WEBVTT 00:00:17.000 --> 00:00:32.000 This year interactive signal. We're really excited for this time to spend with you all. And real quick, a little background on why we're doing this. So we recognize there is so much content to consume in this space. 00:00:32.000 --> 00:00:49.000 It is really hard to keep up with. It's even harder to understand what is signal and what is noise. So what we wanted to do is kind of put together our thoughts. We wanted to share the content that was sparking the most interest, the most debate internally for us. 00:00:49.000 --> 00:00:59.000 And we really want to make sure that part of this is helping you as marketers understand, okay, now what do I do about this? What are the implications? 00:00:59.000 --> 00:01:18.000 So today, Will is going to kick us off with his POV every month in our all hands. Will does this for Sear as a whole. So I'm really excited to kind of share it with the broader community. Then I'm going to take over again. I'm going to go through a few new stories. And what I'm thinking about them, and what I think you all need to take away. 00:01:18.000 --> 00:01:35.000 And then we'll have time for questions at the end. So we're hoping this can be kind of a ask me anything segment where you can chat with me, chat with Will, get your questions out there. You should use the QA within Zoom for submitting questions. 00:01:35.000 --> 00:01:38.000 All right. So with that said, I'm going to hand it over to Will. 00:01:38.000 --> 00:02:02.000 Truck and roll so I can get them to, uh, the stuff you put together, Lisa, so let's rock and roll. So, um… I decided to look at… the themes that are coming across all of our sales calls, to understand, like, when we're having calls with prospective clients, what are we hearing, right? It's so funny how often we go to market without being like, well, wait, we have all these transcripts, like. 00:02:02.000 --> 00:02:15.000 What are the people buying the thing saying, right? So here's some of the quotes that I'm seeing, and I imagine that some of this is going to resonate with all of you, because it's what we're seeing the trends say across the people that are calling us. Next slide. 00:02:15.000 --> 00:02:28.000 are bored. You know, we're starting to hear the word bored, so our AI is, like, picking up on, like, whoa, this is being mentioned more. All of a sudden, people are saying their board is getting involved in this, which is interesting, because we're also seeing… next slide. 00:02:28.000 --> 00:02:41.000 And not only the board is there, and they're asking them to do more stuff. But then I don't have the metrics that the board wants me to have that they want on how is this going to make me money tomorrow? It's like… When am I gonna give you the money? 00:02:41.000 --> 00:02:52.000 Two, I'm putting tons of pressure on you, and three, my metrics are horrible, so I can't help you to easily compare it to every other thing that you've been investing in. Next slide. 00:02:52.000 --> 00:03:04.000 This one is like the banger, right? This is a quote from a prospect. And I think what people want to do now is, yes, they want leadership and yes, they want execution work. 00:03:04.000 --> 00:03:14.000 But they're like, execute what? Like, I need somebody to first banter with me. This stuff is changing so much. I actually want somebody who is a thought partner with me. 00:03:14.000 --> 00:03:24.000 And will tell me when I'm wrong and push back. Let's get in the mud, let's figure this out together, versus, hey, I've figured it all out. You go execute for me, so we're seeing this shift to. 00:03:24.000 --> 00:03:32.000 I don't have it figured out, which means I need a different type of person to help me try to solve this problem. Next slide. 00:03:32.000 --> 00:03:48.000 So, as always, I like to use my own site for all my experiments and tests and all my, like, thoughts, so when I think something's hot, I'm like, maybe we should do something differently. So if you were to look at our Geo page on generative engine optimization, you see these three questions we say we're getting? 00:03:48.000 --> 00:03:54.000 Do any of those sound like the questions that the… Prospective clients were asking? 00:03:54.000 --> 00:04:11.000 This is the new job to be done, my friends, right? Take and harvest the transcripts where your actual people who pay you are saying, this is what I'm looking for, this is what I want, this is what I'm struggling with, and next slide. We're going to start to turn that into our content, and this is what it looks like. 00:04:11.000 --> 00:04:19.000 Right? Because if you don't start to take the things about your business and put them out, AI will find things about your business from other people. 00:04:19.000 --> 00:04:34.000 So we're actively now launching pages on our site that are taking things that we're hearing from clients internally, and we're saying, let's get them on our website. Before, we just would shout out internally. Now we're trying to change it into things the LLMs can pick up on when they talk about our brand. 00:04:34.000 --> 00:04:39.000 Next slide. and my friends, I'm gonna send out the warning flares. 00:04:39.000 --> 00:04:44.000 Because your sea levels are seeing case studies like the one on the left. 00:04:44.000 --> 00:04:53.000 which this company says we tripled their AI brand visibility in one week. Who would not want to pick that freaking company? Right? I would. I'd be like, oh, a week? Let's go. 00:04:53.000 --> 00:05:06.000 Now, if you look at that same company and what they did, you can see that they're launching 10 or 20 or 30 posts all on one day that all follow the same template that all basically say the same thing. 00:05:06.000 --> 00:05:14.000 Next slide. And what that case… and it's interesting that one of the leading GEO providers in our industry. 00:05:14.000 --> 00:05:17.000 has this case study up today on their website? 00:05:17.000 --> 00:05:30.000 And what they have not shown you is how they got that visibility, and they have not shown you that the client has lost their visibility in Google search, which means Google's already found a way to penalize it in search. 00:05:30.000 --> 00:05:43.000 I can imagine that they're also saying we also want to get this out of AI. So then what do you have? Yes, you got your results quick, but now you've got a whole brand damage problem on the other side of not showing up potentially for your own brand or other things. 00:05:43.000 --> 00:06:01.000 It's a new world, and that's why we launched this series. We want to help people have… I want to show you real things. Alisa wants to show you real things that we're seeing, what's causing the most banter in our company of 180-ish people, and bring those to you all so we can all kind of learn together. Next slide. 00:06:01.000 --> 00:06:14.000 So most companies do not want to actively show people a slide that says they're the worst agency they've ever worked with. But you guys are used to sear right? Transparency, right? We got one bad review in 24 years, and this is what it is. 00:06:14.000 --> 00:06:33.000 And that was starting to show up on answers about our brand. Remember, AI wants to be balanced, so it will go deep looking for cons and it'll look at 100 pages, whereas a human back in the day, these like low quality sites, no one would visit, they never visited them, right? No human wants to visit 100 pages to try to make a decision about an agency. 00:06:33.000 --> 00:06:43.000 They're going to visit the top 10. Well, now in AI, they'll just keep going through 100 pages until they find something to balance things out about your brand. Next slide. 00:06:43.000 --> 00:06:47.000 And that is why we posted publicly what our employee retention was. 00:06:47.000 --> 00:07:01.000 And now all of a sudden, the answer changed. We now own the answers about our brand, and this is one of the areas that I'm so focused on. Everyone wants to be like, you should show up for geo agency. I'm like, no, I should make sure that when somebody knows my brand already, which I've spent 25 years trying to build. 00:07:01.000 --> 00:07:05.000 that the answers are congruent with who we actually are and what we believe. 00:07:05.000 --> 00:07:15.000 And I think, Elisa, that's my last slide. Go to the next one. Yes, this is my last slide. So yes, clients want to move from a proactive partner. 00:07:15.000 --> 00:07:19.000 Clients are smelling punch list execution a mile away. 00:07:19.000 --> 00:07:34.000 Be careful building plans for the next 3 to 6 months. You may have to start to get used to building plans for the next three to six weeks. Learn because things are changing and then pivot consistently. All right, with that said, Elise, I think the next one is yours, I think. 00:07:34.000 --> 00:07:35.000 There you go. All right. 00:07:35.000 --> 00:07:53.000 That's right. Awesome. Thank you, Will. So as a reminder, if Will sparked anything for you, if you've got a question, drop it in the Q&A. All right, I'm going to get into the next section of this, which is the the six stories that really got our gears turning last month. So we're going to talk about our updated click-through rate study. 00:07:53.000 --> 00:08:08.000 We're going to talk about 2 studies by Kevin Indig, great source of information. Love the stuff he publishes. One of them is the impact of brand recognition in search. Another is his study on ultimate guide. 00:08:08.000 --> 00:08:23.000 style content. We're also going to talk about Bing's long-awaited AI search data. They've finally given us this data. We're really excited about it. We're going to talk about conversational commerce, or the agentic web, and some interesting things Google is doing. 00:08:23.000 --> 00:08:35.000 And then I'm going to end things with Stanford's AI report, which I think is a really, really good source of information for all marketing leaders. 00:08:35.000 --> 00:08:51.000 Alright, so let's get into it. So this is our research. You might have seen this before. We've been doing this research for, I think, the past 2 years, and we have seen this long decline as a result of zero click marketing where click-through rate is fading and declining. 00:08:51.000 --> 00:09:06.000 And for a long time, this was happening because AI overviews are entering the search results, and people are getting the information they need right from the SERPs. We saw some promising data in the most recent update, though, after 18 months of decline. 00:09:06.000 --> 00:09:13.000 We're starting to see some regrowth, starting to see some stabilization. So what does that mean for you? 00:09:13.000 --> 00:09:30.000 We think you need to cautiously embrace your own new normal. We love providing this data, but this aggregated data is meant to be directional. It's meant to help you think your data is going to look different right? So we want you to embrace your own new normal. 00:09:30.000 --> 00:09:49.000 And we want to see, are you seeing the same stabilization in your own data? Even better, if you start to segment your own data and look at, okay, what is my my informational queries, my comparison queries? What is my click-through rate look like for this? We get really, really good data from Google. 00:09:49.000 --> 00:10:14.000 Let's leverage it. It is… it can be a pain, right, to log into search console. So… In terms of what you should do about it and the implications of this, my number one piece of advice is enable easy access to your own data. If you're a marketing leader and you're like, man, I would love to see this data, but I've gotta talk to my agency about it. I gotta, you know, ask my analyst to pull a report. 00:10:14.000 --> 00:10:32.000 Then, that mindset should evolve this year, and AI tools are helping us through MCPs and APIs are allowing us to query this data and have it more readily available at our fingertips. What this helps you if you're a marketing executive or marketing leader is with forecasting. 00:10:32.000 --> 00:10:50.000 Right? Because for all the investments you're making, we know you've got to defend those investments to your CFO or to your board. And if you have been using old or standard click-through rate data to populate these forecasts, then they're highly likely to be erroneous. 00:10:50.000 --> 00:10:58.000 So, implications on forecasting, even on resourcing and strategy planning here. 00:10:58.000 --> 00:11:11.000 Alright, the next big story we want to talk about is kind of two studies combined, one from Kevin Indig, one from this year team, and the headline here is that buyers choose familiar brands and search for brands by name. 00:11:11.000 --> 00:11:33.000 Alright, so let's get into the the so what of this brand recognition is a really critical search factor. This is what we are learning in the 2 studies that really piqued our interest this month. First, the one from our team at SEER, where we actually watched how people are prompting LLMs, and about half of professionals start. 00:11:33.000 --> 00:11:50.000 with a brand in mind in their prompt. And it's not that it's necessarily a branded query, right? They're not saying, hey, tell me about Sierra Interactive. They might be saying, hey, I'm looking for an agency like Sierra Interactive, and they're setting the tone, or they're setting the vibe based on a brand. 00:11:50.000 --> 00:12:08.000 they're familiar with. Another really interesting study we saw was by Kevin Indig and Citation Labs. They found that 88% of AI mode shortlists are accepted as is. So right now, AI mode is still it's not overly popularly consumed. 00:12:08.000 --> 00:12:24.000 Right? It's kind of a smaller feature on Google search. We expect this to change. But what Kevin Indig found is that if you do a prompt and you're given maybe four brands, as long as you recognize those brands, it's kind of taken as face value. It's like, okay. 00:12:24.000 --> 00:12:36.000 Because these are the four options that I should consider. That means it's very important to be in that initial consideration set, arguably more important that your brand is recognized. 00:12:36.000 --> 00:12:54.000 What this makes us think about, and I've been thinking about this a ton, I might have talking even to some of you about it, is how are we measuring this? Maybe you work at an organization where you're doing brand awareness marketing. You probably have some kind of function to measure your brand strength or that impact. 00:12:54.000 --> 00:13:06.000 of that work. Does your search marketing team have access to that data? And if you don't have any kind of tech in place, brand lift study or brand strength data. 00:13:06.000 --> 00:13:31.000 Maybe you should turn to your search data to populate this. So if you're running paid branded campaigns. Can you see what the what that your impressions look like on a month-to-month basis? Similarly, monthly search volume can be a really good proxy for this data. And don't just look at your brand term. Don't just track your interactive, for example, but see your interactive SEO. 00:13:31.000 --> 00:13:51.000 paid search, anything that is a priority product or service within your organization, get a sense of how that's growing or how that's falling. This, I think, is going to be really important for diagnostic work. So if you've got a board or your C-suite saying, hey, why aren't we visible for X, Y, and Z? 00:13:51.000 --> 00:14:08.000 This is going to be one of the reasons, or one of the ways you answer that. Also, as implications for measurement, success measurement, test measurement, and your tech stack in general, again, maybe you already have this data, but it's owned by a different team. Really good opportunity to break down those silos. 00:14:08.000 --> 00:14:30.000 All right. The next story, and another Kevin and Dig joint is the ultimate guide format is on its way out. So we all know the ultimate guide format and writing one piece of content where you're trying to capture as many different long tail searches as possible. And for a while this worked really well, and it was rewarded. 00:14:30.000 --> 00:14:48.000 by Google search. What Kevin has found is, at least with ChatGPT, density is not the move. It's not moving the needle. And we thought this was especially interesting because Danny Sullivan also recently made some waves with a presentation he gave where he outlined. 00:14:48.000 --> 00:15:04.000 commodity content versus non-commodity content. So this is a big shift, and especially if you've got the same strategy in place for the last 2 or 3 years, this type of content is probably part of your content strategy, and I think you now need to rethink. 00:15:04.000 --> 00:15:14.000 Okay, what are we doing? What does our audience really need? And how are they looking to consume it? And how are these models rewarding the different consumption methods? 00:15:14.000 --> 00:15:33.000 So what are the implications on this? Certainly on your content strategy, also on your resourcing. Who's creating your content? Who's restructuring your content? And my advice here, what I think you should do, audit any of your content that is over 2000 words and has over 20 subheadings. 00:15:33.000 --> 00:15:57.000 And I don't want you to rewrite any of that content. I want you to take a sample of it. Let's say 25% of these pages and restructure them as fresh pages. See if a different page structure, a different content structure allows you to be more visible in AI search or traditional search. I think this is the best way to test in this field. 00:15:57.000 --> 00:16:11.000 Studies from Growth Memo and Kevin Indig are great, but you always want to see how this kind of work changes within your landscape, with your competition, with your audience. 00:16:11.000 --> 00:16:31.000 All right, the number four is story. We're really excited about that. Bing is now offering AI search reporting updates right in webmaster tools. This is something Bing actually promised, I want to say in 2023, and we were really, really excited about it. And then, after a little while they're like, actually, I don't think we're we're going to release it. 00:16:31.000 --> 00:16:37.000 So we were surprised and really excited to see the announcement in the last month. 00:16:37.000 --> 00:16:56.000 So what does this mean for you? I think at a high level, AI search has far more variables than traditional search, which makes it complex, which makes it complicated, makes it really difficult to diagnose what's happening, makes it really difficult to report on what's working and what's not working. 00:16:56.000 --> 00:17:15.000 And a big part of the reason, a big part of the issue with all of these variables is we don't have a lot of quote unquote official information, right? We're using third party AI rank trackers. We're trying to build solutions to mine some of these grounding queries. 00:17:15.000 --> 00:17:30.000 So when you get a set of data straight from an official source, it's really exciting. And if you are a B2B marketer, you want to pay extra attention here because this data is coming from Copilot. 00:17:30.000 --> 00:17:40.000 So if you've got an audience concentration that's using the Microsoft Suite that you know is using Copilot, this is going to be a treasure trove of information. 00:17:40.000 --> 00:18:02.000 What is most exciting for us as well is thinking about how many of your grounding queries are branded. So as a reminder, what is a grounding query? You know, Llms are going to function kind of with 3 main steps. They're going to look first at their pre-training data. So what do they know about your brand, your products, your services, your your category. 00:18:02.000 --> 00:18:17.000 Then there's a layer of post-training where there's kind of invisible criteria that these models are looking for to evaluate what is the best response. And then all that information is often then validated with web searches or grounding queries to identify. 00:18:17.000 --> 00:18:33.000 Okay, what, what's the freshest, most up-to-date information on this topic? And so if you see branded grounding queries, what that suggests to us is that your brand is part of that consideration set. It's part of kind of that swirl of generating. 00:18:33.000 --> 00:18:55.000 The response. So our belief, our hypothesis is that if we were to track the percentage of brand branded queries, branded grounding queries over non-branded queries, if you see an increase in that, you should see an increase in your overall AI search performance, more traffic, more visibility. 00:18:55.000 --> 00:19:15.000 And of course, more conversions. Alright. So what are the implications of this again? Performance measurement. Nobody gets paid from earning grounded queries right? But it can be part of how we identify what's working and what's not. So a big part of the the value here is in the diagnostics. 00:19:15.000 --> 00:19:32.000 Also in your overall content strategy, there is gold in here. And remember, Llms do not search the way that humans search. Very rarely do you see a lot of overlap between the grounding queries or the query fan out selection and your tracked keywords. 00:19:32.000 --> 00:19:47.000 So you can find really valuable information in this data set. And also that can impact your content strategy. So what do you do about it if you don't yet have this set up, make sure you've got Bing webmaster Tools access today. Make sure your domains are verified. 00:19:47.000 --> 00:19:53.000 And check out the new reports, check out the guidance that we're getting from Bing. 00:19:53.000 --> 00:20:10.000 All right. Fifth story I want to talk about is it's basically conversational commerce or agentic commerce. And for many of us as marketers, we were aware of what agents are right and the value of agents, and what we can do to. 00:20:10.000 --> 00:20:33.000 automate some of our workflows using agents. The general population is less aware of agentic workflows or what all of this means. And so I was really intrigued by the fact that Google released this new capability for Gemini. Let Gemini browse for you. That's going to allow Gemini to compare products across tabs. 00:20:33.000 --> 00:20:51.000 Build a plan and execute a plan. So what are the implications of this? Chrome basically just enabled conversational commerce for billions of users. We have to remember that is the value, that is the power of Google. They have immense exposure across the board, right, with the amount of. 00:20:51.000 --> 00:21:06.000 The, you know, the planet's population that is using their products. So when they launch a feature in Chrome, it has the potential to really skyrocket the awareness of a capability. 00:21:06.000 --> 00:21:31.000 So what does this mean for you? How do agents today crawl and access and engage with your websites? Many of us have no idea, right? And so I think the first thing that you should do is simply use Gemini's capabilities. Now, you probably have to do this on a personal through a, you know, a personal Google account. It's probably not activated. 00:21:31.000 --> 00:21:48.000 Even if you do have an enterprise Google Workspace account, but go through the flow. What are the actions that you expect your audience to take on your website? What's a really important user flow? And can you emulate that using these agentic capabilities? 00:21:48.000 --> 00:22:03.000 There's a few others that I like. You can use perplexity for this as well. Perplexity was one of the first to really try to push this idea of agentic browsing. Aura.run, O-R-A.run. That's another good tool. 00:22:03.000 --> 00:22:10.000 to use to identify how agents can access your site, where they're getting stuck, where they're getting lost. 00:22:10.000 --> 00:22:25.000 This spans beyond just your marketing team. It's got implications that really start with your CTO or your CIO, because all of a sudden you got a very big new audience where you're not just building this website for humans. 00:22:25.000 --> 00:22:45.000 You now have to make sure that agents can access it in a really effective way. This has implications for your website architecture. It's got implications for the UX of your web experience. And again, it's got implications for resourcing. Who owns this work? Who's going to lead this work? 00:22:45.000 --> 00:23:08.000 These are open questions in many organizations. Alright, my last story that I want to touch on before we move to Q&A is this Stanford AI Index report. And I believe this is the 9th report that they've done. So our headline here is there's a sharp split in AI sentiment, depending on whether you're using AI at work. 00:23:08.000 --> 00:23:29.000 Or, you're encountering it as a consumer. So this is a really fascinating topic to dig into. So on one hand, you've got your AI insiders, and this is largely a B2B audience. This is probably an audience who's been using a tool like ChatGPT maybe since 2023 or 2024. 00:23:29.000 --> 00:23:38.000 You're exposed to it, you get it, you maybe are aware of kind of the jagged edge of AI and what works really effectively and what you have to be careful for. 00:23:38.000 --> 00:23:53.000 So buyers kind of get this. B2b buyers trust the capability messaging. You don't necessarily need to sell most B2B buyers on the value of AI. Very aware of it, right? The public, on the other hand, doesn't really know. 00:23:53.000 --> 00:24:09.000 what to think about AI, but sentiment is pretty, pretty bad, right? And a lot of that is in the overall messaging. People think of AI, they think of the environmental impact, they think of the ethical impact. They think of things like data centers. 00:24:09.000 --> 00:24:26.000 None of that is wrong, by the way, but it is really clouding their vision, or I should say, changing their perspective, especially relative to a B2B audience. So if you're going after a B2C audience, you have to be careful about this. And if you're using AI to create. 00:24:26.000 --> 00:24:47.000 social assets or write content, you have to make sure that the user or your audience understands that you're using it safely, and they understand what the payoff is of that. You know, why should they trust your your use of AI, whether that's in your marketing materials or in your actual product? 00:24:47.000 --> 00:25:03.000 Alright, so I've got a really easy takeaway for this one. And I think there's implications across the board, because this is one that hits both sides of the coin for me. There are massive changes to how people are searching and discovering information using AI platforms. 00:25:03.000 --> 00:25:18.000 There's also massive changes happening to what we are trying to get our teams to embrace to be AI forward. This report is really, really dense. There's a lot of great information in it that I think if you're a marketing leader, it's going to help. 00:25:18.000 --> 00:25:34.000 It's going to help you contextualize more of the sentiment of what's out there. So my recommendation on the step to take here, take the report and drop it into Notebook LM. If you have not yet used this tool, it's one of the most popular tools. 00:25:34.000 --> 00:25:59.000 For, you know, for the use of generative AI, what you can do is take this report and have it create a podcast format audio so you can ingest the information that way. You can create a video, you can create a question and answer guide, or you can kind of skim the report and just have a conversation with AI about the report to make sure you're better understanding it. 00:25:59.000 --> 00:26:05.000 And the implications it has on your team, on your marketing strategy, all of it. 00:26:05.000 --> 00:26:26.000 Alright, so I have six things for you all to do next week. I know it's Thursday. Let's catch our breath. Let's wrap the week. But if you really want to get ahead and say, okay, what are the things I need to be doing? I need to be thinking about. These are our recommendations. Pull your own segmented click-through rate study, see how your click through rate has changed over the last 16 months. 00:26:26.000 --> 00:26:39.000 Add or enable brand lift studies to your measurement stack. At least start the conversation. See if you've already got something that your organization is paying for, maybe something your Google reps can help you with. 00:26:39.000 --> 00:26:55.000 Number three, remember the ultimate guide is quote unquote dying. Audit, audit your Ultimate Guide content. Audit your pages that have over 2000 words and see if a little change to content structure can help increase your visibility here. 00:26:55.000 --> 00:27:13.000 Number four, make sure you've got Bing webmaster Tools set up today. A lot of us have ignored Bing for a long time, but there really is a treasure trove of information here. If you are a B2B marketer, you have got to get this done, because I can almost guarantee your audience is searching on copilot. 00:27:13.000 --> 00:27:32.000 for your products and services. Number five, you're going to want to audit how Chrome powered by Gemini engages with your website. Again, this probably has to be on your personal Google account, but think about the activities and the actions you need your audience to be able to take on the website and see how well. 00:27:32.000 --> 00:27:43.000 Google's agents can do that. And then lastly, load that Stanford AI report into Notebook LM and get a recap that you can engage with. 00:27:43.000 --> 00:28:00.000 There's a lot of information here. So our goal is really to share as much of this content and information as possible. But if you're feeling overwhelmed, if you're like, man, how am I going to do all of this in addition to everything I thought I was going to do, I do want to share. We have recently launched this new offering, Seer Advance. 00:28:00.000 --> 00:28:22.000 And our belief is your best marketing is still ahead of you. We want you to be able to focus on that. If you're having difficulty getting your team to be AI forward. If you've got roadblocks with your people, your process, your technology, we're hosting workshops, we're offering partnerships, and Will and I are doing live sessions to help bridge this gap between you and your team. 00:28:22.000 --> 00:28:31.000 So, give you a second to scan the QR code if you are so inclined to learn more about our offering here. 00:28:31.000 --> 00:28:37.000 Okay. Now it's time for Ask Us Anything. 00:28:37.000 --> 00:28:42.000 Let me say. Let me find Q&A. 00:28:42.000 --> 00:28:43.000 Matt Graves. I love it. How are you doing, Matt? 00:28:43.000 --> 00:28:46.000 Bring it on. Let's go. 00:28:46.000 --> 00:28:52.000 Okay. Will I'm gonna get. I'm going to tee this one up for you. Okay? Because I'm going to read it live. 00:28:52.000 --> 00:29:11.000 So the first one's for a friend of the pod, Matt Graves. He said, as Sear has seen enterprise SEO, Geo AEO tools. Are there any that can reasonably estimate or forecast a domain's AI visibility as well as competitor domains? 00:29:11.000 --> 00:29:41.000 forecast. So again, I could be wrong, right? Like… this stuff is so new, so I'm not saying that I'm right, I'll just give you my perspective. Um… I would say there's no way you're going to be able to estimate or forecast a domain's AI visibility for a few reasons. One, I think… we've done the studies on this, but I think a lot of us need to start tracking how many words and how many unique brands are mentioned per answer, because, for instance, we've seen things where, like, ChatGPT has doubled their visibility, or doubled their, um. 00:29:41.000 --> 00:29:57.000 Number of words per answer, which means they showed double the freaking brands. Like, I don't want 17 different brands for… to find a contractor to build a deck on my house, right? So it's like these things are in such a state of flux. Another thing. 00:29:57.000 --> 00:30:06.000 is, let's say somebody built a predictor, and I bet you there's a lot of… so here's the problem. A lot of people are asking the question, so I think a lot of people are going to build tools to answer that question, even though they know that they're shitty. 00:30:06.000 --> 00:30:18.000 Because they're like, well, if the market wants it, we're gonna build something, we gotta give them something, right? So… I just saw that Lily Ray showed how ChatGPT was using YouTube a ton in citations. 00:30:18.000 --> 00:30:21.000 And then about two or three months later, it fell off. 00:30:21.000 --> 00:30:37.000 And then Google AI Overviews is now using YouTube citations. All of a sudden, it went from, like, nothing to, like, boom, it's, like, super high. So I think when we're kind of trying to chase and predict, the things are literally moving so much, so fast, that how could you build a prediction model. 00:30:37.000 --> 00:30:49.000 On anything that's changing at this speed and this much ferocity. So, Matt, that would be my answer there, and if you have a follow-up, feel free to type it up, and we'll try to get back to it on our next go-round. Elisa? 00:30:49.000 --> 00:31:11.000 Nice. All right. Great stuff. All right, Josh Steinle has the next question, and I'm going to tea. It's 2-parter. I'm going to take the second part. Will, I'm going to give you the 1st part as marketers. What might work well today for Geo that could come back to bite us a year or 2 or 3 down the road? 00:31:11.000 --> 00:31:16.000 I think it was what I showed you before, and it's not even 2 or 3 years down the road. 00:31:16.000 --> 00:31:23.000 It's happening right now. So, you know, some of the leaders in our space are doing very low quality. 00:31:23.000 --> 00:31:32.000 tactics to rank higher in AI. And because of the pressure that you all are getting, like you guys are part of our community, like we need you. 00:31:32.000 --> 00:31:45.000 We need to carry the flag and give you the examples, so then you can carry that flag inside your organizations and be like, do you see how they got these results next time somebody asks you, why am I not seeing answers in a week? You're like, I can do that. But then 6 months later, do you see what happened? 00:31:45.000 --> 00:31:59.000 Maybe we don't want that for our brand, so we're gonna try to keep feeding you guys these examples, so that we can stay focused on doing the real work of being seen, believed, and chosen, which is how money gets made, instead of just focusing on visibility, visibility, visibility. 00:31:59.000 --> 00:32:17.000 Yep. Yeah, absolutely. And I mean, to put a a fine point on that, too. We've been pretty consistent about the the listicle format, the self-serving listicle, I think, is going to is already being penalized by Google, and I think we're going to continue to see that. 00:32:17.000 --> 00:32:23.000 And in terms of the flip side of that, so what is going to work long term? 00:32:23.000 --> 00:32:41.000 I mean, I think it's it's very similar to what we've been saying consistently for years and years and years. What is your audience actually need? What is actually going to help your audience? And how can you translate that into a content modality that they find comfortable? 00:32:41.000 --> 00:32:58.000 And make sure that is a part of a system, right? It's not a one-off thing, it's not one activity where you think, okay, what does my audience need? Let me create some content. But how do you create these systems? And that's something we're trying to do at SEER. Will already mentioned it with the. 00:32:58.000 --> 00:33:17.000 the sales transcripts. We now have this pipeline of data where we can pull transcripts for all of our sales calls. There is a treasure trove of information in there. But as Will pointed out, it requires a system to figure out exactly what you do with it. 00:33:17.000 --> 00:33:18.000 Hello? 00:33:18.000 --> 00:33:30.000 And we have a hard time keeping up with that, right? So I think the secret here is creating systems that have your customer's best interests at heart. I think that is hard to, uh, it's going to be hard to disrupt and it's going to be hard to quote unquote, not work. 00:33:30.000 --> 00:33:37.000 in an AI world. Alright, I want to move to the next question. I saw a good one here. 00:33:37.000 --> 00:33:40.000 Also, Elisa, there's some good ones in there. Yeah, yeah, yeah, yeah, yeah, we might have to skip around a bit. 00:33:40.000 --> 00:34:01.000 Yeah. Alright. So let me. Oh, another another friend of Sears, Zane Hussein. So, well, let me tee this one up for you. If you are making a business case to get more resources to track AI visibility. But all you have is an AI visibility score and a tool like Bright Edge. Are there proxies to indicate the value? 00:34:01.000 --> 00:34:11.000 Why do you… why are you waiting for the webinar to hit me up with this? You got the… you got the… you can DM. All right, so I'm trying to make the business case to get more resources. 00:34:11.000 --> 00:34:22.000 to track AI visibility. So like most things, my training always teaches me to ask more questions. So I can't because of the format. So I'm going to try to fill in some holes here. 00:34:22.000 --> 00:34:27.000 If all I had is an AI visibility score. 00:34:27.000 --> 00:34:34.000 I don't even like that, for this reason. My first question would be, what prompts are you tracking? 00:34:34.000 --> 00:34:42.000 Because it is so easy to move someone's visibility score if you can convince them that these are the things to track. 00:34:42.000 --> 00:34:51.000 Right? So… The first thing you have to do is make sure that the prompts that you are tracking are even somewhat representative of what people actually type in. 00:34:51.000 --> 00:35:12.000 which you will be shocked. If you built your prompt… if you built your list of prompts without some kind of routing in what actual customers are saying or doing, there's a higher likelihood that you have made this visibility number a vanity metric, right? Because you're not rooting it, and like, what we're hearing that clients or customers say this, and I'll tell you. 00:35:12.000 --> 00:35:20.000 Um, our UX team just did a study for a client where we asked people to do a task using ChatGPT, the same task. 00:35:20.000 --> 00:35:34.000 The min number of words used was 4. The max? 119 for the same task. And this is the challenge in the space that we're in right now, but I think if you have the UX data, you can start to see, are my prompts. 00:35:34.000 --> 00:35:39.000 even correct. And I think this idea of, but I need a single score. 00:35:39.000 --> 00:35:52.000 I honestly believe, and I did a video about this a while ago, if it's me, I'm overlaying whatever that visibility score is over time, which is how much am I being seen, right? 00:35:52.000 --> 00:36:10.000 The next step in being seen is somebody believing that you're the right answer for them. So you need a believability metric to offset those. So if your visibility's going up, up, up, up, up, yay, it's great, but then you've got another metric that's like, this is flat than what you know is you're tracking things and visibility that is not impacting the next step in the funnel. 00:36:10.000 --> 00:36:12.000 That's the way that I would look at that. 00:36:12.000 --> 00:36:33.000 Yep, I like it. Alright, I've got a really good question. I'm going to answer this one from Kathleen Roarbaugh, who asked, How is AI pulling from social media posts? I mean, it's this is something that is happening, and I learned something really interesting at the financial brand forum, and Jean may be in the audience. 00:36:33.000 --> 00:36:52.000 But she had mentioned she's a marketing leader for a credit union, and she had mentioned that she's got an AI intern who's doing some some research trying to understand how these models work and how their brand is being visible. First of all, I just think that's a genius move. If you cannot secure the headcount. 00:36:52.000 --> 00:37:08.000 Then try to push the need for an intern to come in and just assist the marketing department with a little extra research and make the case for a tool or headcount. But what they found is that their Facebook page was impacting their visibility, especially on Meta AI. 00:37:08.000 --> 00:37:23.000 And I want to use this opportunity to say I do think in 2026, you are going to need to care about meta AI if you are a consumer brand. They have made so many missteps. They've been really public about it with AI. 00:37:23.000 --> 00:37:43.000 But things are shifting. I think they are going to to figure it out to some extent, and even if they never have the best model, if they have something serviceable, you've got to remember the amount of exposure and visibility that Meta has with your audience, especially if you're targeting a younger audience. But really across the board. 00:37:43.000 --> 00:37:59.000 And so these models are looking at all of your digital assets, your entire digital footprint. What's what are you posting about on Facebook? What are others saying about you on these models? Anything that is publicly crawlable, I think is fair game. I would hesitate to say. 00:37:59.000 --> 00:38:22.000 Oh, LinkedIn is really big. You see all of these studies like, hey, LinkedIn is the most cited AI search AI search website. These citations are so volatile that by the time you put your team together and say, okay, here's our plan for going after LinkedIn, a month has gone by and there's a new, you know, now it's back to Reddit or it's to YouTube. So my recommendation is. 00:38:22.000 --> 00:38:44.000 Leverage social media, bring your social media team into the fold on all of this, but do it with your audience's best interests in mind, and if they're engaged in YouTube, go all in on YouTube, right? I think it's going to have a big impact on AI search. Don't try to chase the game of what is the most highly cited social platform. 00:38:44.000 --> 00:38:48.000 Because there's just way too many variables in that. 00:38:48.000 --> 00:38:56.000 Well, Lisa, one other thing I'll just add to that is don't get into marketing whack-a-mole where you read a study and you move your team, and you read a study, and you move your team. 00:38:56.000 --> 00:39:00.000 I think instead, the way that I would look at it is like what Elisa said, like. 00:39:00.000 --> 00:39:11.000 If it was me today, I'd be like, okay, I need some presents in LinkedIn, and that may move up or down, but it is a good source of intelligence and information that's training these data sets. Same for YouTube. 00:39:11.000 --> 00:39:31.000 Same for Facebook, same for Reddit. Now, that's going to shift over time as to which one's doing the most and which one's not, but if you don't have a steady strategy that then hedges on all of them, you're playing whack-a-mole, and by the time you get one team stood up, to Elisa's point, it's like, oh, well, now it's this, and then you're really gonna not have results to show, so be… Be careful. 00:39:31.000 --> 00:39:49.000 Yes. We had a good question from Ryan Forsythe. They asked, how do you go about prioritizing prompts to target similar to how you can see search volume and keyword difficulty in SEO? This is a really interesting question, I think. 00:39:49.000 --> 00:40:06.000 The basic guidance I have here is there is not a quick win source for this, similar to what we have with SEO, and that is a good thing, because if there is a quick win, then all of your competitors are on an even playing field. 00:40:06.000 --> 00:40:30.000 Where I have found the most insightful. information for the least amount of effort and money is using a tool like outset.ai, and you're able to do AI moderated user research with that tool. And I think we've mentioned it a couple of times now in this conversation, literally get get your audience. 00:40:30.000 --> 00:40:52.000 to search for the things you know they need. See how they prompt. See where they start. See their entire journey. Are they starting on ChatGPT and then they're going to Google? That's going to give you ammo for so many different reasons. First of all, your content strategy and how you are tailoring your own content, what information you're answering. 00:40:52.000 --> 00:41:11.000 but also measurement, because if you're consistently seeing that people are using ChatGPT to discover your brand and then going to Google to validate that, see what your reputation is like, and then click through, you can't expect ChatGPT referrals to tell the story of your success in AI search. 00:41:11.000 --> 00:41:30.000 It's got to couple that with organic search. So really, really just a treasure trove of information. I'm talking about $1,000, $2,000, not a huge investment. And you can learn so much more, then your competitors who are trying to use some kind of proxy from their own AI search tracking tool. 00:41:30.000 --> 00:41:39.000 That's really just like a, you know, a dressed up version of monthly search volume from Google. 00:41:39.000 --> 00:41:50.000 And Elisa, what I'll add to that is, I think, um, I've been saying this for so long, you might have seen this, Ryan, but I have this slide that says, like, GEO is not a performance channel. 00:41:50.000 --> 00:41:56.000 And it hit me one day, and I said, wait, the math of SEO was, take the search volume. 00:41:56.000 --> 00:42:03.000 Find a click-through rate curve. Make the assumption that all these different rankings, if I can get them, I'm going to get this much traffic. 00:42:03.000 --> 00:42:08.000 value the traffic at X per visitor against paid search to create a value for the C-suite. 00:42:08.000 --> 00:42:23.000 Then you can say, well, typically, from that content, we now get this many people to sign up for our newsletter. That creates a value. This many people that add to cart, this adds value. This many people checked out. Now, we're so used to that stack, and we've trained our executives. 00:42:23.000 --> 00:42:33.000 to judge us on that stack. So then when we start saying like, oh, Geo and SEO are similar, they're like, great, I'm gonna… and then we get mad when we don't… it's so interesting. We get upset. 00:42:33.000 --> 00:42:39.000 That we try to cover our asses and say that GEO and SEO are the same so that we stay relevant. 00:42:39.000 --> 00:42:56.000 And then when the leadership says, great, I should judge you the same way I judged SEO, now you want to start to back out of it and be like, well, it's not really like that, and I don't get this and that. And they're like, that sounds like bullshit. So I think one of the biggest things here. 00:42:56.000 --> 00:43:02.000 is what I love about this AI search moment that we're in is it's finally going to probably force. 00:43:02.000 --> 00:43:06.000 People always looking for… uh, Lily Ray says it best. 00:43:06.000 --> 00:43:26.000 Loopholists aren't strategists. And SEO for a long time created a lot of, uh, me included, that found loopholes and got away with stuff that now you've got to understand your customer more, and when you can't just say, I'm gonna run a script to pull the MSV and this, this, and that. And the other thing that I'll say. 00:43:26.000 --> 00:43:33.000 that Elisa, um, kind of really… Elisa and John in our company have really been kind of getting me to think about, is, like. 00:43:33.000 --> 00:43:48.000 brands, like, or, uh, uh, uh. What prompts the target, it's like the best prompt to target is the one of you versus another competitor that your customers are likely to use. And then read that answer and say, how much of that is true or not? 00:43:48.000 --> 00:43:53.000 Because that person, you spent all this money to get that person to know your brand. That's a lot of money. 00:43:53.000 --> 00:43:59.000 They actually now know your brand, and are using your brand in a prompt, and we're over here focusing on. 00:43:59.000 --> 00:44:08.000 uh, best checking account, instead of focusing on who has a lower overdraft fee, brand A or brand B. That's the thing. 00:44:08.000 --> 00:44:09.000 Yep, absolutely. 00:44:09.000 --> 00:44:21.000 Track those brands. Oh, track your branded prompts first, and make sure you own your visibility there, and that you have as much influence on the citations of those answers as humanly possible. 00:44:21.000 --> 00:44:26.000 100%. I think I'll have a blog post coming out about this, all about brand accuracy and the importance of that starting point. 00:44:26.000 --> 00:44:28.000 Good. It was your idea, it was your idea, Lisa. 00:44:28.000 --> 00:44:36.000 All right. I want to answer one final question. I want to get everybody out of here at 12:45. That's when we said we were going to end. First of all, I'll say. 00:44:36.000 --> 00:44:39.000 Oh, blue. 00:44:39.000 --> 00:44:57.000 This has been great for us. Any feedback you have will be very, very helpful. We want to keep doing this. We want this to be as valuable as possible for all of you. So the question that I want to end on is Jonathan Will asked how important is having AI governance across the company, not just in the marketing department? 00:44:57.000 --> 00:45:15.000 And what are ways to ensure security, especially when dealing with clients that have sensitive information? Jonathan, I think this is so critically important. And I'll say this as briefly as I can. I do not think it is going to be possible to be successful in marketing. 00:45:15.000 --> 00:45:32.000 If you are not an AI forward marketer, at least in digital marketing, the landscape is becoming way too fragmented. Everything that once was relatively simple and still difficult to do, but relatively straightforward, is getting less and less straightforward. 00:45:32.000 --> 00:45:46.000 Prompts are more complex to track and understand than keywords. There are more levers to pull with AI search. Nobody is getting additional headcount, more headcount than you had in, say, 2020, right? 00:45:46.000 --> 00:46:04.000 I think the only way to be successful in marketing is to truly embrace being an AI-forward marketer, and I can almost guarantee somebody in your organization is going to screw it up and get your tools removed if AI governance is not clearly set, if people aren't trained appropriately, if policies aren't set. 00:46:04.000 --> 00:46:24.000 Because it only takes one mistake for the board, for the C-suite to be like, screw it. No, we're going to hope it's a fade that passes, and we're going to take away access for everybody, and that is going to absolutely destroy your ability to hit your goals and reach your audience in this new era. 00:46:24.000 --> 00:46:36.000 This is going to sound brutal, because I'm literally here, guys, like, I'm doing parts of jobs that used to take me hours before, and the quality is good, and I'm doing it. I'm doing it, so I'm seeing this. 00:46:36.000 --> 00:46:41.000 The best thing you can do… be around a lot of AI-forward people. 00:46:41.000 --> 00:46:55.000 Because it's just too hard, that's why we did this webinar. It's too hard to try to know everything going on, so we're gonna try to be a trusted curator, and try to bring down some of the stuff that we're seeing, but you need to be in circles of people constantly experimenting, and if the company is preventing experimentation. 00:46:55.000 --> 00:47:20.000 And the people around you aren't thinking, I'll get DoorDash tonight, but I won't spend the same $20 to get my personal account so I can learn and be relevant. You need to get in a new circle of people where you can banter around and share ideas, because I think it's like career… It's kind of career suicide to hang around with a bunch of people that are like, nah, this ain't gonna be a big deal, because you'll go down with them, and then by the time you go to catch up, it's probably too late. 00:47:20.000 --> 00:47:31.000 Great. All right. Great questions. We're going to review the questions we didn't get to and see how we can answer them in social. So keep an eye out for that. Thank you all for joining, and we hope to see you again. 00:47:31.000 --> 00:47:36.000 Thank you, guys. Appreciate y'all.