Name: Powering Agentic Experiences: How To Serve Your Content to AI Agents and Platforms
Uploaded: 2026-10-27T16:30:47.617Z
Duration: 59 min 28 s
Description: Powering Agentic Experiences: How To Serve Your Content to AI Agents and Platforms

Transcript for "Powering Agentic Experiences: How To Serve Your Content to AI Agents and Platforms": Hello, everyone. I'm gonna wait a couple seconds while people start flowing in here. If you wanna put in the chat where you're joining from, I know we might have a few repeat guests from the other parts of our series. I'm joining from Saint Paul, Minnesota, Austin, Minnesota. I feel like you don't normally see the Austin to Minnesota called out, so that's awesome. Greensboro. Great. I was just in NYC over the weekend. That was a lot of fun. Great to see so many peep... Paris. That's awesome. Oh, all the Minneapolis people. We should meet up sometime. Okay. Great to see so many people from around the world. So I wanna welcome you to today's webinar. It's the third of a series of three webinars in our agentic commerce readiness series. The goal of the series is to provide you with a practical playbook to prepare your products for agentic shopping. And today, we're gonna focus in on how we serve... How to serve your content to AI agents and platforms featuring our partner, Otterly dot ai. But before we get into our amazing, host and presenters here, we are going to go over a few housekeeping notes. So this webinar is gonna be recorded, and a copy will be sent to you in the days following the webinar. On the right hand of your screen, or not on the right hand, but on the tab you're currently in, you'll see the chat. And then a couple of tabs over, you'll see, the docs and q and a section section. For questions, please put them in the q and a section. It'll help us keep everything, organized and make sure we get into any questions we don't get to answer at the end, in the days following the webinar. That doc section has, the link to the rest of the series. So if you've liked what you see today or you or you haven't seen the remaining parts of the series, you can always click there to catch up. We also have closed captioning today... Available today on the webinar in a variety of languages. All you have to do is, select the, bottom... The closed captioning CC symbol on the bottom right hand corner of your screen. You can select a language and turn them on from there. So without further ado, I'm gonna bring Nathan up from Salsify to kinda go over who our hosts are for today and get us started. Awesome. Thank you, Ray. Thank you everyone for joining. Nice to see you all. I see a lot of familiar names in the chat here. And hopefully for those who joined us for part one and two, you're familiar with my face. I'm Nathan Tuplow, director of product marketing here at Salsify. We're joined by Rick from Otterly AI as well, and Marissa will also, from Salsify will give a demo later on in today's session. So we're super excited for this session, super excited to have you all join us here. And, with that, I will I will jump in. And, again, for those who joined us in parts one and two, you've you've, seen this slide. You like to remember this slide, but I like to start here because it just kinda anchors us all on, like, what the situation is today. And and there is a new era that we're all in, this agentic era. And the question you're likely faced with as a brand or retailer distributor is where is your growth gonna come from, over the next two, three, four, five years? You still have the physical shelf in store shopping. There are still those traditional, digital shelf channels, ecommerce channels. But there's this new shelf, the agentic shelf. And it's changing the way that products are being discovered and found online. And the challenge here is that it's not like people are buying more because of this agentic shelf. It's only changing the mix, and it's it's changing the way that traditional, discovery and buying works. So the key to driving growth in this next era is to ensure that you're winning more at bats and that you are relentlessly optimizing your product content and your strategies across every one of these shelves to drive discoverability and and purchasing. And this new agentic shelf is dramatically changing the way that products are are discovered. Right? AI shopping has already become mainstream. 60% of consumers report already using AI to perform shopping research. We've seen an explosion in AI referred traffic, almost 400% year over year increase in AI referred traffic to, to retail channels. And it's generating higher quality traffic as well. AI traffic sources, or AI referred, traffic converts 42% higher than non AI sources. So when you're able to get someone to visit your website or one of your PDPs from an AI channel, it dramatically increases the chance of that person actually converting. So it's really, really important that you are found, via AI channels to drive, continued conversion and growth, across any one of these shelves. So how do you do that? One of the ways to do this is, what's called answer engine optimization or AEO. This is the practice of optimizing your content to be understood, chosen, and recommended in AI powered search engines. And this is the new battlefield for shelf space. Right? You might hear this be called different terms. There's a GEO, generative engine optimization or LLO... LMO, AIO, ACO. I think there's some micro differences to all of them, but for all intents and purposes, these are all part of the same thing. And the common term that we're hearing is is AEO. But you wanna make sure that your products and your content are the answer that the machine that the machine trusts. So when consumers go and ask really detailed, nuanced questions about, like, what product's good for my use case or, you know, do certain products meet certain types of, you know, criteria I'm looking for that the machine is able to find your products to understand if they fit that answer and then ultimately recommend them to that buyer. So when you think about this new reality, you now have three distinct audiences that you need to create and optimize for, and you've gotta kinda do it simultaneously for all three. You still have the consumer. Right? Shoppers are still human, and you need to make sure that your product content is, high quality. It's compelling. It it resonates with with your buyers. Right? So the consumer, you still have to keep them in mind. You also have the algorithm. Right? There's still a ton of search that happens through traditional search channels. So you wanna make sure that your product content, has the right keywords in it, that it's meeting SEO best practices, and that it's being found through traditional search channels. But then you also have the agents. Right? This is that that new channel. And you have to make sure that your content is structured and, you know, deep enough and context rich enough to help these agents answer questions in dynamic conversation. So it requires a new depth and breadth of the type of product content that you have to store. And, again, you've gotta create content across all three of these audiences and really work to optimize it, to make sure your, products are found and discovered across each one of these shells. And and the the time to act on this is now. We saw a similar trend when, you know, during the rise of the digital shelf and SEO best practices, the the brands that were able to move first and figure out the playbook early on gained an early advantage and were able to build and steal market share. And it's the same dynamic happening now. It's, you know, there's similarities to the playbook, but there are a lot of differences and these things are moving very, very quickly. But the brands who are able to figure out that playbook and begin to work and optimize their content for the agents as well are gonna gain that first mover advantage. So when we take a step back and think about what this means for product experience management, you know, and and, you know, what's required for this new era, there's a few different things that we think about. First, your PXM platform has to be able to store and organize every bit of data that's required to go to market. This includes the product specs, the logistics data, the marketing content, but also this new layer of of context and and, you know, FAQ content that's really important for answer engines to ultimately, understand and recommend your product. So your system has to be able to manage all of this together in unison. It has to enable you to work at machine speed. Right? The stuff is moving so, so fast, and the faster you're able to respond and iterate and, you know, be agile on the market, the better positions you're gonna be. And, really, the only way that we're seeing people be successful doing this is by leveraging AI to your advantage. So ensuring that you can use AI in your day to day to speed up your work, to move faster, and ultimately respond to changes more effectively. And finally, your PXM platform has to be able to power this product content out to any different shelf and any different channel to ultimately influence those buyers across all three of these shelves, the physical, digital, and agentic. And this ultimately ensures your products are found and recommended, and and ultimately purchased to drive growth. And this is exactly what Salsify does and is built for. We are a unified PXM platform that's purpose built around these three key pillars. First, we serve as the system of record. We allow you to create that central source of truth for all of the product content that you have as well as that additional context that's required to to fuel these agentic experiences. We serve as a system of action that allows you to orchestrate and accelerate work across both humans and nonhumans, those AI powered agents as well within your system. And finally, we serve as the system of delivery that enables you to reliably deliver this product content to any any different destination across the physical, digital, and agentic shells. And when you bring all these pieces together into a unified platform, this creates that intelligent engine that allows you to work at machine speed and drive growth. And with Salsify, you can create that trusted source of truth for high quality AI ready data. We've got intelligent automation built into the core, and it allows you to leverage agents and AI powered workflows to dramatically increase your team's capacity. And finally, it doesn't just stop at data management. Right? Our platform is connected to the commerce ecosystem. We have a number of connections and retailer partnerships that allow you to publish your content out, federated across all three of these different shelves to drive discoverability and purchasing. And in the first two parts of this webinar, we focused on the system of record in part one, and we focused on the system of action in part two. And for this webinar, we really wanna focus on the system of delivery. Right? And talk about how Salsify enables you to publish content out to any different destination and particularly how we help you influence this agentic shelf and AI agents to better understand and recommend your products in conversational commerce. So we're gonna dive into that in a little bit, but I wanted to pause and do a quick poll. I wanted to get a sense of where the audience is at right now. So if you open up the the poll and the polls tab, I think Ray is firing that up. We wanted to know, what channels are you currently using to influence AI shopping agents? And I'm not just talking about the general channels that you publish to, but which channels do you have? Maybe call it a dedicated strategy or dedicated content that's being delivered out to market, particularly built to influence AI shopping agents. So we've got a number of different channels listed there. Please select any that apply. I'm really interested to see, what you all are using currently. So I see votes coming in. I'm gonna give it, like, maybe ten, fifteen more seconds to let you all get your answers in. Oh, there we go. Alright. So I see brand.com d two c websites. That's a primary one. Direct, to retailer channels and portals. Those seem to be seem to be the top two. Digital catalogs, definitely getting some votes as well. Not a lot for AI channels, discussion forums, not a lot for PR and media. Looks like most of you are doing something, and it's mostly spread across, call it brand.com and, and retailer channels. So cool. Thank you for taking part in that. Super interested to see, and we'll we'll see how the votes kinda tally up as you continue to answer. So, anyway, just to get into AEO, and starting to put together a playbook for how you really drive, answer engine optimization, I think it boils down... Just to oversimplify things, it boils down to three main buckets. Right? Number one, you have to create the content that these answer engines are looking for. Right? It has to be structured in a different way. It has to be purposefully built for these answer answer engines. You've gotta create that foundation of content that these agents need. Second, you've gotta go and federate this content out to any different destination. And you think about an LLM, like, they're not just using one source. Like, they are indexing across dozens and, you know, hundreds, even thousands of different sources, to figure out and understand different things about the web and, obviously, about about your products. So it's not enough to just simply publish to one channel. You really need to figure out how do we get this data across every different channel, make sure it's consistent, make sure it's accurate, and make sure it's, you know, discoverable by these AI agents to help them sort of index and understand your products better. And finally, you have to go and measure and optimize. Right? Like, this stuff, again, is moving very, very quickly. You've gotta be able to iterate. You've gotta understand the impact that you're driving and obviously make changes and and refine your strategy... Your strategies moving forward. So let's dive into each one of these at a at a high level. So the first step, creating. Right? Like, as we mentioned, AI engines need a lot more context about your products to inform the answers and ultimately recommend your products. And traditional ecommerce copy isn't sufficient. Right? It sort of treats products as a set of features and lists, and there's bunch of attributes, size, color, materials, weights. That's all important. Like, LLMs use that to to verify data and ensure accuracy, but you have to go deeper. Right? A, LLMs and AI agents, they don't think in lists. They really think in scenarios. Right? You think about the types of questions that are getting asked to them. They really need to think in scenarios. So think about, like, you know, if you sell backpacks. Right? Like, is this backpack good for wet weather hiking? Like, those are the types of questions, and those are the types of answers it has to be able to validate and reply. So it's not just looking for the term waterproof, let's say, but perhaps a phrase that says, yes. This backpack is great for wet weather hiking. So that's the kind of content that we have to start to create. So Salsify introduced two new accelerators to our intelligence suite. What these do, they purposefully build this content and structure it in a way that's built for LLMs. So you can start to build this foundation of content. It looks at your existing data stored in Salsify, will auto generate things either as use case content or, structured FAQ and q and a pairs that can then be stored as their own attributes within Salsify and ultimately published out or syndicated out to different channels. So, again, this gives you that... It it helps you start to build that foundation of agent ready, agent specific content that can be used to inform agents, give them the context they need, and help influence them to understand and recommend your products. So next, you've gotta go and publish this content out to the different destinations. Make sure that bots can ultimately find it. So there's a few different ways that we typically recommend that you do this. The first one would be, digital catalog. Salsify has a digital catalog tool. Earlier this year, we introduced, an ability to basically make that catalog optimized for agentic discovery and agentic search. And what that does is when that feature is toggled on, it creates a set of metadata and markdown files that help agents better understand how to index and read basically the the content in that catalog so they can interpret it better. So it's basically building sort of the infrastructure that allows an agent to more easily crawl and understand the the content in a catalog. And catalogs are a great tool for this. Right? It's a great way to curate out content for any different audience... In any different audience. That could be an internal team of salespeople or customer support reps. That could be a retailer or distributor that just needs access to some simple product data, or you could create a catalog that you specifically curate content and deliver content for AI agents. And, again, with this new option, it allows you to, increase the ability for that agent to basically understand and interpret that information. So catalogs are a great way to do this. Ultimately, you wanna make sure people and agents find these catalogs, so linking them on your brand.com site is a great way to help kinda federate that data further and make sure people can find it. Now on the lines of brand.com and d two c storefronts, this is another great way to get that content out. As you create this FAQ content, the use case content, making sure that appears on your brand.com and on your d two c storefronts is another good way to federate that content out and help agents more easily access it. One, quick call out here. Sometimes web teams and, you know, kind of DevOps or security teams will block crawlers by default on your website. So a good thing to do is check with, you know, potentially your web team or whoever owns your website or perhaps security protocols at your company. You know, are you blocking, these agents by default? Are you blocking crawlers by default? So if OpenAI does come to your website, you wanna make sure they're actually allowed to index that that content, and you can whitelist certain crawlers, to access your site. So one sort of, minor tactical call out there, but we've seen brands, you know, introduce this content and then ultimately discover that bots are being blocked from from crawling their website. So definitely something to tie off and make sure, as you're as you're going through and trying to federate this data out. Now the last channel I just wanted to mention are traditional AI and retail channels. This is another great way to publish your data out. There's a lot of infrastructure that our our customers already have in place to do this, and it's another good way to start to influence agents and federate your your AI content out more effectively. Number one, OpenAI, they have a a merchant program. I believe it's closed and they're currently accepting registrations. We have not seen them really open this up broadly, but we do encourage you to register as an OpenAI merchant. This allows you and will allow you to eventually feed data to them directly, when they open that stuff up. So that's one thing you can do immediately. Just make sure you are registered as an OpenAI merchant. And then there's a number number of other channels that you can use to publish to. The Google Merchant Center, that's a direct connection within Salsify. That's directly used to influence, Gemini and the Gemini model. You can publish through the Amazon selling partners API. Again, that that that, channel and that API directly influences Alexa, their AI assistant, and the same thing for Walmart with SupplierOne and Sparky. Publishing through the, SupplierOne portal or through those APIs allows you to directly influence Sparky. Some of these channels... I I know Google Merchant Center and maybe some of the others have opened up new properties and attributes that are specifically for q and a content. So they're starting to accept that sort of structured, you know, purpose built content for agents through their feeds. So these are good ways, again, to leverage the existing infrastructure, leverage some of the connections you already have to, publish and federate this data out and start to influence, different sorts of agents and AI shopping channels. So those are some tactics around publishing. I also mentioned, measuring and optimizing, and I am super excited to have Otterly dot ai on the call here today. They're one of our partners. They really specialize in measuring, agentic channels and, helping inform you with best practices and, you know, to optimize your content for AI discovery. So I'm gonna go ahead and turn it over to Rick here. Rick, I'll invite you on stage, and I'll let you take it from here. Rick, if, can you come on stage? Sorry. There we go. we go. Hi, guys. Yeah. Hey, everyone. I'm, Rick. I'm connecting here from Europe, Belgium. I'm the Geo SEO content strategist at Oderly dot ai, and, we're gonna be covering, the measurement and optimization bits, of the process. Otterly dot ai is an AI search intelligence platform. So, we are trusted by over 40,000 marketeers, from around the world, agencies, brands, and we help them with their AI search visibility, measurements. You're gonna kinda see how the process works. So, everything is from a to z, from researching your prompts all the way to tracking, how your brand is evolving over time in terms of AI visibility, up to, the last layer of, getting actual recommendations on how to improve your AI visibility. A little bit more into detail of how it works. So, again, AI search prompts, we help you, find, what prompts are relevant for your products and business, help you monitor, over time. So this basically means, how visible are you exactly in AI search. And then, next to that, the next step is actually doing a content audit. Is your content AI friendly enough, to be cited by these AI search engines? And finally, the recommendations piece, that part, you get recommendations on how to improve your existing pages. AI search, Nathan mentioned that AI search has changed a lot, especially in terms of measurements. One key issue is that AI search is basically a black box. Right? So with existing tools, it is very hard to actually see what people are typing into, ChatGPT or Claude or, Perplexity. So it is pretty much a black box. When we look at our own situation, this is data actually from, our own brands, Ottery dot ai. When we look at data from our web analytics tools like, Google Analytics, it is basically telling us that, 0.1% of our conversions are happening from platforms like Cloud. The reality could not be, any different... Any more different. Right? So when we actually ask our clients, with contact forms, by doing self reported attribution, we see that the the gap is huge. Right? So here, on the right side, we can kinda see the the reality, When we asked, our customers, how did you actually hear about our brand? It it is not 0.1%, but it's over 10%. And we can see the same thing happening, on other AI search engines as well. So, another interesting fact about our brands, are you aware how much revenue is coming from these AI search platforms? Self reported attribution is is, of course, the first step. There's some data again from, our own brands. Around, 25% of our revenue comes from, Claude and ChatGPT alone. So, quite, eye opening, if if you are not, aware of this. The average user, Nathan also, mentioned it, spends a lot less time, in in making a decision. They are already coming to your website very much in... Informed, and they spend 25... 15% less, less time. So these are just examples of how, important AI search is. And most importantly, to remember is that AI search is primarily a brand channel. Right? People go there to ask for recommendations on, what are the best running shoes, for example, or in a certain situation for for the winter, what backpacks should I use for, extra long hikes, and so on. And next to this, also, next to the organic way, of appearing into into AI search, we are also seeing, these AI search engines changing. Right? So, the emergence of, chat GPT ads, now being recently introduced. And this is data actually from a recent research, that we published. This is happening in, in this research piece, yeah, in in more than 17 industries. So from finance to, governance, hospitality, and so on, ChatGPT Ads is, is definitely there. And this is also our view of how how to optimize exactly for, AI search. We call this the AEO or the GEO strategy house. You start, of course, by scoping your own situation. What is your current AI visibility right now? This is called the geo audit up to what prompts should I be tracking, correctly. Right? Here, I'm, definitely referring to, bottom funnel and middle funnel keywords, questions, that users are asking, for example, what are the best alternatives of a certain brand, or, what are, the the type of content pieces you should, create, in relation to problems. And then, of course, the final layer of the foundation is reporting. And part of that house, includes, on page optimizations, but also off page. We saw it here as in the vote. Right? How are you exactly optimizing? Is it by doing PR? Is it by optimizing your content? I'm sure, Marissa will, show a little bit, how that works. And let's continue to the demo. Marissa. Alrighty. Hello, everyone. Let me get my screen share pulled up here so that we're not just looking at my face really, really big. Right behind the scenes. I know you can hear me. I need that button. There she is. Perfect. Alright. So great to see you all this morning, or maybe afternoon depending on where you're calling in from. But I'm gonna take us through the Salsify platform, really building off of kind of what Nathan just opened with that overview that he gave and giving you guys some visuals as to what all that stuff means and how it actually happens within the Salsify platform. So I'm starting us here on the Salsify detail page. This probably looks familiar to a lot of you, and it probably looks a lot like your current data. This traditional ecommerce copy really thinks about products as that list of features. Your sizes, your dimensions, colors, weights, things like that. And while that is really important for human browsing, like Nathan said, agents don't really think in lists like this. They really need to hear those scenarios. They are trying to answer tricky human questions like, what is the best backpack for a three day weekend hiking trip in rainy weather? I'm hoping to walk the Camino De Santiago next year, so I can say that I have definitely made a search like that many, many times recently. So if your product data is not addressing those kinda niche situational variables, AI is gonna just move on. So what do we do about that? How does Salsify help to kind of mitigate that problem? The intelligence suite works to close this gap by leveraging your existing product data to automate that content creation, including things like question and answer content, as well as use case content. AI search engines really do prioritize this q and a format over almost all other content types because it directly matches how shoppers are querying them. So the q and a content generator, Salsify analyzes your catalog data alongside historical retailer q and a's to produce structured, optimized q and a pairs for each item. This gives the AI platforms precise answers delivered in their preferred format. And then the use case content generator is gonna give that really deep narrative depth that the models require to understand your product. So structuring things across the intended purpose, event context, and audience profiles. So like many of you have seen in our previous webinars throughout the series, your intelligence suite workflows come equipped with these structured, prebuilt prompts. These really enable your team to quickly generate that high quality AI optimized content while maintaining agility and consistency because you don't have to DIY this every time. Specifically, this AI text generation task is using Salsify IQ to power that q and a content generation. So by searching major retail channels, extracting those common questions, and cross referencing answers with your Salsify product attributes, the system is gonna deliver shopper friendly context that's anchored in your catalog data. Then what happens, is every item that moves through this workflow triggers a review task for your team. This keeps a human in the loop at every stage so that you have the flexibility to really refine and adjust content as needed. I can click in and make sure that all the proposed updates are what I want them to be before they're saved back as trusted product information. So in this example, I've really got that compilation of comprehensive questions and answers covering key details about materials, applications, the compartments, durability, storage capacity, all that stuff that I'm gonna wanna understand as a shopper. And then this makes sure that those AI systems have a really precise understanding of your catalog when they're answering those shopper questions. Next, your team saves the final version only after they've confirmed the accuracy of those entries. And then selecting this resolution is gonna create the link between that human review and then the governed approved product context. Now you can also use the use case content generator. This is gonna put together that narrative layer to gain more understanding of the product. So I've got my audience profile, like who might want it, the event context, when are they gonna use it, and intended purpose, how are they gonna use it. This is all great. But once we create this, how do we actually make sure that the bots see it? Right? I think that's probably the question that most of you, came to this for, to understand that. So achieving AI visibility comes down to maximizing your... And Nathan laid out three ways that we kinda think about that. And that's exactly what I'm gonna take you through, which actually really closely aligns with what you all said in the polls as well. So rather than guessing responses, the models have to crawl a lot of sources to retrieve verified data. How do we make sure that you have an input on that data and what they're crawling? A common mistake that brands are making is relying on a single channel for that distribution, but we know that there's really three main places to look for this. So starting with what lies in our direct control here. Falsify catalog sites. This is where we have that dedicated toggle to optimize your sites for agentic search. Activating this setting is what enables the agents to discover your catalog site and then leverage that content to address shopper queries. So we plug in your use case and your q and a content, and then we can connect this to really establish authority and making sure that those AI systems are recognizing a single unified source of truth across your platforms. What it looks like is kinda like this. We see all that use case data is laid out. The q and a's are listed here, and this is what the agent will be crawling to actually figure out when someone is asking a question, what to return back to them. Moving on to, like, your own digital storefront, so yourbrand.com. Because these are actually stored directly on the individual item record once the intelligence suite has created them, you can publish straight to your site. So you could use something like the q and a content, below the fold in enhanced content. You could use some of the use case data to help tailor your features and descriptions a bit more. Really, the opportunity is kinda endless because you can use this however you want to see it. But then finally, we've got our syndication channels, and this is where we can distribute these dedicated AEO focused attributes out to their intended destinations. So for example, the Google Merchant Center here, they've got this enable, feature you can enable for AI shopping, which is gonna make your items eligible for a Genetec checkout on Google. We toggle this to yes, and then people are actually able to complete their checkout directly in Gemini. So we're distributing your data across all three of these places, and that's gonna establish this kinda airtight way where AI agents can locate your brand across every listing channel, get it in front of the customer, and let them actually have that conversion. So we've created the content. We've published it. And now we have to actually optimize it and measure it. So I will turn it back over to Rick to show us how Otterly helps with that, and gets you the data that you need to kinda track your progress. Thanks, Marissa. Let me just quickly share my screen. There we go. This is the Altery AI platform. Right? So, right now, what I am showing here are our, industry reports. Right? So we have one about consumer goods, health care, hospitality, travel, crypto, and so on. And here in this, selection, I've selected consumer goods as an example. Let's say that we are, Walmart. We're tracking, various different brands in in these industry reports. We can also see, the movement over time, how this, brand compared to competitors is, being cited, by AI search, platforms. And more importantly, as also we can see, this evolution over time. Right? So are you... Is your brand getting, covered more or not? What is also super interesting is to, be able to see what domain sources, an AI, platform, actually sites. Right? So it it can be anything from news and media. It can be, branded content. So, if you wanna have a look at your competitors, this is possible. It could be, community and forums. Right? So, we see here also Reddit is is one of the top 10 sources in in this situation. We see YouTube is also being cited. So there are many ways, to be cited. And in this example, let's say that we are Walmart. How do we actually know if our content is is working? Right? So, if I click this, then you will see, we are, here on a page with all the Walmart, URLs. Right? And we can see here, the total amount of citation, citations that each page, receives. So, for example, we can see, what is working really well, what is not, what this AI like, what does it not like. Apparently, this page right here, received, in the last ninety days, almost 90, citations. It's one of the top cited sources right here. So we know that whatever is happening on this page, the AI, really loves it. Right? So, the way that the content is formatted, all the information is there. We can look at some other examples like this one, for example. Another, example of a landing page, that contains a lot of this, snackable content that is really attractive for, AI. And this, the... These sort of things, they influence your, AI visibility. Being, present on Reddit, for example, is also so off page and on page. This combination is really important. And here you can also filter, is your brand being mentioned on this, Reddit, Reddit post? We can see people are asking, certain, questions around home bars and what spirits they should be, should be... Which one should be recommended. Right? So we saw also here, a lot of these products are are or are not being mentioned, and, we can kind of track this within the threads. But also on your, own page content, we are also able to see, for example, what are... What is working with your competitors. Right? So having this ability, to, get these sort of recommendations of how to improve your content exactly, this becomes very vital and also how to increase your, AI visibility, being present on, a lot of different channels. This is how to improve your, AI, feasibility. Next to this, we we also have the ability to, track, bot traffic. Right? So this metric, basically focuses on brand coverage, but we can also see, for example, different ways of measuring your AI visibility. One of those ways to do that is also by looking at, agentic traffic. Right? So we are precisely able to see which pages are being crawled. Here we can see, this example, copybarroworld.com. We can see that ChatGPT, has visited certain pages over 1,800, times, Claude, 116 times. We can also see which agents exactly are crawling, your website as, not all agents are are there, by a user. There are, for example, ChatGPT has three different, types of bots. Right? And one of them, ChatGPT user agent, that one is the one that you want. That one is the on demand, fetcher. Basically, it means that, whatever... Whenever someone is typing something into ChatGPT, this is the bot that will go ahead and fetch, the... Your information. So, knowing these sort of things, allows you to, see what is working and what is not working in AI search. And, this, this is the the short demo of, alterdy.ai. And we'll work through, some of the questions here. Am I... Can you can you... Everyone hear me? Okay. Rick, Marissa, are are you guys on stage? Here we go. There's Marissa. There there we are. Okay. Awesome. So let's, let's move into q and a. We've got a bunch of questions here. And as a reminder, again, use the q and a section, in your panel to ask any additional questions, about anything you heard here. I do wanna start. We got a few different questions about catalogs and how they're used. Like, why would you create this content if consumers can already see it or, like, how does it actually get accessed by, by AI agents? I I think the overarching idea here is that, like, a lot of this is a numbers game, and you saw this in Rick's presentation as well. Like, the more that you can federate this data out to different channels, the better chance it's being indexed. The more repetition helps reinforce that to agents. So some of this is a numbers game. Catalogs are one way that you are able to, again, federate that content out and make it accessible. And with the new AI discoverability option, it just further helps those agents better understand that context. So it's a great way to expose that content. Cross linking that catalog onto your .com site is another way. Again, it gives, the crawlers another link to go and access and and, you know, additional, pages to to to find it on. So, again, it's a it's a way to expose that content and then, you know, figuring out ways to cross link it and and maximize its ability to be found helps improve agents' ability to, again, find that and, you know, index that content. So I just wanted to clarify that piece because there were a few questions about, like, where do catalogs fit in. Yes. You can have your catalogs. They're great for consumers in some situations. They're great for internal users. They're great for retailers. But, again, it's another tool in your tool belt, and you can either use those existing catalogs and optimize them for agentic search, or you could create a a new catalog that's specifically built for agents that you wanna cross link in specific places. So there's lots of different ways to approach this, but when we say catalogs are a channel to help help support this, that's how we're talking about it, and that's how we've seen our our customers typically use it. I wanted to ask, we've got some other questions about, kind of optimization and, you know, which q and a content, to to prioritize. Someone asked about only having five q and a, answers available, or how do you ensure your use case content is relevant to what users are searching. Rick, I might toss this one to you. Can you just tell us about how to... You know, it's great to create this content, but how do you optimize it? What types of insights can you pull back to either figure out which questions you should prioritize or how to optimize your use case content? Yeah. So we basically, do a sort of gap analysis. Right? So we have a a built in recommendations tool within the platform, and, the platform will basically tell you what it likes and what it doesn't like. It can be certain things like having comparison boxes, right, on your landing page. It can be having, I think you also mentioned that, maybe an FAQ. So these sort of differences, and and, unsurprisingly, I've been conducting a lot of these experiments in the background with the team, what is what is working, what is not working. Not surprisingly, the type of answers, the the format that you get a response from, ChatGPT or crawled is also the way you should optimize it. So, for example, if you are, asking, compare, my existing phone with the latest, iPhone, you know, what is the difference? You would you would get as an answer a sort of, comparison table. Right? So making it easier for, the AI search platform to to, to understand your content, that that is the first step. But, of course, there are, certain types of content that do give you a benefit. I think, Nathan, you you mentioned it also in your presentation. For example, alternatives, content, revolve... Revolving around solution problems, these sort of things. Perfect. Thanks, Rick. We've got a few other questions here on, citations. So there's one question. Can you know what volume of search for a brand or category goes through an LLM? You know, it's kind of like a black box now. And then I might bundle this one in as well. How does Otterly get all the usage and tracking data including citations and crawling? So maybe just talk about, some of the citations stuff that you guys you guys support. Yeah. So the... Basically, how it works is, we have, we are tracking seven different, AI search engines from ChatGPT, Cloud, Google AI overviews, Google AI mode, perplexity Copilot, and Gemini, and we, ask that prompt, for a specific location. So let's say that your user or you're in... You're interested in knowing how... What the situation is in The UK or somewhere, in Canada or The US. You can see... Simulate, exactly that question in that location, and that question is asked, over time. Right? The same question, the same prompt, and with a certain amount of prompts you can, get a sort of idea of how many times your brand is cited or mentioned in the prompts and this is how we kind of analyze the data. And next to that, we also have, the the crawler bot data. This is, a piece of, if you have, for example, CloudFlare, you can hook it up directly with Audley AI, and we we pull the data from CloudFlare, and we're able to see, the server logs. Right? So the server logs meaning, the the traffic coming from these, AI bots. So, using these, different methods, we are able to determine, whether or not your brand is, cited, more, than, any of the competitors that you list, within the platform. Perfect. Thanks, Rick. We have another question here about, q and a content. Do you need to be concerned about having the q and a content repeat, what is already on the page, like stuff that might be in paragraph format? Marissa, I might toss this to you. Do you have any guidance on, like, how to reuse content or what things should be maybe purposefully different from what you already have? Yeah. So there's a lot of customization that you can make within the workflow that creates this, but by default, it's gonna look to create new q and a pairs. So trying to really differentiate from your existing content so that it isn't duplicative. But you may find, for example, when you enable the web search feature, That's gonna let the workflow actually go out to your existing PDPs at some of your retailers and see what customers are asking in q and a on the site, maybe see what they're mentioning in reviews, like what common themes there may be. And if they are mentioning things that are in your content, you may see that there are some similarities there. And that just means that it's probably worth doubling down on because they're still kind of wondering about those. But regardless, you're gonna get that q and a format from that response. So it's always gonna meet that criteria. Awesome. Thank you. Rick, this one's probably for you. There's a question on agent to conversion. How does it link up? Does it require checkout through the brand.com site, or can you actually kind of account for that or route transactions through traditional retailers like Amazon or Walmart, to track checkout and and fulfillment? Any any guidance Yeah. My best recommendation... I know I know this is a sticking point there? even in the old, SEO days, is, is really doing self reported attribution. Within, the checkout, you can add a certain, how did you hear about us, section. Right? This is the most... What we've seen, in the data, this is the most reliable way of seeing that sort of, click through rates. Let's call it AI click through rates between, the total amount of citations that you get and actual users that are buying your, products. So, that is still my my best recommendations, so self reported attribution. Alright. Awesome. Another one for you, Rick. Does inventory availability affect agentic search? Like, if your product is out of of stock, do you know how that might impact its ability to be recommended by an agent? Yeah. So it depends, you mentioned it before, for example. Right? ChatGPT has this, merchant, program. Right? What, that enables you to do is to, directly link up ChatGPT with your API, and and then it would, directly know without having to send, indexing bots, whether or not your platform has an has an info... Inventory, has stocks for certain products, and so on. So in that case, yes, it is pretty important to, regularly update that as, we are also moving more and more in that direction where, AIs are going to be fed directly with APIs from, certain stores. Perfect. Thanks. Marissa, this one's for you. Can you just talk about the visibility on catalogs and, like, the optimized for agentic search, how you can track if the catalog is being crawled? And this person also asked, like, they might assume that the... It might be more likely to be crawled on a PDP versus the catalog. So maybe just talk about some of the analytics and visibility there. Yeah. So our catalog sites do have some pretty good built in analytics that you can look at. If you're not using the AI discoverability today, but you are using catalog sites, you've probably seen this with site views and things like exports and understanding which human users are looking at your catalogs. There's also a bot traffic indicator now as well. So you'll be able to see our bots accessing the data, and actually using that because, really, it's a great way to get that data in front of them without having to make huge tweaks to your website and other places. So you will be able to see that bot traffic within the, the engagement section. Okay. Great. And, Marissa, another one for you here. Around the use cases and q and a's, like, can they be fed on the back end to, LLMs, like on Amazon or, or others? I know we've talked a lot about, like, exposing this to different, like, web crawlers, but how do you kinda feed it directly to different channels? Yeah. Couple different ways. I mean, there is... There's, of course, the direct syndication. Like, publishing through the Amazon channel is gonna get the data to Alexa. Publishing through supplier one to Walmart is gonna get it to Sparky. But if you wanna get really specific with it, you can include things like the q and a data in an FAQ block in your enhanced content that you may send to those retailers. So actually surfacing it directly in that specific format, not only just for the sake of having it be crawlable by those agents, but also in case you do have, you know, human viewers who are wondering that same information. So definitely a lot of different ways that you can kind of repurpose and distribute this content. And then the point of that being is you've got kinda consistency across all those different areas, and it just makes it more and more trusted and verified data. Great. Thanks. Another one for you, Marissa. Do bots on catalog sites count towards monthly views? They do. Yes. You will see that in the full, usage total. Which I'd say is probably a good problem to have. Right? Like, you want you want these bots to be crawling crawling your, your catalog site. So, and we don't see... Like, there's not, like, thousands of crawls. Like, these bots aren't, like, constantly crawling. But if you get a bot to come through, you know, once or twice a month, like, that is a good thing that reinforces this stuff to these to these LLMs. Yes. So that is a a good problem to have. What else do we have here? There's a question, Marissa. This might be to you. Can we connect these AI workflows to, any insight area and Salsify to measure results? I'm not quite sure what we mean by insight area, but maybe just speak to how some of these workflows can be used. Yeah. I mean, there's, gosh, there's so much opportunity with them to create all kinds of content. So when I'm thinking insight areas, I'm thinking maybe, like, the product detail page, insight reports, or something like that. So if you're seeing feedback that, you know, your titles aren't enough characters, your descriptions aren't enough words, you could definitely use the intelligence suite to really, really quickly action that and then create new content that does meet those requirements. If you're thinking more measurability after the fact, say, post publish understanding impact on checkouts and things like that, that would probably be more Rick's side of things, with Otterly, I think. Cool. Thanks, Marissa. Rick, I know we're getting to the end of our time here. One more question for you, Rick. Is there a way to determine why someone is ranking higher or lower in the GEO set versus competitors? And then, you know, how do they take the right corrective actions? Yep. So it's, it really depends on the situation. The correct approach is doing a gap analysis. So, for example, seeing what, content, is being cited by your competitor that that you are not, being cited for. And you can quickly do a sort of, within the recommendations tool, it automates that process. Right? So the manual process would, of course, be, taking your competitor's page that gets a lot of citations, comparing that with the content on your page, and seeing, what the gaps are exactly. Right? So this is, this is the process, and, again, these sort of recommendations, get automated with... Within the tool itself. Awesome. Thank you, Rick. Well, that brings us right at time basically to the end of our webinar. So thank you, Marissa and Rick, for for joining us. Thank you to the audience for all the wonderful questions. Hopefully, you, learned some good tips and tactics to take back to your organizations. And, again, thank you both for participating. And, to all the audience out there, we hope to see you on an upcoming webinar. Thanks for spending some time with us here today. Thanks, guys.