Video: How AI Sees Your Brand | Duration: 1876s | Summary: How AI Sees Your Brand

Transcript for "How AI Sees Your Brand": Hello, everyone, and welcome. Welcome, welcome to this webinar. We will be discussing how AI sees your brand, and this is, one part in a three part webinar series around AI discoverability. So I'm one of your hosts. My name is Jodi. I am the CMO of WordPress VIP, and I'm joined today by my friend and partner and the CEO of 97th Floor, Paxton Gray. So just a little bit of housekeeping before we get started. As you might expect, we are recording. This is a live webinar, but we are recording, and you will get the replay afterwards. So feel free to to look out for that and share it with a friend or colleague. Go ahead and drop any questions you have in the chat on the right hand of your screen. We will do our best to take those questions at the end of our session today. And you'll also, see a poll, that my colleague Mike just flashed on the screen. That'll give you the ability to weigh in to this critical question. And then we will be sharing the results of the poll later in this session. And then finally, as I mentioned, this is session one of a three part series. This one is really about, diagnosis and what AI is saying about your brand right now, why it says that, and how to measure the distance between what you really wanted to to understand about your company and your offering and what it actually does understand and reflect back to your audience. So our agenda today, we've got about thirty minutes together, so we'll be packing a lot in. We'll start by reviewing the AI reputation gap, then we'll transition into talking about the mechanics of how AI understands the truth or facts about your brand, and how to become authoritative in the AI era. And then we'll wrap by summarizing five concrete asks that you should take back to your team to start to really understand, how you're doing with AI discoverability and start to move the needle. And then we'll take some q and a, as I mentioned. And with that, I'll hand it over to Paxton, who's gonna get us into some of the meat of the presentation. Great. I'm excited to dive in here. So, you know, AI has changed the game, especially as it relates to search. We've been doing SEO for a little over twenty years now, and I have to say there's not been a bigger update than what's happened with AI search. We went from 10 blue links and trying to get shown there, and now what we need to do is get recommended. And so how we go about that changes, but we need to start with setting some baselines. The reason this is so important is because there's often a gap between AI and what it says about your brand and who you are. And the reason that matters so much is because this is some research from g two, but half of all b two b buyers start their research in AI chat. Two thirds say that AI actually changed the outcome of their buying decision, and 80% say that AI made their decision happen faster. So it's extremely important that you understand what your AI search strategy is and that you have that strategy dialed, because of the big impact it has on buyers. So starting with this reputation gap, it's really hard to know what to do unless you know where you are, so you have to understand where you are first. A great example of this is a, a client that we worked with, and they are in the venue management software space. They're one of the big players there, and, being such a larger company, one of their growth strategies was to actually go down market and to work with these new upcoming smaller venue places. They did some preliminary research and saw that AI did not think about them in that way. In fact, what AI said is that their business is like the IBM of venue management software, which is not what they wanna be. He said they're very well established, but they're clunky. They're hard to use, and they're not for small venues. So clearly, there was a gap between who they thought they were and who they wanted to be, which is hard to see at first, but it prevents, presents such a great baseline and a place to work from. So that that's so getting that baseline, understanding what that gap is is crucial before you even start building out what your AI strategy is going to be. And I'll pass it back over. to Jodi to talk about some of that build out. I love that example. I thought that was a real shock for them. And I think it's it actually it segues nicely into into this section because what surprises a lot of marketing leaders that I talk to is that AI models aren't just reading your most recent messaging or what is currently live on your website. It's reading what we call the total information environment. So definitely your website and what's live today, of course, but also, you know, that microsite that your your business, you know, launched in 2019 and is still up or a legacy subdomain that nobody sunset or an old product PDF, that's that's no longer up to date, press coverage, Reddit threads, etcetera. So there's this whole world that, AI models consume to try to understand your brand. And the fact is we see this when we look at when, we we have cross compared different sites for a same customer, and tried to understand what AI saw from that customer. So we evaluated a single company, six sites, the same brand, six sites, and the and the AI comprehension varied dramatically from site to site, on the high end, 88% down to 59%. And the point here is is when AI is consuming all this information, it's normalizing or averaging what it sees. And an an averaged brand is a diluted brand, so something that you really need to keep in in mind, which leads us to ask number one or recommendation number one that we have for you and your teams to have your team actually inventory all your properties, every domain, subdomain, microsite, PDF, and ensure that all those assets are telling a consistent narrative and telling the narrative that's most up to date and reflective of your current positioning, messaging, and offer. Alright. So next, we'll get into the mechanics of how AI interprets your brand and why it does that. And as Paxton Gray said, there's often a large disconnect between what you, the marketer or digital leader or executive, want AI to understand about what you do and why you're special and what it actually does and what it reflects back. And why is that? Well, someone once told me that an AI crawler is a brilliantly fast reader, but it's incredibly literal, which I think is a is a good mental model for how to think about your content and what it sees and cares about and what it does not. It does not care about your beautiful design. It reads structure. So it it reads headings that say what what pages are about, semantic HTML, that marks, you know, what type of content this is this is. Is it a product page? Is it an article? Is it your nav, etcetera, your schema? And we've seen at at WordPress VIP when we've worked with different clients that enterprises fail this test pretty consistently when it comes to creating the right structure to be understood. In one of our assessments, an enterprise brand that we are working with, had 16 different product pages with no markup at all. So as to a human, it looked fine, beautiful even each page. But to a model, it was, very difficult to decipher. So in other words, as many of us know, your website now has two audiences, and you need to make sure that you've got the structure that that LLMs can really understand. So that leads us to recommendation number two, and that is get a structural audit of your website. Make sure that you've got one h one per page. You've got a clean hierarchy. You've got schema on every page so that you're speaking directly to how, these crawlers read and under and understand and decipher. So this is template level work. It's a little bit boring back end work, I would say, to the to the average marketer, but it's super important, if not necessary. And it also doubles as an access an accessibility win, for your website as well. So now the the layer below the structure. The question is, can AI even find the content at all? So in one of our our our research studies, we analyzed 28,000,000 requests of real server traffic for one of our customers who's in the pharmaceutical space. And we saw that there was a really big disconnect between what humans could see and what AI crawlers could see. So AI crawlers saw just 2% of what humans saw, in terms of a con of content. And when we zeroed in on product pages, which are arguably the most important pages because that's where all of your buyers are getting information about your offerings, it was even less, point 7%. Remarkably, one set of product pages saw, 233 site visits from human beings, but just six from AI crawlers. That's not 6%. That's six crawlers, so basically invisible to to AI crawlers. And why did this happen? The good news was it's very simple. It's very fixable and certainly something that you can be proactive about so you don't end up in the same boat. And that was that these product pages weren't on the site map at all, and they had content that was locked inside JavaScript, which really renders empty boxes to most crawlers. So it's a long way of saying just because something exists and a human being can see it doesn't mean it's discoverable by AI. Your content can be live. It can be on brand. It can be perfectly written. It can be beautiful, but it can be functionally invisible to the systems that your buyers are now engaging with first. So our third recommendation is to audit your sitemap against what you are promoting and what you want to be reflected in in in in content and make sure that you have text fallback behind every JavaScript component on your website. And with that, I'll hand it back to Paxton who's going to talk a little bit more about content strategy. Great. So you've got a good technical base. AI can actually understand and read your content. So now you need to think about what does the content say. And AI has unlocked for a lot of brands the ability to produce just, like, tons and tons of content, and we've seen a lot of blowback from that. However, AI search does represent a huge potential for a lot of the content that you can create for these, for your customers. And the main reason is because a big difference between traditional SEO and AI search, goes into the prompt. So in traditional SEO, you look at keywords, you kind of develop your strategy based off of those keywords, and you build out content. Prompts changes the game because you have now have an extremely long tail of very specific prompts, and the way that AI search uses those prompts, sometimes when they don't know the answer, they will then go and ask and query search for multiple queries based off of one initial prompt. And so what this means from a content perspective is the more you can answer all of those secondary query fan outs, the more likely you are to show up in AI search. So the game becomes really understanding your audience and being very clear on who you're for, when is the right application for your solution, and then answering every single little question that may be asked by the audience. And so, customer understanding and research becomes extremely important to your content strategy. So it's this marriage of great, great content, understanding who you are and, who you're for, as well as a really great technical base. I mean, I'll I'll say this. Jodi might not, but AI was trained on WordPress VIP. It represents the vast majority of the Internet, and that's what AI trained on. And so what better platform to provide a great technical base? And then once that technical base is married with really great content and great message for the right audience, that's what AI success looks like. So, so now we need to we've got our our baseline. Now we need to establish authority. So how do we track progress to define authority, and track progress in AI search? This is another thing that is vastly different from traditional SEO to AI search is how do you measure impact and how do you measure movement. So I think there's a lot of different takes on this. We've consolidated down to really four different areas. One is referral and AI direct traffic. Two is brand mentions. How often is your brand mentioned as the solution to someone's problem? Three is citations. Is AI citing your content and drawing from you to answer people's questions? And then competitive comparison. How do you stack up against competitors in all these areas? It's not enough to just look at these metrics in the silo. You need to see how you compare to these brands these other brands. So this is what some of the what we would call leading indicators look like. So this is all dummy data. None of this is real. But if I am Jobvite, you know, I might look at to see how often is my brand mentioned by AI, in this case, within this query, 205 times. And this is all within the, tracking prompts around applicant tracking software, ATS. I'm also seeing across different platforms how often I mentioned, and then I'm looking also at how often I'm cited, and then all of that goes into kind of a general share of voice within the concept of applicant tracking systems. How much what percent of that attention do I command? So in this case, they've got a lot of room to grow. Workday is, you know, absolutely crushing. They're taking away a third of that impression share. So this is what I would focus on to establish, like, what we need to work on and whether or not we're making traction. Now the other side of that coin are the lagging indicators. As you move up the ranks and those mentions, citations, and share of voice, what you should start to see is more LLM referral traffic. You should start to see more branded search. As you show up more as mentioned as a solution to someone's problem, people will start to Google your brand more often. You also see increases in direct traffic. Often, people will have a conversation with AI, then they go straight to your website, and that will look in your analytics like direct traffic. And you should also see conversions happening faster. In the traditional SEO world, they're doing research. They're looking at your site. They're looking at others, and that's the research process. Now the research process is being moved into AI, and so they're coming ready to buy. And so as you succeed here, you should start to see those conversion rates increase and the time to conversion start to decrease. So these are just some of the main indicators that I would start measuring that are very different from the traditional SEO space, and and kind of require a little bit of a psychological shift when it comes to AI search strategy. Alright. We'll pass it back over to Jodi. Awesome. I love this I love this question of measurement because it's one of the, as Paxton said, one one of the most meaningful changes, from SEO to GEO. And, just to say it plainly, current measurement tools, mine included here at WordPress VIP, they really measure humans because they run on JavaScript in the browser. And AI crawlers don't execute JavaScript, which means this this big and fast growing reader of your website is their their engagement is really invisible to most of the dashboards that your team is using to report on on such behavior. And I definitely have seen that and hungry for, okay, great. So, you know, organic site traffic is going down. But how is that being counterbalanced with AI crawlers who are who are, sourcing me for different queries and answers? So when the the truth is in order to really understand what's happening from an AI crawler perspective, there's really one source of truth, and and and that's in your server logs. And if if you're a marketer on the call, you may not be looking at your server logs every day. But if you do, there's some pretty big discoverability gains that you can extract. We worked with a customer, for a global enterprise company, and they had a plug in installed for regulatory compliance, which is totally normal in the enterprise space. But unfortunately, it was accidentally rejecting about 15% of all ChatGPT requests. So in the time period that we measured, there were about 68,000 denied requests because that plugin was treating AI crawlers like human visitors from blocked countries. And then we actually did the same audit for ourselves at WordPress VIP. We rank relatively high in terms of technical AI readiness, but our server layer was rejecting one of Anthropic's crawlers. So, the question you need to be asking your team is show me AI crawler traffic in our server logs, and that may may or may not be something they're doing today. They may be really focused on some of the more traditional dashboards like like GA and Adobe and others. And if you do do this, we'll share a little bit later. We do have a complimentary assessment that we're offering, attendees for this webinar. And if you do discover that something like this is going on on the back end for you, don't worry. It's very common, and it's something that it's it's very difficult to to understand or glean without looking at the source of truth, and not the the traditional measurement solutions. This is just a quote from from our customer saying that that this audit was a was a game changer. They wouldn't have found it on their own using their normal measurement tools. The next point I wanna make on all this, that measurement and really just in general, is that, is that this is not a one and done project. It's a it's an ongoing owned cadence. As many of you may know, AI models have a very hard and, recency bias. So content goes stale really, really quickly. It's very impatient. I think Paxton actually said this in a different webinar. It's very impatient. And it has a huge recency bias. So you have to make sure that your content is is up to date because content that's stale will fall will fall out of answers pretty quickly. And additionally, if you're if you're measuring AI baseline and then fixing, retesting, reporting, that whole cycle, you need to make sure you're doing that on a regular clip. We do it about every 90 days here at WordPress VIP, but we certainly have customers that do it more frequently than that, just to make sure that they're always gonna have a up to date understanding of where they are relative to where they wanna be. And the second point here is just around ownership. This can be a tricky one for a lot of organizations. Parts of this work sit in different teams oftentimes. Schema can live with the developer. You know, crawler access can live with security. Sitemaps live with web ops, if you have a web ops team. And, of course, narrative lives with brand or the content team. So there's there's not necessarily a single obvious owner, but it's important to name one for that accountability so that someone can can take accountability for performance and bring the right folks in in the room to make sure they're doing the right things and optimizing regularly. So that's our our final recommendation is to name an owner and make sure that they're retesting and updating regularly. And with that, Paxton, would you like to wrap us up with some of the the summary of the recommendations? Yeah. I'd love to. Backing up, you know, what Jodi said, the impatience is something to contend with, but it's also a major opportunity, which means the faster you can implement these asks like, AI search moves so much faster than the traditional SEO strategies used to. So in summary, some of the big asks, I think, to walk away with today would be number one, inventory. Everything you have, make sure everything's consistent across the board and is on message with what you want to say and what you want AI to understand about your brand. Audit the structure of the site. Make sure you've got great schema. Make sure you apply the correct schema, fields. You wanna look at your site map. You wanna make sure that, you have a text fallback for when JavaScript won't work because the LMs do not play great with JavaScript. Show, AI crawlers. Number four, show AI crawlers in the server logs. Make sure you understand what they're doing, where they're getting stuck, any bottlenecks that you're causing, for these, AI crawlers. And then, really, like Jodi just said, name and owner. It's really important to make significant progress. You need some accountability, so having somebody own that and work with these multiple teams that impact AI search is gonna be really important. So, those are the five asks. Now it's time we wanna move on to looking at the poll, the results of the poll. So, some some interesting callouts here. Visibility and AI generated answers, that's a clear priority for all our people here today, which is great. That that's where we're starting today is the visibility and understanding the impact overall. The second, content optimization, strong second, and, brand representation and measurement, is is very close behind as number three. The lowest concern of people here today is ownership across teams. So that's great. Maybe that's because, everyone feels like they own it, which is would be a great problem to have. But, anyway, number one is is visibility in AI generated answers. So, great great thing to focus on. Excellent. Does that, sorry, just to just to discuss that for a second, does that track what you hear from your customers in terms of the top two? Absolutely. Yep. It's, it reminds me of you know, we like I said, we've been doing SEO for twenty years. In in the old days and for a long time, demand was kind of driven by, hey. The CEO is Google is our name and we don't come up, or the CEO Google's this word and we don't come up. We need to invest in this. We're seeing the exact same trend happen, but now it's the CEO is asking some things, in their chatbot, and we don't show up. And so, really, this is like a a brand new wave, some different tactics, some same tactics, but I'm seeing the cycle repeat again, and it is in you know, there's a lot of people rushing to, kind of grab this attention right now. Yeah. It's interesting that ownership shows up, at the end because it feels like if we're worried about some of these areas like visibility, and, you know, content optimization, Having a strong owner with rich expertise in this area obviously would would help, answer both of those questions. But having said that, this is probably just a stack rank and people are just calling out what they're what they're really focused on. And I would say number one is starting to become a board level conversation. Boards and. executive teams are certainly asking CMOs and heads of digital, CTOs sometimes about, you know, what are we doing? How are we showing up? Do we have control over this? Do we have expertise, etcetera? So I'm not surprised at all at number one because I think it's coming up at the very highest level and then, of course, trickling down. Yeah. I think the real tricky thing right now that our industry will get passed at some point, but it's one thing to see, are we showing up? And then it's another thing to see, is showing up having an impact on the bottom line? And that is a big gap that I don't think there's really great answers for right now. There's a lot of correlation and not causation. And so those brands that win are gonna be the ones who have the guts to kinda say, I can't see connect those dots completely, but I know that if I show up more, that will come back to grow our bottom line even if I can't connect those. dots. And so there's there's a lot of people, I think, waiting on the sidelines until they can connect those dots, but the ones that move first are gonna be the ones that win. I resonate with that so much because there was, like, this whole many day many decades where, you know, marketing really was about brand. It was about reputation. It was about, you know, you know, showing up and having people be aware of and preference for your brand, your solution. And then, you know, we enter the the age of kind of, Eloqua and Marketo and, you know, that that whole push around while, you know, marketing is part of the revenue team, and we can attribute revenue, you know, our investments to revenue and true ROI. And that lasted for a decade or so. And now it's almost like we've come full circle, and there's a return to, hey. Influencing and brand reputation matters a lot, but it matters differently, of course, because you're trying to influence machines and people at the same time. So I I think, we're in the early stages of being able to really measure the cause and effect. But until that happens, definitely encourage people to embrace the return to brand. And I say that as a former demand marketer myself, as the new reality and something super important. So I know we have more content to cover, but I just wanted to to discuss this the results a little bit. Yeah. No. I completely agree with that. Oh, this is me. I, meant. I. mentioned it before, but we we we are offering, an an assessment. If any of this work, particularly, I think, on the the more technical side, not all CMOs that I talk to are are very technical. They they obviously own the website in in many cases, but, they're eager to have kind of this this this rich technical expertise. We we do do this for customers, and we're offering a complimentary assessment from the WordPress VIP side and the 97th Floor side. So it'll be both, you know, some of the the the stuff that we talked about today in terms of access and structure and what AI actually sees in the server logs in the 97th Floor. We'll be evaluating more on the strategy side, where your mentions cited, how your how your stack ranking relative to your competitors, and then we'll give some personalized recommendations. So if that's of interest to you, it is complimentary. So feel free to to scan this QR code and request an assessment. And I believe Mike will will drop the link in in the chat as well. Okay. Cool. Well, we've got about three minutes for some some questions. And one of the one of the questions that came up is, we kinda touched on this a little bit, but how is AI discoverability different from traditional SEO? I've got a perspective on this, but, Paxton, I'd love to hear from you on this one because this is squarely in your in in your wheelhouse on the content side. Yeah. Yeah. Absolutely. So first off, there's a lot that's similar. It's not like a complete revamp. However, there's a lot that's very, very different. One of the things that, we talked about was measurement, so kind of building a new suite of KPIs to see and track how we're performing, that's different. Another thing is research. It used to be so easy to go in, do some keyword research, build strategies based off of that, and now the visibility into prompts is just not what it was before. And so the research we're finding is shifting from keyword research really to audience research and solving true audience problems, kind of what you were saying before. It's like a a return to real marketing principles is the way to succeed in AI search. So that's another. And then another that I would say is, SEO is often thought of as relegated to the site, which is true. AI search, your site matters a lot. You know, you will not win if if they can't understand your site, if you don't have great content on your site, but it's also much more comprehensive than just your site. It's how you show up everywhere across the web. That corroboration is really important for AI search. So those are the main three things. I always tell people a great analogy to traditional, SEO, the movement to AI search is is thinking it's more like a billboard. How would you measure the success of a billboard? You would see, are people searching my brand? Are people coming direct to my site? Do they convert better? That's how you would measure a billboard. That's kinda how you should measure and think about AI search too. I'd love to hear, really love, that. yeah, some of your thoughts on the differences. Yeah. I mean, there's a lot like you said, there's a lot that's core. We've got the like, on one side, content optimization and content influencing, and then the on the other side, we've got kind of the infrastructure and some of the more technical side. So in that, it's very in that way, it's very similar. And I feel like a lot of the things that we would recommend doing in order to to optimize for, showing up, you know, accurately and, and showing up in a favored way, are are very much the same. But there are definitely nuances, and no one model is the same. So I think that's one of the the bigger ones, and and things are changing. So I think thinking about, SEO and or GEO as kind of an evolution, an important evolution, it is different, but it's not like it's, you know, a wrench versus an apple. But it's certainly not apples to apples, but it's it's also not a wrench versus an apple. Some of the same principles, some of the same players are involved, but there are some very important differences that you need to be aware of. So I think that that, takes us to time. I know other people asked some questions, some good questions. I'm sorry we didn't get to them, but we certainly will, in our follow-up folks. Just a quick mention. Several of you, I think most of you, are actually, registered for our for our next, webinar. But if you haven't, we'd love to see you there. It's on September 22. And in the meantime, thank you so much for joining us for this thirty minutes. I hope you at least got a couple of takeaways, and we will see you hopefully in a few weeks. Thank you. so a great day.