, and you'll get many, many different flavors of that, right? Um I have one that I use for myself that's a little bit different, I think. And and there's three stages along that maturity curve, and it breaks down in between there, of course. But the three primary stages on the low end of that curve, and this is like digital, B2B and digital, right? Or digital and B2B is do no harm. Like that's the basics, right? Just make sure that your data's clean. Don't screw it up. Don't make life difficult for your salespeople, don't make it difficult for your customers, right? Get the basics right. If we're gonna talk product data, have a product data, had a product catalog on your site that has complete, current, and accurate information, right? And because honestly, for for Schneider, about 75% of the traffic to our our sort of owned digital footprint goes to our product experience, right? They want product information, so it just needs to be right and easy to get, all right? That's kind of the do no harm stuff. So don't mess it up. In the middle is what I would call ease of doing business, right? Which is all right, how are we gonna make sure our customers have a job to do? Right? How are we going to make it how are we going to enable them in that job and make it easier for them to be a customer of Schneider and be effective in the delivery of whatever it is they need to do? So we need to be easy to access, easy to get support. Again, it builds on the do no harm, good quality, accurate information, but how do we create a user experience for them and a journey for them that just is seamless and makes it simple for them to engage with us and get what they need? Right. So that's in the middle. And and and then on the on the on the high end of that continuum is what I would call uh commercial impact or growth, right? Now, what are we using our digital, our digital platform and our digital go-to-market for to actually grow the business, right? Get new customers, have a really solid acquisition engine in place that brings in and captures demand and qualifies it and creates actionable insights for the sales organization so they can go off and pursue that, right? How are we, how are we not just giving good product information for our customers, but how are we actually helping them specifically understand if it can meet their needs through like configuration, then go get a quote, right? And do all that with it, you know, instantaneously. How can we give them a very robust suite of self-service business process capabilities that allow them to do a lot of things on their own without having to call our customer support team all the time, right? And so you move a lot, and and you can be on this journey, on this continuum at all of those stages at the same time. I think it kind of depends on what you're talking about. But from a strategy standpoint and a priority standpoint, I think you want to move sequentially along that so that you so that you make sure you're taking care of keeping the lights on, right, while you deliver innovation that is going to create differentiation for you in the market in terms of being a good partner, but also allow you to drive that impact and that growth by by engaging the market and understanding the intent there and acting on it. So that's kind of how I think about it. And that's how we prioritize. And then the innovation comes on the right side there, right? I mean, we've got some projects running right now that are really looking at how do we take clicks out of the journey and how do we accelerate kind of our end-to-end lead management capability so that we can we can act on insights faster.
SPEAKER_01And and and you know, at the end of the day, you're trying to help the customers solve their problem. 100%. And you're trying to provide the tools and the capabilities and the knowledge and the resources for them to do that so that they build a stronger, so you build a stronger relationship with them. And they say, Oh, I gotta go to Schneider because they're gonna help me solve my bigger problem. I'll be able to generate this quote very quickly, I'll be able to get these other tools and products. And and and I would imagine, you know, there's a certain number of products that like what you sell, and correct me if I'm wrong, you're selling other manufacturers' products as well as your own, correct? Yes, most of you, yeah, you sell out so you are a distributor.
SPEAKER_03We are a manufacturer, so we are we are making what we make, what we sell, we make. We have partnerships where we sell companion products and things like that, but most of it are so part of it is saying you want that complete solution.
SPEAKER_01It might include other manufacturers and you want to bring that to bear, right? And so part of it is also saying, what's the knowledge? So whenever I talk to organizations and I'm talking to say the product data folks, they need to talk to the people who are doing customer service, customer support, you know, customer enablement, knowledge processes, right? Because once you have perfect product data, well, what else is part of that journey? Right? It's the knowledge, it's the self-service, it's the solution to those problems. And of course, you also need that customer digital body language. You need that customer data to say, well, what are they doing? Who are they? Where what stage are they in? What's their mental model, right? Because are they an engineer or are they a procurement manager? All of those things go into that complete end-to-end experience. So it is really kind of an orchestration of all of those pieces. And I think as you start building maturity, you're learning how to orchestrate those pieces. You're getting more of that harmonization of organizing principles across those disparate departments, across knowledge and content and customer experience and product information. And so that becomes kind of the holistic journey that you're thinking about. And with that, there's the ability to personalize that experience. And so one of the things that we had talked about was, and when you think about this with AI, AI has a promise to take personalization because this is what you need to do it at scale. If you had to hire one salesperson for every customer, you could have a nice personalized experience. You have to have the digital equivalent of that. So, how do you see AI as being able to provide that more dynamic and real time customer engagement compared to sort of segment based approaches that a lot of organizations rely on today? So, talk a little bit about how you see that. That scaling up and being operationalized and being done at that level that we need to uh achieve. And I imagine there's various stages of maturity around that in and of itself.
SPEAKER_03Yeah, absolutely. You know, the the where I think this is all going, at least my kind of conceptual vision of all this, and and I don't think this is particularly unique to me. I think this is everybody's kind of thinking about the same thing, is that you know, we're gonna get to a place at some point where the experience that a customer has on our website, right, or or uh uh in any given kind of digital engagement we have with them will be entirely personalized, right? It'll be created on the fly. Um, it will it will build and personalize the immersiveness of that personalization will increase the more we the deeper that engagement goes, the more we interact with with with with people to the point where no customer, no user actually has the same experience on our website, and the whole notion of like a home page or a landing page goes away, and upon receiving that, upon getting that engagement, everything is just kind of constructed on the fly by AI, right? But the requirement for that is good data, so the ability to kind of understand the individual, who they are, where they are, are they an existing customer, are they a potential new customer, what is their interest, what is their intent, spermographic, demographic, personal, you know, all that stuff layered with or combined with the knowledge that you're talking about, right? So if I start to build some sense of who this individual is and what they're, you know, what might be of interest to them, I need an entire knowledge set and language model that is going to be able to very quickly analyze and understand that and respond to it. And the the the immersiveness of that response is gonna be based solely on the quality of the data that we understand about this person and the quality of the content that we have to deliver back to them, right? But in the middle is an engine that is always on, that is looking at who we're engaging with and saying, all right, what is the it's almost think of it like what you have like like customer, uh a lot of this customer experience software for like customer support we see out there, next best action kind of stuff, right? And some of that, you know, you're we're gonna start by prescribing some of that, right? So there's it's a little bit of a combination between what you said earlier of kind of segment-based approach with pure personalization, right? You need to orchestrate a little bit of that. But as we get more effective, it's gonna be more agentic and it's gonna be more conversational, it's gonna be more real time.
SPEAKER_01Yeah, that's a great point. And I and I like to say you have to have the handcrafted artisan experience at first, right? Because you need a framework, you need a model, you need a template to say this is how we need to interact with these customers. So you frame that and you build that, and it is handcrafted, right? It's based on human judgment, human insight, knowledge of the customer, knowledge of the solution. And then once we have those parameters, that's when we can start to riff on that. That's when we start to turn this over to uh multivariant testing, right? And the ability to vary those different parameters from those different experiences and come up with those next best actions, the next best piece of content, next best product, whatever it might be. But you made a really good point to say, so we're not just invoking AI, right? Or generative AI. We're invoking it based on our differentiated knowledge and expertise, right? It's like taking your very best engineer and saying, I'm gonna give my very best engineer to this customer to help them solve that problem. That is great. Well, you can't afford to hire an engineer for every customer, but you can build that expertise from your knowledge base, from your technical specs, from those documents that are that are being created in engineering upstream, but then don't always make it out to the user experience, right? Because you can't anticipate everything. You may have attributes in your technical specifications that nobody anticipated that a customer is going to need a facet based on that, or an attribute, or a bullet point. And you can't just inundate everybody with everything, right? So you have to be selective and then say, what's a procurement manager care about versus a design engineer versus a maintenance engineer versus somebody who's troubleshooting something. So that all is based on that organizational expertise and that that knowledge of customer solutions, engineering, designs, all of that, which is proprietary, right? You don't want that out in the public, um, in the public domain. You want, and that's where things like retrieval augmented generation come in, right? Where you're saying I have repositories of content that are curated, that are structured, that are componentized, that a an agent can access and then surface to the customer in the context of all these other things. So that is a really wonderful vision, right? But we're still kind of far from reality on that because of all the problems that organizations are having. Everybody kind of feels like generative AI and RAG are a solved problem, but they're not. It's not. It's not from it. And so it's scratching the surface. So a lot of this is it's also started with uh those off-domain platforms, right? So many people are beginning that. So you can use it once you have people, and you can use it to to uh enrich their experience and to provide and anticipate and to solve problems based on your knowledge, right? Based on your your proprietary knowledge that that is competitive differentiation and competitive advantage, wonderful. But then that journey many times is starting off domain with a chat GPT, asking those questions. So there's a certain amount of information you do want out there, right? Because it's so important. So, how do companies how should they prepare the prepare to influence the discovery and those consideration phases that are taking place off those digital properties and less about SEO and more about what are people are calling GEO or generative engineering?
SPEAKER_03Yeah, we're a word increasing. We've we've decided that we just this morning I had a meeting about this, and we we said, you know what, let's call and we were calling it GEO internally. I said, let's call it AI search. Okay, AI search. Yeah, we let's call it AI search because AI search, when you think about kind of the flavors of that, it happens on domain and it increasingly, to your point right now, it happens off domain, right? But you know, the so we have some work running right now that is looking exactly that is looking exactly at this topic. And I'll give you just a quick uh kind of uh story just to that I use to tee this up internally, is I you know, one of the things we do at Schneider is we make commercial great electrical panels, right? So I have a plant or a hospital or something like that, and we make panels to run to manage the energy in that place. And I went to Claude, and I my first prompt was um build me a bill of material, create for me a bill of material for a uh commercial electrical panel for a hospital. Pretty simple prompt, right? Really basic. Well, what it returned was a full priced bill of material which matched very much um a kind of a typical bill of material we would see uh when we have this kind of project. Full full bill of material, fully priced, with all the comp you know, it's a hospital, all the compliance and regulatory stuff in there too. So then I said, all right, you know, where should I buy this? And or where would I buy this? And it lists like all the distributors that you you know that are the top distributors to buy this from and and and why, right? And well, who makes the best product for this? And then it lists like its interpretation at least of the top three products. Thankfully, in this case, Schneider was number one, but it's not always gonna be, right? And so, but that I did that in like two or three minutes, right? And and so it really kind of you know created some urgency on the depth and the breadth that these tools are able to go to in their analysis and recommendation of this stuff. So the point is is these tools and platforms all have preferences for how they like to see this information and and how they want it formatted and framed so that they can easily put pull it, right? QA formats, um, structured data, you know. Exactly. Summaries, summarization.
SPEAKER_01That's right. Things that are semantically meaningful. Go ahead.
SPEAKER_03Exactly. Exactly. So very can quickly make determinations about what this information is and if it is aligned to the to the to the query and the intent behind the query. So we're doing a body of work now where we're kind of um playing with these platforms and using some tools to look inside of them and say, all right, you know, what kind of are the are the are the um some of the top prompts around our product portfolio or our space right now? Where are these tools getting that information? Where are they sourcing it from? What are the citations in the responses that they're giving? And how can we then look at that and look at our content and look at how we have it formatted, how we have it set up, how we have packaged it? Because a lot of times the responses it's presenting is pulling content from all over the place, right? It's not just going to like a web page and saying, okay, there we go. And so how can we package this up in a way that is going to align to the interests and preferences of these tools and influence, frankly, the share of citation that Schneider gets when it is returning a response around a given prompt. And that is in both content that is on our own domain, but also on third-party domains. Because when we look at the data that we look at in terms of where where uh uh these tools are generally sourcing, it's usually from third party domains. It's not the domain of the of the particular topic or company um that it that is being prompted. So that's a huge part of this, is what happens off domain versus on domain.
SPEAKER_01And are those coming from uh content that you're syndicating to say a distributor?
SPEAKER_03That is yes, that is part of it. So we have you know, there's an in being really intentional about where you're trying to influence publication of content off domain on distributors, but also just flat out third party websites that operate in our space that are deemed authoritative. But then, you know, how do you continue to to influence other third parties in in creating content about Schneider? Right? So it's like bringing in our our comms and our marketing partners to say, hey, this is what the data is telling us in terms of where these tools are sourcing content, go out and engage with these folks and get them to do more about us or seed content for them, right? Give them stuff that they can then publish on our behalf. So a big a big piece of this is us not necessarily creating the content ourselves, but getting other people to create it for us.
SPEAKER_01Well, and that end when you componentize content, when you take that long monolithic, you know, spec sheet or text uh doc or or you know product manual, whatever it is, and you break it up into semantically meaningful chunks, uh, that actually is good for your own retrieval on site. 100% shopping assistant box or your own QA. It's good for your search, it's good for semantic search, it's good for lots of things, and it's stuff you can syndicate out, right?
SPEAKER_02Yes.
SPEAKER_01You have uh some content you want to keep behind your firewall or for customers, but then you have some content you want to let out there push it out, just like you syndicate product data, right? You want to syndicate product content and knowledge. And that if it's in that format that the gen that the uh gen AI engines like, right, which is again semantically meaningful quest answers to questions and summaries, then they'll surface it, right? But but it benefits you and your customers and your employees to have that more readily available. So I see that as kind of that whole content operations, product data operations and product content operations that needs to think about this, right? In addition to enriching with metadata. And the metadata is also very important because that can provide additional signals for generative retrieval, right? It's it's a combination of semantic search, vector search, and parametric search. That'll get you to the very specific product that it applies to, or the product model, or the whatever it is, but it gets you down to that level of granularity and nuance to get that very specific answer to troubleshoot or to install or to whatever, you know, whatever it is you need, but it gets that very specific level of precision around that.
SPEAKER_03Well, I want to go back to your point. I you're using a word that is really important in this conversation, which is knowledge. Many, a lot of companies, especially B2B ones, are further on their journeys. So if you think about kind of data sets at a very high, oversimplified level when it comes to this space, you've got product, you've got customer, and you have customer would include account, and then you have knowledge, right? Knowledge is data, but a lot of us are further on our mature, further along on our maturity journeys with product and customer data. We're not so far along our journeys with knowledge because it's mostly unstructured, right? And so, and that doesn't align to the way that these AI tools like to like to consume data, but they eat it up because they don't, you know, Chat GPT doesn't just seek to answer your question, right? Like like Google always kind of has in traditional search. It seeks to answer your question and elaborate on it, right? Which is not just here's the answer to your question, here's some recommendations and analysis on the topic itself, right? And that's how people are more increasingly using it, and that's where the knowledge comes in, right? It's it's it's it's it's content that goes deeper about your value proposition than just feature benefit and attributes.
SPEAKER_01Totally, totally. And I think that is the next frontier, really, because when you have great product information, well, that's only part of the process, right? And you do need that expertise, you do need that knowledge, and that is the way you're going to engage customers, and that is the way you're going to feed these Gen AI tools, right? So, so that is a critical, critical piece. And you know, knowledge processes have always kind of been the redheaded stepchild, right? Uh digital transformations. I wrote an art, I've written I've written numerous articles about that saying, you know, this is something that's really important. This is like the next frontier of this. This is the thing that we don't necessarily look at consistently. And the organizations that have invested in this, especially Create Once Publish Everywhere, where you have multiple channels and you need consistency. You can't just build a bot for this camera. You can't just build content for this distributor. You have to make it enterprise-wide and you have to do it in such a way that you're creating it and then you're you're consuming it across the organization and throughout your distribution channels. But again, you do need that separation of well, what's what are we willing to share with the world?
SPEAKER_02Yeah, exactly. Yeah. Okay. That's a great point.
SPEAKER_01Yeah, you know, and and I think a lot of it you're going to want to share with the world because it differentiates your expertise, right? It differenti your your offerings and your ability to solve these problems. But at the same time, you know, there's some stuff that still needs to be, you know, uh behind the kimono or wherever you want to say it, behind the firewall. So and and and only being used for trusted customers or trusted partners.
SPEAKER_03Right. Absolutely. But I think that circles back to the results and the outcome conversation we had at the beginning. You know, if you start there, you can get some some some more uh concrete context around what should I share when and with whom, right? You can tie, you know, you can have that. That's a good context to kind of govern and help make decisions around the access to that information and how prolifically you're going to put it out there.
SPEAKER_01Absolutely. And then building libraries of use cases for different segments for different types of customers for different products, right? You want to make sure, and libraries of use cases are important because they let you test, right? Because you need to test these things on an ongoing basis. But they also need, they also become assets of increasing value because they can be and tell the organization what certain people do, what customers do, what processes there are, you know, how you're trying to enable those, how you're trying to speed those up, what the baselines are. So to me, building use cases and libraries of use cases, that's not just a one and done thing. That's something that becomes enhanced and enriched over time. Well, this has been great. I've really enjoyed this conversation. You know, thanks very much for sharing your perspectives. Any final thoughts on um organizations going down this path? Because it is it is changing priorities, it is changing how organizations think about their data, their customers, their content, their knowledge. And with all of this happening so quickly, uh what are your final thoughts for organizations going down this path?
SPEAKER_03Yeah, what I'd say is, you know, I I I spent a lot of time talking about the basics and the importance of that. And that that still rings true. I don't, I've never um kind of I will always say that. Um, but I we all need now in this where we are now with um how quickly we're moving through technical innovation and the capabilities that are that are just in kind of their nascent stage of of of being able to deliver some impact. We need to balance our our execution along foundational activities and experimentation. Like we we really need to, even if you, you know, I talked about being outcome focused, even if you don't know what an outcome is, there's capabilities here now with AI that allow you to arrive at one by accident almost. So be opportunistic, I guess is what I would say, right? Responsibly opportunistic. You still need to start with an outcome in mind, but you might not know what per 100% what that's what that might be, but I have an idea. So I would say spend some time here experimenting because the tools are there to do it, they're accessible, and you don't know what you're gonna get. I mean, it there's some very opportunistic things out there now, and we're starting to tinker with some of that stuff, with the notion that we think we think we know we're gonna where we're gonna get with this, we're not sure, but it's worth it to try. You know, and so I ever the plate is there to be able to to experiment and and and really find something, find some nuggets out there that nobody else is doing right now. So interesting time.
SPEAKER_01That's such a great point. I think you have to still have some some willingness to learn and experiment and fail, right?
SPEAKER_02Yeah, absolutely.
SPEAKER_01You can do that much faster and much cheap, much more cheaply. I really want to thank you for joining us today. I mean, this is really great. You know, you we move beyond the buzzwords, you know, to really talk about the business outcomes. And you know, for our listeners, thanks for tuning in to the Early Pay AI podcast. Stay with us for more conversations on how AI is and data shaping uh the future of business. We'll have Eric's contact information in the show notes. He's Eric Real, uh, it's E-R-I-C R-E-H-L on LinkedIn. And what's interesting is you don't have like Eric Real 52, like that. You've been on LinkedIn for a long time.
SPEAKER_02So for a long time, right?
SPEAKER_01But you you got that name. Uh, so that's great. Eric, thank you so much for time, and thank you um uh uh uh to our listeners and and thank you, Carolyn, for uh handling things behind the scenes. And we'll see you next time on the next uh early AI uh podcast.
SPEAKER_00Thank you for joining us on another deep dive into AI innovation. Tune in next time when we introduce another industry expert and discuss how to maximize AI in your world. The Early AI Podcast is sponsored by Early Information Science and Vector. To learn more, visit early.com. That's E-A-R-L-E-Y dot com. Thank you for tuning in.