WEBVTT 00:01:31.000 --> 00:01:47.000 The content is great, but there's so much of it. There's so much happening. So we're trying something a little different this week. We are going to focus on three main stories, and then we're going to continue to have our AMA period at the end. We'll try to save at least 15 minutes for that 00:01:47.000 --> 00:02:03.000 So whatever questions you have, we've got the Q&A feature. Drop it in the chat, drop it in the Q&A, and we'll be super excited to answer those questions live. Lastly, again, we really run on feedback. You're going to see a post-Zoom survey 00:02:03.000 --> 00:02:08.000 Please give us feedback, let us know what's working, what's not, so we can continue to make this better. 00:02:08.000 --> 00:02:10.000 All right, so with that, let's get into it. 00:02:10.000 --> 00:02:15.000 Here is what got our gears turning last month 00:02:15.000 --> 00:02:25.000 First of all, a lot happening with ChatGPT ads. Our team has been running these for a handful of clients. We started to get some early data that I just found really, really fascinating 00:02:25.000 --> 00:02:41.000 Simultaneously, our AI search data provider Scrunch started to pull ChatGPT ads data into their data. So we've got these two different data sets we're reacting to. We're going to share some of the visuals, some of the insights 00:02:41.000 --> 00:02:56.000 from that data. And then kind of talk and riff on what that means for brands. We're also going to talk about query fanouts, branded query fanouts. If you don't know what that is, or if you forget, we're going to start by setting the table and explaining why it matters 00:02:56.000 --> 00:03:11.000 Will and I have been going deep on this. Nick Hagler produced a bunch of research for us that really got our gears turning. I think even peak AI and other AI search data provider has started to comment on this. There's something interesting here 00:03:11.000 --> 00:03:25.000 And it does actually directly map back to the last element that we're going to talk about today, which is measuring geo. This continues to be something our clients are asking about, something prospects are asking about 00:03:25.000 --> 00:03:34.000 We are at this really interesting time where we have so much attention being put on GEO as a marketing opportunity 00:03:34.000 --> 00:03:54.000 But not a lot of clarity on how exactly do we measure this? How do we talk about what's valuable versus what's not valuable? And something Will and I keep coming back to is we see these really no-brainer opportunities where it's like, even if this doesn't map back to a KPI, you cannot tell me this is not valuable for a brand 00:03:54.000 --> 00:04:11.000 So we're going to talk a little about that, what we're seeing, and start to talk about what new KPIs could look like. Some of the feedback we got, this was also interesting, is that people are looking for more interactivity. In fact, a few of you wanted homework, which I love that we've got that vibe in the audience 00:04:11.000 --> 00:04:23.000 So we're going to give you an assignment for you to to share your thoughts with us, and hopefully you can get highlighted on the next episode of the Signal. All right. So with that, let's jump in. 00:04:23.000 --> 00:04:41.000 So we're going to start with those early findings for ChatGPT ads. First of all, I have a little data here. As you know, Hannah and the team do a fantastic job recapping all the information we talk about, sending out the visuals. So don't worry about screenshotting anything live. You're going to get access to all of the stuff 00:04:41.000 --> 00:04:49.000 We've got some interesting data points, though. First of all, when you're looking at a blended click-through rate 00:04:49.000 --> 00:05:09.000 These numbers are considerably higher than I thought we would get with ChatGPT, so that in itself, I think, is interesting. Also, as of June, this is not just an awareness play. We're seeing CPM data, we're seeing CPC data. So there's some interesting data coming in that starts to enable us to benchmark and think about, okay. 00:05:09.000 --> 00:05:19.000 What is the value of this for getting visibility, paid visibility on ChatGPT prompts and topics 00:05:19.000 --> 00:05:35.000 One of the key learnings, and this is coming from Brittany Sager. She did an incredible job, first of all, piloting this with our clients, but then also keeping the agency informed of what's working. One of her biggest observations was to bid to the prompt, not the page 00:05:35.000 --> 00:05:51.000 And what she means by that is we aren't creating ads and creating campaigns for a blank space. We are adding ads to the bottom of conversations. And so what is most likely to be said in those conversations 00:05:51.000 --> 00:06:08.000 Especially what's most likely to be included in the response is highly relevant to what your ad includes. We have some examples of where we're seeing this work really well, and some examples of where we're seeing other competitors really miss the boat 00:06:08.000 --> 00:06:24.000 Now, this pulls in some of the data that we started to see in Scrunch. So looking at all of our tracked queries within ChatGPT for this period of nine days in July, we're starting to see this data come in 00:06:24.000 --> 00:06:36.000 And we've at least got a benchmark here. We're up to 7.7% exposure of ads on the keywords, or the prompts, rather, that we are tracking in Scrunch. 00:06:36.000 --> 00:06:53.000 We also have some early data on what industries are most commonly being saturated by ads. Not really a surprise, B2B software and SaaS, I would expect that to be number one. It is interesting how much farther ahead they are 00:06:53.000 --> 00:07:02.000 Then, for example, cybersecurity, privacy, GRC. We also see comparison and review sites, legal services 00:07:02.000 --> 00:07:17.000 When I look at this data, I see a lot of white space, a lot of opportunity for brands that fall in this middle to bottom group of categories because there's just not a lot of activity right now happening in this space 00:07:17.000 --> 00:07:24.000 Which means you as a brand, if you've got some incremental budget for testing, it's a lot of learnings that can happen here. 00:07:24.000 --> 00:07:37.000 By the way, in June, it's believed that ChatGPT finally crossed that threshold of 1 billion monthly active users. So when you think about how big is the pool of users here. 00:07:37.000 --> 00:07:49.000 That's really big and especially if you're targeting folks in B2B, if you're targeting knowledge workers, if you're targeting consumers who happen to be a higher affinity to AI search 00:07:49.000 --> 00:07:59.000 As always, SparkToro is a really good tool for identifying how much or how little is your audience engaging with AI search. 00:07:59.000 --> 00:08:05.000 All right, couple more thoughts on this, and I'm going to get Will on the mic here. 00:08:05.000 --> 00:08:21.000 But one of the questions we had was how well are these ChatGPT ads aligning to the answers above them? And we saw one example that I think is a really, really good gold star example. So for a prompt, which hotel brands have the best rewards programs 00:08:21.000 --> 00:08:30.000 We then saw an ad for Hilton stay perks and Hilton Honors was named as the number one answer. 00:08:30.000 --> 00:08:48.000 That is a really, really solid piece of brand activation or brand prominence that I think really can move the needle for your audience. How do you measure that? We're going to talk about that in the in the third section here. We also see really a varying degree or kind of a degrading 00:08:48.000 --> 00:09:05.000 scale of other brands missing the mark a bit, and I didn't include the the 2 that were most misaligned. You can check out our blog post. Maybe Hannah can drop a link in the chat to that blog post and see all the information that Nick Hagler put together 00:09:05.000 --> 00:09:11.000 But we also see things like a user is asking for coaching workflows 00:09:11.000 --> 00:09:29.000 That leads to an ad for an AI native CRM that is a huge mismatch, right? That's really interesting because I feel like I'm back in the early days of Google ads where it's not it was not uncommon to see an ad that just fully was misaligned with what user intent was 00:09:29.000 --> 00:09:40.000 Early days of ChatGPT. And so a lot of opportunity to get those early learnings and start to understand what works, what doesn't, start to establish that moat 00:09:40.000 --> 00:09:56.000 Last thing I'll call out here is what we're seeing. This is an example of, I don't know, a little arbitrage in our space where there's a prompt where SEER is recommended as the strongest brand to consider 00:09:56.000 --> 00:10:16.000 And we have a competitor bidding on that response. This is interesting to me, not necessarily because I think it's a high quality way for this competitor to earn market share. I could be wrong on that, but what makes me think about is, man, how much more valuable could it be if we ourselves were bidding on this space 00:10:16.000 --> 00:10:32.000 Thinking about what do we know our audience is looking for? What have we heard in the last 90 days of sales calls, transcripts, pain points about whatever this prompt relates back to? How could we be really intentional with the messaging within the ad copy 00:10:32.000 --> 00:10:41.000 To really put a much firmer stamp of influence on this this surface 00:10:41.000 --> 00:10:56.000 So that's all of the data that I've got for ChatGPT ads. I want to turn it over to you, Will, and get your thoughts on what you're thinking early days with ChatGPT ads and the opportunity here for brands and search marketers. 00:10:56.000 --> 00:11:01.000 Sure. Thanks for handing that over. So the first thing I think of is if you're in B2B, 00:11:01.000 --> 00:11:16.000 You know, it's funny, we're so early days that I think it's the critical thinking about what numbers we're showing that make you kind of, like, poke holes and whatnot, and we'd love to hear those, right? You know that's the job now, is like critically think about these things. One of the things I started realizing with B2B 00:11:16.000 --> 00:11:19.000 when I was looking at it, is I'm like, hmm. 00:11:19.000 --> 00:11:38.000 A lot of B2B companies, I believe, are willing to pay the $20 a month for somebody to get the version that gets no ads. It's like this is the kind of thing that you also have to start to think about when you look at these metrics, right? Like, oh, it makes sense that B2C might be higher because a lot of people that are using ChatGPT to find products and things like that 00:11:38.000 --> 00:11:44.000 That could be anybody, right? But if you have an account through your employer 00:11:44.000 --> 00:12:01.000 and you use that account for things, then you don't see the ad. So keep that in mind, at least for now is the 1st thing. I'm a little bit more willing to just put people, it's not even putting people on blast. If I'm going to tell you that Hilton did a great job at something, I'm going to be like, oh, here's the other company that's not. So like, don't follow them 00:12:01.000 --> 00:12:07.000 Right? And what we saw was, like, Monday.com was, like, showing up. It's funny, because I've been tracking their 00:12:07.000 --> 00:12:21.000 Their ad presence in paid search against organic, and it's like, oh, it's the same playbook. It's like broad-based, like, everywhere, right? And it really is not poorly matched. So what I think is, is the same companies 00:12:21.000 --> 00:12:26.000 who have seen poorly matching ad relevance in paid 00:12:26.000 --> 00:12:41.000 tend to bring that same scattershot strategy over here, but yet people have given them so much more rich information about what they actually want, and they're just kind of ignoring that, is one of the things I saw there. The other thing that we noticed when we looked at this data is it's back to arbitrage 00:12:41.000 --> 00:13:03.000 Just like Elisa was saying. So think about it. There are industries where everybody drags their feet and they're a little slower, they're more regulated, whatever. And that's why we ended up with sites like NerdWallet, right? We ended up with these middlemen because banks were like, well, I can't produce this content, I can't produce that, I gotta check it out. So these people arise from the ashes to be like, because you can't move fast enough 00:13:03.000 --> 00:13:08.000 Aya now will sell you leads so you can hit your numbers 00:13:08.000 --> 00:13:27.000 What that makes me think of is, and what I'm seeing in the data is you're seeing all these sites that aren't brands literally showing up for these branded searches because they're like, oh, none of them are here yet. None of them have decided to take up the real estate, so I don't have to compete with them, even though they're better answer and typically like ad quality scores and costs will be like, well, shouldn't we show the brand? 00:13:27.000 --> 00:13:42.000 But if you're not there at all, shouldn't we show the person to sell the leads back to you? And they're also selling those leads to everyone else as well. So remember, like you might want to especially get there and own your brand. And the last thing I'll get everybody thinking about is 00:13:42.000 --> 00:13:45.000 And I don't have answers for this 00:13:45.000 --> 00:13:52.000 But Google gets a lot of clicks on ads. A lot of times when I look at ChatGPT ads 00:13:52.000 --> 00:14:11.000 The only thing I think about is I have to ground my client in what they're already used to. You anchor people in things they're already used to. So it's like, how do these numbers look against Google, I think is the biggest thing. And one of the things that you have to keep in mind when you think of that is Google's gotten really good 00:14:11.000 --> 00:14:16.000 And showing the ads at the top in a way where nobody knows their ads. 00:14:16.000 --> 00:14:33.000 So if you start to try to compare click through rates and things like that, you've also got to think of the audience, like B2B might not be seeing the ad, whereas we can't say there is no program for Google where I can pay to never see an ad. So you've got to think that that's going to affect the strategy, that's going to affect your numbers if you're comparing these two 00:14:33.000 --> 00:14:40.000 And then the other part is the placement of the ad. So these ads are at the bottom, which means, at least for now. 00:14:40.000 --> 00:14:42.000 I got a really good answer above 00:14:42.000 --> 00:14:59.000 You got to come with something quite different. Like I didn't even get to see your brand. Whereas in Google ads, I get to see all the brands, I get to see your bad messaging that doesn't really match what I wanted, but I get to see all of it. I get to accidentally click on it. Like all of those things are different in this world when you're comparing the two. So that's a few things that I would just drop 00:14:59.000 --> 00:15:05.000 For people that are kind of starting to experiment in this space. 00:15:05.000 --> 00:15:22.000 That's really good stuff. Thank you all. The one thing you made me think of there too is who within your organization is most likely to lead the charge in buying these ads? And maybe it's your performance marketing team. I would make the case that this is a really good investment 00:15:22.000 --> 00:15:43.000 for your organic search team or your geo team, because this is an opportunity to buy data to better understand how many people are searching for different topics. You don't necessarily get the exact prompts that people are searching, but you do get fairly good data on how many impressions you're getting across different topics that you're targeting 00:15:43.000 --> 00:15:53.000 So if you have a C-suite who's like, I'm not investing in this until I know what the search volume is, this could be a really good way to get that data. 00:15:53.000 --> 00:16:09.000 All right, I want to talk about branded query fanouts because this is maybe the most thought-provoking thing that I have seen us talking about at SEER recently. So first of all, what are we even talking about? What is a branded query fan out? What is a query fan 00:16:09.000 --> 00:16:25.000 This is a very high-level model of how I think about AI search operating. A user has a prompt, right? And this applies to ChatGPT. It applies to Gemini, all of the LLMs. In this first layer, there is a corpus of data 00:16:25.000 --> 00:16:40.000 that these LLMs are trained on, pulled from Internet archives, pulled from all kinds of sources, right? So if you are a big brand in the 90s and the early aughts, you are probably outweighted in that corpus of data. That's good if you're that kind of brand 00:16:40.000 --> 00:16:56.000 That's bad if you're a very new brand. Then you have this next layer, and we've been calling this invisible criteria at SEER. So these models don't just go by what the pure data they were trained on. There's now reasoning that is baked into these models 00:16:56.000 --> 00:17:11.000 And there's a level of post-training alignment that happens where the model is trying to better understand, okay, if the user wants this, and I know this about the topic, what else do I need to consider to make this a really good response? 00:17:11.000 --> 00:17:27.000 Then the third layer is live web retrieval or validation. And often this is going to happen if you have a very niche topic, if you're looking for the most up-to-date information, there's going to be a point where the LLM runs 00:17:27.000 --> 00:17:38.000 search queries to pull back the freshest possible data. So what we care about and what we're talking about right here are what are those search queries that the LLMs are using 00:17:38.000 --> 00:17:46.000 Query fan out, grounded queries, all of this then combines to create the answer, the generated response. 00:17:46.000 --> 00:17:57.000 So let's take a query like this. As a VP of growth looking to future proof our search strategy, I want to know which agencies are leading the conversation on AI search rather than just reacting to it. 00:17:57.000 --> 00:18:11.000 And we see the answer here. We're like, all right, great. Seer is included. I pull rank. Yes, Mike posts a ton of stuff. Lily Ray posts a ton of stuff with AMCIV brain lab siege. All these are our players I would expect to see. 00:18:11.000 --> 00:18:31.000 But if you are working in Geo, what your mind, I think immediately goes to is where did this come from? Is it about the pre-training? Is it about the post-training? Is it about the web retrieval? What is causing this? Because if you remember, a year ago, these were not the brands that were mentioned. And obviously, I'm biased, but I think we were still 00:18:31.000 --> 00:18:45.000 This was still the right answer a year ago, but this was not what the models were including. So what we are learning is we are now seeing branded queries within query fan out. So what do I mean by that? 00:18:45.000 --> 00:19:02.000 What we did is run the same prompt, the prompt I just went through. We ran it 30 times and looked at all the query fanouts because every time a prompt runs, you're going to see different query fanouts. It's a probabilistic system. It's not deterministic. And this is one of the many implications of this 00:19:02.000 --> 00:19:07.000 And what we found is iPoolRank as a website 00:19:07.000 --> 00:19:26.000 was mentioned 97% of the time for this same prompt run 30 times. Mike King as an entity was named 63% of the time. If you remember, Mike was one of the first people to write about, I don't even remember what he called it at the time. It was like pre-branding of what all of this is 00:19:26.000 --> 00:19:42.000 But his agency came out with some content pretty early on. And I have to believe that's part of why he is so highly cited in these branded queries. We also see Seer, we see Amsiv, we see Search Engine Land 00:19:42.000 --> 00:20:00.000 So this is the model saying, okay, I've got all my pre-training data. I've got my post-training assumptions, and either through some combination of those things, I know in order to validate this query, I want to see what Mike has to say. I want to see what search engine land has to say. I want to see what Lily Ray 00:20:00.000 --> 00:20:13.000 And Will at SEER have to say, this is really interesting to me. It's really interesting to Will too. So I want to turn it over to you, Will. How does this get your brain cooking? What do you think about this concept? 00:20:13.000 --> 00:20:17.000 Yeah, one sec. There we go. Yeah, like this 00:20:17.000 --> 00:20:21.000 To me, this is like, everybody should stop for a second 00:20:21.000 --> 00:20:24.000 Look at what you're seeing 00:20:24.000 --> 00:20:25.000 And be like 00:20:25.000 --> 00:20:32.000 How does that change everything that I believe and think about how to optimize a site for visibility? 00:20:32.000 --> 00:20:34.000 Because 00:20:34.000 --> 00:20:44.000 Somehow, what we're seeing with GPT-5. And all the new models, I believe in 5.6, we haven't done the research, but I built something that just hit 5.6 for this. 00:20:44.000 --> 00:20:47.000 Like, they're still sticking to this concept of, hey. 00:20:47.000 --> 00:20:54.000 Let me actually go back a second, because I need to connect the dots for you guys, so that you can connect these dots when you're in meetings 00:20:54.000 --> 00:20:56.000 When Google 00:20:56.000 --> 00:21:12.000 Came out and was Google, right? And they dominated everybody. What made that happen was the fact that they realized that there was certain things were too easy to spam on Lycos and Alta Vista and Hotpot, right? People could just put the keywords in the keyword tag and boom, they ranked. So Google came up with a new thing 00:21:12.000 --> 00:21:23.000 What the AI engines have definitely realized is, wait a second, in a lot of places, if you just created a roundup listicle and promoted it in enough places, you were the answer. 00:21:23.000 --> 00:21:27.000 Now, what they know is we need to grow our monthly active users 00:21:27.000 --> 00:21:46.000 And they've already seen this happen before with Lycos and Hotbot and Excite and AltaVista and Northern Light and Infoseeq and all those other engines got beat because people were like, you know how easy it is just to start typing in Google instead of Excite. So because of the ease of switching over to another model 00:21:46.000 --> 00:22:02.000 These companies are like, oh, that lesson's already been learned. So I got to fix this spam problem early before people just start being like, well, I guess I'll just go to ChatGPT then because Claude is spammed the holy hell. So every one of these folks are trying to figure out how do we create a higher quality answer 00:22:02.000 --> 00:22:11.000 And what ChatGPT is doing is they're saying, well, maybe what I can do is figure out what brands are highly associated with this prompt, just search their site 00:22:11.000 --> 00:22:22.000 You can't beat SEER Interactive or iPoolRank for a site colon, I pullrank.com and a prompt. You can't because it's only searching my site, right? 00:22:22.000 --> 00:22:26.000 So I think for many of us, the hardest thing is kind of a little bit what Margie mentioned 00:22:26.000 --> 00:22:29.000 Which is 00:22:29.000 --> 00:22:38.000 Companies that have invested for a long time in helping out people in their community to learn have now built a moat 00:22:38.000 --> 00:22:50.000 Tough. Find some niche that we're not talking about or Mike's not talking about or Lily's not talking about or Dejon's not talking about because we've already got that moat and you can tell we had a mention from search engine land today 00:22:50.000 --> 00:22:59.000 through some of our research, and we had a mention from search engine land on Monday with some of our research. We're not going to stop producing what was already in our ethos 00:22:59.000 --> 00:23:09.000 So I think companies are going to have to find niches if this is the real world that we're in for certain types of prompts, I think companies are going to have to find niches to try to be like, well, where are they not paying attention? 00:23:09.000 --> 00:23:26.000 Right? Why do you think we're doing all this research on the integration of paid ChatGPT and organic? Because it's something the models already know about us and I want to build a moat around it before somebody else starts to talk about it, right? So like these are the kind of things that I think are just really, really, really important to be thinking about 00:23:26.000 --> 00:23:34.000 How do I become a brand that AI hardwires into this topic? And 00:23:34.000 --> 00:23:40.000 Yeah, and I think the hardest part is when somebody says, what's the ROI of that? The answer is I have no freaking idea 00:23:40.000 --> 00:23:42.000 But it's everything 00:23:42.000 --> 00:23:57.000 Right? And I think that's where a lot of companies who want to go, well, I want to be the I want to be the company and the fan out. It's like, that is six, seven, eight, nine, 10 years of publishing things that people would say, shouldn't you keep that to yourself and keep it proprietary so nobody knows? And you go, no. 00:23:57.000 --> 00:24:13.000 We're going to put it out there, right? And it's years of that that led to this, and I think most companies are going to reward us as marketers, quarter to quarter, which means you'll never get the opportunity to really do this, and to see it come to fruition unless we change the way that we're rewarded and the KPIs that we track. 00:24:13.000 --> 00:24:14.000 Yep. 00:24:14.000 --> 00:24:15.000 I love it 00:24:15.000 --> 00:24:32.000 Will, the one thing you mentioned, too, that I want to double click on is the idea of, you know, if you're competing with us at SEER, maybe you need to go more niche. The other thing is, think about spaces where there is a clear market leader, and you've been able to win based on your dominance in the space 00:24:32.000 --> 00:24:52.000 I think even that is a risk in an AI search world, because if you're for everybody, you're for nobody. These systems are really trying to be matchmakers and say, Okay, you know there's here's the prompt you said, and you're a VP of growth, but obviously there's personalization that has a factor in this, and they're trying 00:24:52.000 --> 00:25:09.000 to think, okay, for you specifically, what are you looking for? And if you are, you know, let's say you're one of the big hold coasts, like, well, who are you for? What specifically do you offer is baked within your culture you're offering that is going to be the best possible fit 00:25:09.000 --> 00:25:24.000 So I'm very interested to see how this plays out. We work with some of these such type brands. Right now, we see them dominating in AI search. A year from now, I don't know if that's going to be true unless they really start to clarify their UVP 00:25:24.000 --> 00:25:25.000 and who they are for. 00:25:25.000 --> 00:25:41.000 It's funny, I have a recurring email coming into my inbox to remind me every week that Sear needs to publish somewhere on our website who we're not for. It literally comes in my inbox every week and I'm waiting to be like, all right, now I'm ready to write that because 00:25:41.000 --> 00:25:57.000 Someone did the research on this a while back, and if people really want it, I'll just go through my Readwise and find it. I can just ask Claude. So it's not that much work. Someone was like, I just caught our company doing this the other day, right? It's like, I want every statement on Sears site to be backed up by 00:25:57.000 --> 00:25:58.000 a receipt 00:25:58.000 --> 00:26:10.000 Because what happens is everybody's going to go to Claude to write some marketing copy. And it's all going to sound the same. And AI has gotten really good at being like, if I go across a bunch of brands 00:26:10.000 --> 00:26:16.000 And they all sound the same, then I'm just gonna take that out anyway, and I'm just gonna put in the things that are unique. 00:26:16.000 --> 00:26:33.000 How many of you are going to ask questions about SEER and go, and you're going to see B Corp, you're going to see the work we do in the community. You're going to see our numbers on our account turnover. And you're like, oh, why, why would they show those things about SEER? Because I'm not afraid for the world to know what my account turnover is or my client retention is 00:26:33.000 --> 00:26:42.000 Right? So I'm trying to figure out what are the unique things about me that when everybody else keeps writing the same bullshit copy that sounds like everybody else, and it's like 00:26:42.000 --> 00:26:56.000 We're great for everybody and we do every client and we do all of this. I'm like, we need to make sure that every sentence on our website, that should be what we aspire to be. Like, there is a receipt. So if somebody ever said to us, you say this on your website. 00:26:56.000 --> 00:27:13.000 I could go, okay, here's the receipt. Boom. We actually have done that. We actually did work with that client. We actually did have them for this long, right? It's almost like a ledger of, like, saying that what we actually… what we are saying on our website is true. So, yeah, it's like… it's just, like, one of those moments where, like, you're watching marketers 00:27:13.000 --> 00:27:28.000 write copy that sounds like everybody else's, and then going, why don't I show up in AI search? And I'm like, why would they show you, one, why are we hallucinating? One of the biggest reasons why it's hallucinating is it doesn't know who to attribute your brand to, because you sound like everybody else. 00:27:28.000 --> 00:27:45.000 And it's funny how we want to blame the LLM for having hallucinations, because that absolves us from being part of the problem. And it's like, you don't post anything that's unique enough for me to know what attached to your brand, so I'm misattributing stuff to you. And we go, they suck. And it's like, no, maybe we suck, and we control our websites. 00:27:45.000 --> 00:27:59.000 So let's do… let's fight the good fight within our organizations to try to, like, be like, guys, like, you may not want to hear this, Mr. or Miss CMO, but this is the reality. We are doing nothing 00:27:59.000 --> 00:28:10.000 To put a receipt on anything that we said on our homepage about who we're for, on any of our services pages. Well, we don't want people to know that. No, that's what we actually have to write. 00:28:10.000 --> 00:28:11.000 Yeah. 00:28:11.000 --> 00:28:26.000 Right? So that's what I believe, man. I get fired up about this , because it's like, man, like, like, like, we're so early that we've got a chance to correct the ship and to do the right work as an industry, and I'm watching us not have the conversations with the C-level to be like, this is the reality. I can't go call Sundar and be like, can you change 00:28:26.000 --> 00:28:28.000 Your algorithm for me, please. 00:28:28.000 --> 00:28:45.000 Yep, yep. And it's a new reality, right? Like there's a time, even three years ago, if you had enough budget, if you had enough influence, even if you did not deserve to be the number one answer, you could manufacture a way to get to that number one result in Google 00:28:45.000 --> 00:29:03.000 You would have to be a little more risk tolerant, right? And it's not necessarily the work that we've ever done at SEER, but it was very possible. I don't think that's possible in an AI search world. Today, July 16th, 2026. Yes, I think the listicle shift. There's stuff you can do 00:29:03.000 --> 00:29:07.000 In the future, I do not think that is going to be possible. 00:29:07.000 --> 00:29:23.000 One thing I think people are going to be wondering based on that segment is, okay, how do I find the query fan-out data? How do I get this? So I have three tips for you. Keywords Everywhere has an extension. I've used that, I like that 00:29:23.000 --> 00:29:39.000 There's this SE ranking tool that's free. I think Hannah can drop a link in the chat so you can check this out. And then our friend Nick Eigler, he just offered this to me in Slack because right before this session, I was like, Hey, I pull this data for me. And he's like, hey, here's how you do it yourself 00:29:39.000 --> 00:29:55.000 So super easy. Again, we'll send all this out. Don't worry about taking notes and screenshotting. But there's a lot of ways to tap into this data. What is most exciting is the vast majority of marketers do not even know this exists. If they know it exists, they don't know how to access it 00:29:55.000 --> 00:30:01.000 So this is an area where if you truly are on the cutting edge of all of this, you really do have an advantage. 00:30:01.000 --> 00:30:18.000 So the last session we have here, last segment is measuring geo. So I have a few kind of provocative slides here to get us thinking about how do we answer this question? How do we go to our CEO if they are not a believer in geo and AI search 00:30:18.000 --> 00:30:24.000 And get the required budget to do the testing that is required to win in this space. 00:30:24.000 --> 00:30:40.000 And what I believe is you're going to get some questions like this. Let's say I went to see… let's say that Will doesn't run SEER, and I see this opportunity where SEER is the number one recommended agency for this prompt, and I'm like, man 00:30:40.000 --> 00:30:55.000 I think we can have a much bigger impact on our audience if we pull in those insights from our sales call transcripts and we speak to the pain points we know our audience has. It's a really nice one-two punch. What's the question I'm going to get 00:30:55.000 --> 00:31:01.000 Well, what incremental value can I expect if we bid on this prompt? 00:31:01.000 --> 00:31:03.000 I don't have an answer for that 00:31:03.000 --> 00:31:20.000 Here's another one. So if Mike sees this, I think he's going to be like, yeah, like this is great. I won. Let's imagine Mike is answering to a board. What attribution model should Mike be using to measure the impact of this? And Margie had a good quote in the chat I saw 00:31:20.000 --> 00:31:36.000 Well, what if a lot of this has to do with how present you all are on the conference circuit? So how do you take the investment somebody like Mike or Will has made in speaking many, many, many days of many, many, many years to lead to this outcome 00:31:36.000 --> 00:31:42.000 I don't know that that's the question we should be asking. I think it is the question you're going to get though 00:31:42.000 --> 00:31:58.000 And then, what's the business value of high-intent Copilot impressions? Because now we're starting to get this data. You can go into Bigmaster Tools, you can see the grounding queries, you can see how many times your brand is cited, what is the value of that 00:31:58.000 --> 00:32:10.000 So, Will, what do you think about when we talk about, and I'll go to the next slide here. This is kind of our… what we think you all should do with this information. When we think about measuring GEO, 00:32:10.000 --> 00:32:13.000 And the balance between 00:32:13.000 --> 00:32:30.000 We all know the business needs to make money. We know we need to have confidence in the investments we are making, but you cannot measure everything, right? And performance marketing, I think, has been on one end of the spectrum for a long time. And I think we have to start winding it back 00:32:30.000 --> 00:32:34.000 To be more middle of the road or treat it more as a brand investment 00:32:34.000 --> 00:32:37.000 How do you do that? How does our audience do that? 00:32:37.000 --> 00:32:55.000 I think it's just a weird time, guys. Like, let's just be honest. Like, it's a weird time. Like we don't have any of the metrics we used to have. Trying to track this stuff is really hard. So I think that we first need to have the conversation. And that's what Elisa and I are trying to arm you with is some of the internal banter 00:32:55.000 --> 00:33:07.000 That we get to have at SEER, because we've got 150 people bantering this stuff internally and try to bring it to our larger community. Because any answer you come up with on your own as to how to track AI visibility is wrong 00:33:07.000 --> 00:33:11.000 You have to get other people's perspective on this. The world is changing too fast 00:33:11.000 --> 00:33:22.000 to stay on top of it all. We all admit that, so it's like, you gotta, like, try to get your brain cooking. So what I'm what I like to do is show my clients the old world. Remember, everything is framing 00:33:22.000 --> 00:33:42.000 We have sold SEO to clients for 20 years, framed around, I know your monthly search volume, I can figure out what a click-through rate curve is from AWS ranking or whatever that is, and I can start to build you an ROI model. So people have been buying our services for 20 years. I know exactly what I'm getting from SEO 00:33:42.000 --> 00:33:54.000 Now, the same people who used to buy SEO are now buying GEO, and they're applying the old frame to it, and it's your job is to be like, you know how this used to exist? That doesn't exist anymore 00:33:54.000 --> 00:33:56.000 Like, if you find it. 00:33:56.000 --> 00:33:59.000 Tell me where to find it because I can't find it 00:33:59.000 --> 00:34:08.000 Then you take, like, alright, click-through rates. Like, look at what we're showing today, like, so everything is proof points. So I want to give you proof points 00:34:08.000 --> 00:34:10.000 Show someone 00:34:10.000 --> 00:34:14.000 What happened to your traffic click through rate 00:34:14.000 --> 00:34:18.000 The day that ChatGPT changed the layout 00:34:18.000 --> 00:34:27.000 Because they changed the layout and some of our clients got 400% more traffic to their homepage. So did you do something that week to make those numbers go up? No. 00:34:27.000 --> 00:34:42.000 ChatGPT changed their layout. That's an example to show an executive to go, if I did that, would that be good? And they'd be like, yeah, that's what I want. And you're like, I have no control over that. And they go, oh. And you're like, yeah, yeah, yeah, yeah, I had no control over that. There's nothing that I did to make that happen. 00:34:42.000 --> 00:34:47.000 Right? And you start to try to give them real solid proof points to be like 00:34:47.000 --> 00:35:02.000 So what KPIs could we use? And then you can sit on their side of the table and say, the last thing I want you to do, Mr. or Mrs. VP or CMO, is come up to the cease to come up to the C-suite with your visibility report and everybody's going to love you for the first three or four months 00:35:02.000 --> 00:35:13.000 Because that's what they asked for. But that's not really what they're asking for. Because their frame is that, oh, just like SEO, if my visibility goes up and my rankings go up, my traffic goes up, my conversions go up 00:35:13.000 --> 00:35:16.000 So they're thinking visibility is what we need 00:35:16.000 --> 00:35:23.000 Then you show them the visibility report, and then 6 months later, they're coming to you being like, yo, that thing didn't do this thing 00:35:23.000 --> 00:35:39.000 And now you're on the defense trying to backpedal into like, well, I was on this webinar with Will. It's like, no, no, no. Take that today and start planting the seeds today. And then what I would also recommend you do is to proactively build a report that shows visibility 00:35:39.000 --> 00:35:55.000 And then lay another line in that's like traffic or newsletter signups or maybe conversions that says like, hey, if my visibility goes up, executive, what do you think that's going to do for the business? Because they're all going to ask for visibility 00:35:55.000 --> 00:36:05.000 So you have to ask the question of what do you think that the visibility increase is going to do for the business? And then when they say more traffic, you get to go, oh no, that's not actually going to happen. 00:36:05.000 --> 00:36:23.000 Oh, more conversions. Well, that's not going to happen. So then what you do is you say, would it be helpful for me to give you a report that shows those two things against each other, so we have an early warning signal that says, hey, if my visibility's going up, but I'm not seeing people sign up for my newsletter more, I'm not seeing them add to cart more, I'm not seeing 00:36:23.000 --> 00:36:34.000 dig deeper on our site, then that is a vanity metric that's not actually tied in the reality of what our customers are doing. Now, if they both go up together, it's not a vanity metric for you, because you pick the prompts that actually are driving the outcomes. 00:36:34.000 --> 00:36:46.000 But if you pick the wrong props and it drives a lot of visibility, and it doesn't drive the outcome, you want to be ahead of this. So you want to already have a report that overlays those two things against each other in one graph 00:36:46.000 --> 00:37:00.000 Sea levels have very low attention spans, because everybody wants our time. So we have to be like, what is the one slide you can show me that tells me if this thing goes up that the other thing I already told you I believed is also going up. That's the way that I would approach that. 00:37:00.000 --> 00:37:16.000 I love that. And that's actually… it's a perfect segue. That's what we want to hear from this audience, and this is something we riff on a lot, but even going back to the example I showed of the percent mentions in… of branded query fan-outs in target prompts 00:37:16.000 --> 00:37:31.000 That is not a metric anybody is tracking today. What I keep trying to think about is, what are those metrics 5 years from now? I'm going to be talking to, you know, the next group of interns at SEER who are born in, you know, 2015 or whatever 00:37:31.000 --> 00:37:46.000 And I'm going to be talking to them about all the crazy ways we used to try to measure success, some of which I think are still going to exist, a lot of which we're going to say it was crazy that we tried to do that. We really, really tried to hold on to these attribution methods 00:37:46.000 --> 00:38:06.000 That worked fine in a pre-privacy world where you could cookie everything and really measure everything to the nth degree. And then we refused to let go of them. And so we made the wrong investments often. Think about your branded paid search. There's a huge value in doing branded paid search correctly and protecting the brand 00:38:06.000 --> 00:38:23.000 But that doesn't necessarily mean that that paid search is your most valuable channel. That brand equity has to be built in some other way in order to be valuable enough to protect it. So we want to hear from you all about what are those metrics that you're starting to think about, you're playing with 00:38:23.000 --> 00:38:29.000 Things like branded search volume and things that you can measure, maybe not easily, and we can help 00:38:29.000 --> 00:38:41.000 But you can measure it. You think it does define, yes, this is good or no, this metric is bad. And it's not going to directly connect to ROI. It's not going to directly connect to the bottom line impact 00:38:41.000 --> 00:38:53.000 But it is going to be one of those early leading indicators. We can look at it and communicate to a board to say, yeah, if you believe in this channel, here's one of the metrics that says, yes, you're doing good. No, you're doing bad. 00:38:53.000 --> 00:39:15.000 All right, I want to move to Q&A. We got really, really good questions in here. First one comes from Rob Leslie from George Fox University, and I believe he asked it previously in a chat with Hannah. Rob wants to know, is it really worth it to chase citations and mentions for non-branded informational AI prompts? 00:39:15.000 --> 00:39:28.000 Only once you've tightened up your brand. And that's not what anybody wants to hear. But what I like to the way I'd like to say this to executives is how much money do you spend on your brand? 00:39:28.000 --> 00:39:42.000 And for a lot of people, it's a lot of money, right? And it's not just like putting your name on a stadium. It's having customer service reps that know what they're doing, right? That's part of your brand, right? They represent your brand and they spend 00:39:42.000 --> 00:39:44.000 Millions a year on my brand. It's like, okay. 00:39:44.000 --> 00:39:52.000 When I typed in you versus your top competitor, here are all the things that the AI have to say about that 00:39:52.000 --> 00:40:08.000 How much of this is true and not true? And they'll be like, well, that's not true. And I'm like, okay. So if somebody already knows your brand and your competitor's brand, all that branding stuff worked. They knew your brand from the jump. They didn't go like, who's the best geo agency? They said, I'm picking an agency for these reasons between this one and this one 00:40:08.000 --> 00:40:26.000 So you have to look at that answer and go, do I want to start off on unbranded stuff, which is this quagmire of models are changing, fan outs are coming, there's already entrenched players. While I ignore the like, wait, there's already people who know my brand 00:40:26.000 --> 00:40:42.000 Let's make sure that when ChatGPT gets helpful and just serves up weaknesses about my brand that they didn't even ask for to be balanced, you know, I should probably make sure that those things are representing my brand appropriately, and if they aren't, that's the first work I should do, while I'm also 00:40:42.000 --> 00:40:50.000 playing in the field of how do I show up for non-branded prompts as well? But I think 00:40:50.000 --> 00:40:56.000 The other thing that will help you not be so scattershot, and I think only brave consultants will do this 00:40:56.000 --> 00:41:01.000 They will pull their client aside and go, between you and me, we know we're not good at everything 00:41:01.000 --> 00:41:17.000 So for my non-branded prompts, let's take the thing that you and I both know, we fucking crush on. And let's start with those because eventually the AI will catch that we're not really good at this thing, and then we're not going to convert or drive signups or drive add to carts 00:41:17.000 --> 00:41:33.000 Anyway, so if we can't win on something that you and I know we're really good at, how the freak am I gonna win on something that we're at? So let's do brand, let's focus on the areas where we are really good at, and our customers have confirmed we're good at that, not our marketing team, or our brand team says we're good at 00:41:33.000 --> 00:41:38.000 No, our customers said we're good at it, let's focus on that first 00:41:38.000 --> 00:41:39.000 I love it 00:41:39.000 --> 00:41:58.000 All right, I'm going to take the next question. And the question is basically, you know, we get this data, we get these inputs for long tail keywords and short tail keywords, and how important is it to include those keywords verbatim. And my answer is it's becoming increasingly less important 00:41:58.000 --> 00:42:15.000 These models really deeply understand the semantic connection between different ways to phrase things. The proof is in the pudding. You can always test this, right? If you stray too far, and it's like, well, the brand wants to talk about it this way and call it that 00:42:15.000 --> 00:42:35.000 But my audience is calling it that you want to try to lean towards what your audience is referring to as much as possible, at the very least, include both. It's worth testing, but my short answer to that is we are increasingly getting away from the need to include these verbatim keywords 00:42:35.000 --> 00:42:52.000 All right, the next one I'm going to give to you, Will, what are your thoughts about the possibility that users will begin to push back against ads in ChatGPT responses with the thought that they might view the responses as less correct or true and more sponsored 00:42:52.000 --> 00:42:56.000 I think you already have the answer to that. It's called Google. 00:42:56.000 --> 00:43:10.000 And they used to show you where the ads were, they used to not have ads, and then they added ads in, then they had them with a blue background and then like Google has grown a lot, haven't they, since 1988, 1998 when they launched, right? 00:43:10.000 --> 00:43:12.000 So we already have the proof 00:43:12.000 --> 00:43:19.000 Oh, I don't trust ads. Yeah, sure you don't. Like, how did they grow their business as much as they've grown it faster than anything else on the planet 00:43:19.000 --> 00:43:23.000 With a bunch of people saying, I don't trust ads. They click on them anyway. So the proof is already there 00:43:23.000 --> 00:43:39.000 Yep. It reminds me of Kevin Indig's study, I think we covered a month or two ago, where on one hand, people are like, I don't trust AI. And on the other hand, whatever Gemini says, they give them three brands, and people are like, okay, those are the three brands I should consider for this query 00:43:39.000 --> 00:43:40.000 Yep. 00:43:40.000 --> 00:43:43.000 What people say and what people do are very different. 00:43:43.000 --> 00:44:02.000 All right, the next question I have is on measurement. So Will, I'm going to tee this up for you. Chrome Auto Browse went live this week and it's built for what our clients sell appointments, quotes, bookings. There's two problems here. When an agent gathers quotes and the human only ever sees the winner 00:44:02.000 --> 00:44:11.000 Brand, design, mid-funnel content stops mattering for the losers. What do you optimize for when the buyer is a bot? 00:44:11.000 --> 00:44:24.000 All right, so this reminds me a lot of like voice search in the early days. So there was a very specific thing. So actually, it's funny, like what we're doing right now is a microcosm of what we did at Sierra when voice search came out. We got a bunch of people in a room 00:44:24.000 --> 00:44:36.000 And we sat around, it was like 6 of us, and we sat around and recorded and challenged each other, and challenged each other's thinking, right? And what we realized was, while everyone was like, we got to do voice 00:44:36.000 --> 00:44:55.000 We were looking in Google Webmaster Tools at misfiring OKGoogles. So we had data on whether or not people, whether that number was growing. Now, all the haters were like, well, those were only the misfires, and they're this, and they're that. I'm like, well, what data do you have to consult with your clients on this? This is the best data I have. I'm very good at playing with the data I have 00:44:55.000 --> 00:45:10.000 And now crying in my beer about the data I wish I had, right? So we knew, like, this voice search ain't taken. The misfires across all of our clients just were like flat, flat, flat, flat, flat, flat, flat are not growing at the level that people had us being so fearful of 00:45:10.000 --> 00:45:26.000 So what I would say back to you is, I would network with other people who are in my space and figure out what's the data point that would tell us that, like, that's starting to take. Do you see a certain query that triggers it 00:45:26.000 --> 00:45:43.000 And you're like, well, I rank for that query, and I show up in AI for that query pretty well. Am I seeing that when this thing showed up on the day it started showing up, that all of a sudden my funnel started giving me… You're going to get the same conversion, but it's going to show up as a Google ad automated conversion thing 00:45:43.000 --> 00:45:51.000 Or did it not really pan out that way, and people still like visiting websites, right? So I think what you've got to do is think through what's the metric 00:45:51.000 --> 00:45:53.000 Or metrics 00:45:53.000 --> 00:45:56.000 That would help me to steer my company at this time 00:45:56.000 --> 00:46:00.000 How do I find non-competitive other people 00:46:00.000 --> 00:46:14.000 to pool our data together to look at trends to make sure that I'm not some weird outlier? And then how do we share that information together so that we can make more informed decisions? Because at SEER, what we do is we just look across all of our clients instantly, because all their data is stored in BigQuery. But 00:46:14.000 --> 00:46:19.000 But for you, you've got to find a way to bring that group together so all of you are smarter to work together as a group 00:46:19.000 --> 00:46:35.000 That's what I would do, and I know that's hard with legal and where's our data and blah, blah, blah, blah, blah, but you're like the alternative is, I'm working on one website at a very turbulent time and could completely blow the strategy because I don't get to see how other people's data looks 00:46:35.000 --> 00:46:36.000 I love it 00:46:36.000 --> 00:46:37.000 It's a great question 00:46:37.000 --> 00:46:47.000 And there's a second part to that question that I'm going to take. Part two is Agentic sessions look like human sessions in GA4. So our attribution is about to be trash. 00:46:47.000 --> 00:47:05.000 This is a really, really good question slash comment. And I would argue that our attribution in many ways is already trash. This is a difficult problem. We are working on this at SEER in a variety of ways, but something we're close to launching is an Agentic readiness 00:47:05.000 --> 00:47:16.000 audit. Note it's not about Agentic commerce. This is not just about if you're worried or concerned that bots are going to be acting on behalf of people to purchase things 00:47:16.000 --> 00:47:32.000 This future is already here. I think Cloudflare reported recently that more than half of Internet traffic is already happening through bots right now, and it is really difficult. A lot of these bots don't self-identify. And as these models get better and better 00:47:32.000 --> 00:47:49.000 And websites get better at trying to gate and allow for what users or what bots they want to allow. These bots are really good at just masking themselves or identifying as something else. So it is a serious problem. I think the way you tackle it, the way we're trying to tackle it 00:47:49.000 --> 00:48:02.000 Is to look at a variety of metrics, both analytics data and other signals. We're working with John Lovett, Tracy McDonald, some really, really smart folks at SEER are working on solving this problem. 00:48:02.000 --> 00:48:20.000 All right, the next question I have is a really good one. Somebody said, one thing I've been noticing in my own geotests is that some highly capable companies aren't recommended by LLMs, even though they're well established. How do you determine if that's a positioning issue, an entity understanding issue 00:48:20.000 --> 00:48:37.000 or a technical geo issue. I'm going to give my thoughts, and then Will, I want to hear yours too. So this is a very first of all, I love this problem, and this is the kind of thing where it's like, this is really interesting, because I've seen this happen too. Sometimes what you find is that brands early on blocked 00:48:37.000 --> 00:48:53.000 ChatGPT or Claude, they're bots, from accessing their data. You might find that it was a very well-established brand or a long-tenured website. IMDB is a really good example of this, where it's like, this website is 00:48:53.000 --> 00:49:11.000 Built for bot crawlers, but it is a critical part of the internet ecosystem. Google has found a way to get what it needs from IMDB that those kind of considerations don't exist in the AI search world. So to answer the question, the first thing I think of is 00:49:11.000 --> 00:49:21.000 Did somebody something up and prohibit these models from even understanding what the brand is? Is there some kind of blocking happening 00:49:21.000 --> 00:49:31.000 Before I move on to the next things on my checklist, so Will, assuming it's not a technical issue, where do you start troubleshooting? 00:49:31.000 --> 00:49:38.000 Oh man, like this is such a great question and I don't have an answer. So let's 00:49:38.000 --> 00:49:44.000 This makes me want to go do some research. So let's table that one. 00:49:44.000 --> 00:50:00.000 Because it's such a good question about, like, you know, if I am… so I think the only thing is, we all think we're leading our industry. So, the first thing I would do is I'd be like, I would take the top 5 publications in my space, or I would take something like, you know, whatever it is, and be like, are they mentioning me more or less? 00:50:00.000 --> 00:50:18.000 Like, than they used to. Like, am I actually putting stuff in… because we have a hard time admitting that our is , so it's like, are they mentioning me more? It's like, well, I wrote more content. It's like, well, do people want that content more? Yes or no? Like, you know, will they pick up on that? But yeah, that's a really good question, so let's table it, and I'll record a video to you after I do some of my own digging on this and think about it 00:50:18.000 --> 00:50:29.000 And I'll send it back to you, because that's a really good one. Thanks for the… thanks, that was a stumper. Good one. 00:50:29.000 --> 00:50:30.000 Totally. 00:50:30.000 --> 00:50:31.000 That is good. Whoever submitted that, if you want us to dig into the specific brand that you're talking about, shoot it our way and we'll use that. 00:50:31.000 --> 00:50:33.000 Totally 00:50:33.000 --> 00:50:45.000 All right, cool. All right, let me see. Where are we? I want to hit one more because I like to leave everybody with a little breathing room before you go to your next meeting. So let's do 00:50:45.000 --> 00:50:47.000 Let's do this one 00:50:47.000 --> 00:51:02.000 Small business. I don't know how relevant. Okay, I hope there's a small business owner slash marketers in the room. If not, let us know in the feedback in the poll we are going to send you. So one question here 00:51:02.000 --> 00:51:17.000 Related, I guess, to everything we're talking about, what about small businesses, whether they're highly niche or localized, you're low on resources, how do you do this work? And I've got thoughts, then I'm going to hand it to Will to close us out 00:51:17.000 --> 00:51:33.000 I actually think maybe this is toxic positivity, but I really think if you have less resources, you have better constraints than organizations who have more resources. Because if you're the only person doing marketing 00:51:33.000 --> 00:51:50.000 And you have way too much on your plate, yes, that is stressful, but at least you don't have to worry about all the bureaucratic of, well, I want to change this page, but I can't because the person who owns this product line doesn't want to call the thing this thing. Living in this AI forward world 00:51:50.000 --> 00:52:06.000 There's so much you can do to automate, to take advantage of the speed through which we can move. So if you are a small business owner slash marketer and you are the directly responsible individual, it is your say what goes on the website 00:52:06.000 --> 00:52:16.000 I would start to reframe that as a superpower and not a weakness. And so to answer the question, it is all about putting out something that is unique 00:52:16.000 --> 00:52:34.000 What do your experts know that your competitors do not? Let's imagine you don't have all the resources in the world, but you really do firmly believe you are the best option for a specific audience. Why is that? What does that look like in the sales process? How do you communicate that 00:52:34.000 --> 00:52:43.000 to your audience, how do you demonstrate that through your content? You want to be as forthcoming as possible with any kind of proprietary insights, research you've got 00:52:43.000 --> 00:52:59.000 And you're going to be if, you know, sometimes it's like, well, do we want to share all of our proprietary insights? What if our competitors get it? Will mentioned this earlier on. That is a bet you have to take. You have to be risk tolerant in this landscape, in my opinion 00:52:59.000 --> 00:53:10.000 And I think the benefit far outweighs the potential risk of your audience or your competitors seeing, you know, the way that the magic happens internally. 00:53:10.000 --> 00:53:17.000 Those are my thoughts. Will, what is your advice for small business owners and marketers? 00:53:17.000 --> 00:53:21.000 First thing is 00:53:21.000 --> 00:53:31.000 I gave this advice yesterday at a conference that Elisa and I were at. And I told the story of like, there was a time where SEER was really small, just like a lot of people. 00:53:31.000 --> 00:53:35.000 And I was, we were probably, like, 3 people 00:53:35.000 --> 00:53:37.000 And 00:53:37.000 --> 00:53:43.000 And I remember it was like 2008 and the economy was shit. 00:53:43.000 --> 00:53:47.000 And no matter who I talk to, when I wasn't selling anything. 00:53:47.000 --> 00:53:50.000 They were like, but the economy's bad, Will 00:53:50.000 --> 00:54:04.000 And it was interesting to watch how many people would show up to tell you, like, feel better. The economy is bad. It's not you. You don't suck. The economy sucks, and that's why you suck. And then you go, okay, well, what can I do about the economy? Nothing. 00:54:04.000 --> 00:54:14.000 It's like, okay. So then I started looking at myself at the end of every week where I didn't make any sales, and I would look at myself in the mirror, and I would say this exact phrase to myself in the mirror. 00:54:14.000 --> 00:54:18.000 Somebody bought your service this week. They just didn't buy it from you 00:54:18.000 --> 00:54:22.000 It wasn't like nobody was buying search. People were buying search 00:54:22.000 --> 00:54:24.000 I didn't get in front of them. 00:54:24.000 --> 00:54:30.000 When I am the DRI, right, the directly responsible individual, I get to say to myself 00:54:30.000 --> 00:54:45.000 No, somebody bought this thing. They just didn't find me, so let me go try to do a better job of being found. Can I build a webpage? Can I go to a networking event that I don't want to go to? You know, all the things that I could do is I could take some ownership of this 00:54:45.000 --> 00:54:49.000 And to Elisa's point, this is the best freaking time. 00:54:49.000 --> 00:54:52.000 Not the worst because in the old days 00:54:52.000 --> 00:55:08.000 Banana Republic, for instance, I always use this example. They had the marketing brand heft to put low-quality content on the bottom of their webpage about how their jeans were ethical, and they started ranking for it over every other brand 00:55:08.000 --> 00:55:10.000 Even though they don 00:55:10.000 --> 00:55:27.000 They don't do much, if anything, to be an ethical genes manufacturer, okay? So they could, because of the heft of their brand, pay stuff at the bottom of a webpage and beat out every ethical gene manufacturer. That's what big brands used to be able to do 00:55:27.000 --> 00:55:32.000 And one of the brands that they used to beat out was this company called Nudie Jeans. 00:55:32.000 --> 00:55:41.000 If you buy a pair of nudie jeans, and I have a few. You can walk into their stores in any city in the world, and they will repair them for free. 00:55:41.000 --> 00:55:58.000 They fly to all their different providers to make sure that they are not using sweatshop labor and that people are being treated right. So they get on planes and fly team members to partners and vendors to make sure that the people in the place are being treated right. So Nudi's been doing all this stuff 00:55:58.000 --> 00:56:06.000 For 15 years, and they got beat by a big brand who slapped some low-quality content on the bottom of a page 00:56:06.000 --> 00:56:08.000 And guess what's happening in an AI world 00:56:08.000 --> 00:56:15.000 They can't do it. Now, Banana Republic doesn't show up with their little thing because guess what the AI is looking for 00:56:15.000 --> 00:56:16.000 receipts 00:56:16.000 --> 00:56:33.000 And it's like you have no receipts for the thing you said on your site. Whereas a company with a real ethos has the ability to show those receipts and win. So I think more than ever, the small business has a chance to be nimbler, faster, try more stuff than to deal with all the corporate bullshit that me and Elisa have to deal with, right? 00:56:33.000 --> 00:56:38.000 So it's like it's an opportunity, but I think it's only an opportunity if you're also honest with yourself 00:56:38.000 --> 00:56:48.000 I am not the best at this. This other place has got better water ice than I got. Okay, well then get better ingredients or find a niche of people that love your stuff and speak to them. 00:56:48.000 --> 00:56:57.000 Right? So yeah, that's what… that's what I would leave you guys with, and the other thing I'll say is just this personalized search is starting to get… it seems like it's starting to get real 00:56:57.000 --> 00:57:08.000 And I can see it being used a lot more. I'm seeing it being used a lot more on my test Gemini pages and whatnot, so I would say, like, start making sure that you're showing up in people's inboxes 00:57:08.000 --> 00:57:16.000 And that they don't unsubscribe from you very quickly, because these things are starting to affect, especially how Gemini is showing answers. 00:57:16.000 --> 00:57:36.000 That's great. All right, that is the show for today. If you submitted a question and we didn't get to it, we always record our responses for Hannah. She sends those out in the recap, so we'll take care of you there. Take a few minutes before your next meeting, go outside, touch grass, try not to take the time to scroll. 00:57:36.000 --> 00:57:44.000 send a message to somebody that you love, reconnect, and have a great rest of your day, everybody. Thank you so much for joining us. 00:57:44.000 --> 00:57:50.000 Thank you.