[00:00:00] Preston Boling: And that's exactly with that story with the mom that created that lovable-- she wasn't showing up on Google. If someone were to Google it, she was not gonna show up. But because the website was basically AI ready, the large language models were able to find that when a very specific niche question was asked. And so then she had customers lining up before she even knew that she had a business.
Welcome t o the Simple and Smart SEO Show
[00:00:21] Crystal: Welcome to the simple and smart SEO show podcast! I'm your host, Crystal Waddell, here to bridge the gap between SEO strategy and real world business success. By bringing you insights, stories, and conversations from the SEO community and beyond.
[00:00:34] Crystal: Grab a coffee. Or your favorite tea. And let's dive into Smarter SEO for your business.
Welcome and Guest Intro
[00:00:39] Crystal: I'm here with another colleague that I met at SEO Week up in New York City. It was a great week.
[00:00:45] Crystal: It was a great week of learning, but mainly a great week of networking, and I was so grateful to have a chance to interact with Preston Bowling.
[00:00:52] Crystal: He is the founding GTM at Searchable.
[00:00:56] Crystal: I'm so excited that he's here to talk about AI agents and the [00:01:00] agentic web. So Preston, thank you for joining me today on The Simple and Smart SEO Show Podcast.
[00:01:05] Preston Boling: Happy to be here
Why Agents Matter Now
[00:01:06] Crystal: you just started over at Searchable not too long ago, right? And when we talked about this episode, oh my gosh, like SEO just changed. And now you're telling me that SEO changing again.
[00:01:20] Crystal: And we're supposed to prepare for AI agents. So I've just... I just wanna jump right into this because is this a real change that business owners need to prepare for? Or is it mostly industry language right now?
[00:01:35] Preston Boling: It's a really great question. Honestly, every industry that I've kinda seen for the last 15 months is headed towards the same direction.
[00:01:42] Preston Boling: If we look at a SaaS industry perspective, like even like the visitors on your website, like 50% or more of them are going to be agents visiting your website on behalf of humans.
[00:01:54] Preston Boling: And so again, within each industry that's looking a little bit different, like if you look at like the home service [00:02:00] industry, the last time that I checked with like the actual visitors on the website is a lot lower than 50%. But as I mentioned, all of these individual industries are trending in that direction.
[00:02:12] Preston Boling: And so I think it's definitely something to keep in mind as like a business owner or a marketing team. That people are finding new ways to purchase and to discover and to learn. And that is through these agents and large language models.
[00:02:27] Crystal: Okay. And I liked how you put it that SEO needs a makeover. And I'm like, oh my gosh, we're still trying to understand the last makeover. And, hearing that websites need to be agent ready. Is this something that you learned at Google I/O? or where did you learn about this? And where did you kinda come up with the thought that people need to pay attention?
Google I O and Agent Ready SEO
[00:02:48] Preston Boling: Yeah, great question. So yeah, Google I/O was one. At a previous company that I was at, our head of partnerships attended that conference. And it was definitely a huge conference.
[00:02:57] Preston Boling: Google basically saying that "Hey, there are [00:03:00] new visitors on your website. You need to be agent ready."
[00:03:03] Preston Boling: And what a lot of that looks like is traditional SEO. And so SEO is definitely not dead in any capacity. Like SEO is mattering a ton.
[00:03:12] Preston Boling: And I feel like there's a lot of noise in the space around like AI search optimization and SEO and like they wanna be separated.
[00:03:21] Preston Boling: But when it's all comes down to it, it's just like search in general.
[00:03:24] Preston Boling: At least that's like our perspective at Searchable. But yeah, you definitely wanna be like agent ready. And so that is a lot of the technical aspect of things of allowing large language models to access your website. And making sure that it is content that they can read. 'Cause they're not able to read through JavaScript and read media yet and any images.
[00:03:44] Preston Boling: And so you wanna make sure that you have that content on your website that's in accessible form for a large language model to pick up on.
[00:03:52] Crystal: And by that you mean text content. Is that right?
[00:03:54] Preston Boling: Exactly.
[00:03:55] Crystal: Okay.
Bots vs LLM Retrieval
[00:03:56] Crystal: So the big idea here is that our websites have always been [00:04:00] built to help people find information. But now they might also need to help an AI agent complete something.
[00:04:06] Crystal: So what do you think that an AI agent might need from a website that a traditional Googlebot crawler might not need?
[00:04:15] Preston Boling: Not something that I've necessarily thought about before, but if you think about if we break this down with a Google bot. It's indexing the web looking at your individual websites to grow its internal brain around what is available and what's not.
[00:04:29] Preston Boling: And the large language models are also indexing.
[00:04:32] Preston Boling: But also they are live retrieving as well. And so people are asking them individual questions, and that model's are basically visiting the website looking for individual answers. And so I think that's gonna be like the main difference there is looking at the methodology of the way that the large language models work and doing a series of searches when they're given a prompt by a human. And trying to find an answer that's correlated, that is trustworthy. And [00:05:00] has authority to it.
[00:05:01] Preston Boling: And then delivering that and maybe citing or mentioning an individual brand.
[00:05:05] Preston Boling: And that's like that new process that we're, like, scoping out as, as marketers.
AI Visibility and Influence
[00:05:10] Crystal: Do you think that being visible in AI search is the same thing as being ready for an AI agent?
[00:05:16] Preston Boling: I definitely think there's a lot of crossover. Because your website and your content has to be accessible for a large language model.
[00:05:23] Preston Boling: In order for it to even know that you exist. And, but then there's also an additional part of being just having enough influence within these large language models as well.
[00:05:34] Preston Boling: And that's pulling in a bunch of other aspects of marketing and earned media, off page, everything.
[00:05:39] Preston Boling: And there's a lot more to it than just making sure your website is agent-ready. And so it is hard 'cause marketing, it seems like it's more complex 'cause it is getting more complex.
[00:05:49] Crystal: I agree with you, but at the same time, if you approach it from a marketing perspective, it's almost like getting back into your comfort zone because it feels, in a way, less technical than technical SEO and that type [00:06:00] of stuff. Because it's "Oh yeah, we're trying to optimize for a human," or, answer these emotionally charged questions versus just making sure that, like, all the technical elements are there and that we're visible
[00:06:11] Preston Boling: Very fair. Yeah, absolutely. 'Cause on the technical end... and I'm not no expert on the technical side of this. This is just being in the middle of search changing and drinking through a fire hose for the last, 18 months around what we should be doing and shouldn't be doing.
[00:06:28] Preston Boling: I like the aspect of the technical SEO and in a world with like AI agents, I'm starting to see, especially from like a marketing perspective, that a lot of people are using these agents to kinda do some of that dirty work on the technical side.
[00:06:43] Preston Boling: Just to free them up on more of the fun approach of yeah, creating content and learning what are customers searching for and what we need to be doing and where our competitors are winning. And all of that individual stuff.
[00:06:54] Preston Boling: And letting the agents kinda do like the mundane work on the technical side.
[00:06:59] Crystal: [00:07:00] Absolutely. So I remember I had a woman named Shay on the podcast, and she talked about this a couple years ago, about the fact that, look, I don't want to give AI the fun stuff. I wanna give AI the boring stuff.
[00:07:10] Crystal: So I can do the creative fun stuff. That's the whole point here. Okay.
Agents in the Buying Journey
[00:07:14] Crystal: So let's use a normal example. Let's say I'm looking for an accountant or a photographer or a product for senior night, okay? So today, I might search, compare a few options, visit websites, decide what to do. What starts changing when an agent enters that journey? And because I make custom things, this is such an interesting thought exercise to me.
[00:07:35] Preston Boling: Yeah. An agent is gonna be basically trying to help the primary person that it's been given the instructions by, right? And I... So I heard an, like an interesting story the other week where a mom was just wanting to get more tech-savvy with AI.
[00:07:50] Preston Boling: And so she and her son had an idea to vibe code a tool.
[00:07:55] Preston Boling: So they went to Lovable, and they built an automated way to build like [00:08:00] customized kids' stories. And you would upload like a picture of your kid and kinda give some context around like what it was.
[00:08:07] Preston Boling: And it would turn them into like a superhero or a baseball player or whatever it might be, and then you could order that book, basically.
[00:08:14] Preston Boling: And this was just an idea that they wanted to do for fun. And then the next few days, she started getting basically like emails like, "Hey, there's visitors on your website." And she was like I'm not promoting my website at all.
[00:08:26] Preston Boling: I haven't done any like technical like SEO or anything around this." And it turns out like the direct traffic was coming from ChatGPT, and people were looking for customized children books.
[00:08:37] Preston Boling: And I thought that was fascinating 'cause it's like only could that have happened in like agentic search, ultimately. And so a lot of that you do have to give credit to Lovable and the way that they're set up is very AI friendly. '
[00:08:49] Preston Boling: But to your point, what would you call like those projects that you work on?
[00:08:53] Crystal: Keepsakes or gifts.
[00:08:55] Preston Boling: yeah.
[00:08:55] Preston Boling: So like those custom keepsakes, ultimately, like if people are typing [00:09:00] that and looking for that if it's known on your website like that's what you are doing, and maybe you will have additional sources that are either validating that.
[00:09:09] Preston Boling: Just because when an agent is coming to your website or it's doing those individual searches.
[00:09:14] Preston Boling: If someone types in "Hey, I want to create a fun little gift or keepsake for my child," and from like this, they're a senior and they're playing baseball.
[00:09:22] Preston Boling: It's gonna do a series of fan-out queries to look at everything that it, A, knows about you.
[00:09:27] Preston Boling: But then also like what it should be considering for the answer. And so we're calling those like the sources.
[00:09:32] Preston Boling: So it's gonna be looking at a number of different sources, and then it will determine what it wants to cite and mention within the response.
[00:09:39] Preston Boling: And so of those individual sources, that's where you want to start to see like your brand showing up there.
[00:09:44] Preston Boling: And so that's like the behind-the-scenes of, like how I'm looking at that.
[00:09:48] Crystal: That was a great answer because I was immediately gonna follow up and be like, "What are the sources?" So thank you for sharing that.
[00:09:54] Crystal: So what then happens if a customer doesn't personally visit 10 websites? Their [00:10:00] agent does the research and brings back two recommendations? Or makes a decision for them?
[00:10:05] Crystal: How does that transfer work? I
Personalization and Attribution
[00:10:06] Crystal: I'm thinking about the way that I research. Say, I'm trying to buy shoes for my son.
[00:10:10] Crystal: If I know the exact pair of shoes, I'm definitely looking across a variety of websites to see how I can get the best deal. Like, how is an agent gonna know that? Is that just part of the prompt, or, how do you think personalized agents are gonna work?
[00:10:22] Preston Boling: Yeah. The fun thing about like the personalized aspect of this is as you interact with ChatGPT or Claude, it takes a memory of like your interactions.
[00:10:32] Preston Boling: And so it will remember of, Crystal, you're looking for running shoes, and maybe earlier on in the year you mentioned like, "Oh, I have a high arch," or, "I have flat feet."
[00:10:42] Preston Boling: And so it will take that into consideration, and so that way when you ask that individual question it will basically probably do a fan-out query of like you're saying, "What's the best running shoe for me?"
[00:10:53] Preston Boling: And it's gonna do like best running shoes for high arches or for flat feet, best shoes for in North Carolina.
[00:10:59] Preston Boling: Best [00:11:00] shoes for if you gave it in any other additional context in the prompt, like maybe it knows that you're training for a race coming up in October or whatever it might be.
[00:11:10] Preston Boling: And so it's gonna take all those personalizations into consideration. And then find additional sources. And so that's a hard part for a marketer is because in a world where everything is becoming hyper-personalized, it's how do you even mimic that? Or track that? And a big part of AI search, it comes with a huge headache of attribution.
[00:11:31] Preston Boling: And it's just y- there's not a, a very successful way to actually attribute where they're coming from.
[00:11:37] Preston Boling: Opposed to just like self-attribution and asking directly honestly.
Content Strategy for Custom Orders
[00:11:41] Crystal: So from the business side, how can I make my website... And I'm probably jumping ahead of myself a little bit here, but I, because things I make are so custom, they need somebody to sign off on the decision or to make the decision.
[00:11:54] Crystal: And let's just say like for senior night keepsake, somebody wants a poster. One of the [00:12:00] decisions people have to make is like, what is the number that you want?
[00:12:03] Preston Boling: Mmm hmm.
[00:12:04] Crystal: Another decision is like, what color do you want the text to be? What color do you want the background of the poster to be?
[00:12:09] Crystal: And then the biggest thing is uploading images. So that I have the images I need to make the poster. And I'm wondering like how can I make my website friendly to an agent looking for a gift.
[00:12:24] Crystal: But then also, com- have that communication back and forth with the client because I know that they're gonna have to sign off on or upload those images.
[00:12:33] Crystal: And that's maybe not something they know that they're gonna be getting themselves into when they start that search
[00:12:39] Preston Boling: Yeah. I think that's a really great question. So if you're thinking about, like, how do I get more customers to do this when there is that additional layer that they need to give that additional information? Ultimately, I'd be just like creating content around, like thinking around, like what questions people are asking.
[00:12:55] Preston Boling: And so if keepsakes and gifts for seniors is like [00:13:00] a big part of the business.
[00:13:01] Preston Boling: That would be its own content piece that I would be pushing out and writing about either previous customers that have purchased from you.
[00:13:09] Preston Boling: And like the different-- all the different areas that they've touched on when it's oh, this was maybe like someone was buying it for their grandson. Or parents buying it for their buying it for their children, or maybe it is a girlfriend or a boyfriend buying it for their significant other.
[00:13:25] Preston Boling: And so there's tons of ways to think about it, but I would just look at like your actual customers and what they are doing. And like most likely, A, like what are they searching for when they're finding your business and how they found you.
[00:13:36] Preston Boling: That might be just asking directly and then creating content around that.
[00:13:40] Preston Boling: Because ultimately then as people start to gravitate towards or like gravitate away from Google and going into either AI mode or asking Gemini or Claude or ChatGPT and "Hey, I have this idea. What should I do? How can I do this?" Google or whatever will be able to find that content and then recommend "Oh, Crystal does really [00:14:00] great work.
[00:14:00] Preston Boling: Here's hers. You'll have to submit... Like it's more customized.
[00:14:03] Preston Boling: Visit this website," and then you should start to see, more direct traffic to that individual site. Which I think in a world we'll see that like agents get a little bit more intelligent and a little more self-serve around I can give my Claude instructions around, "Hey, my daughter wears number three.
[00:14:19] Preston Boling: Here's all these individual pictures of her playing soccer. I want some kind of keepsake or gift that I can give her 'cause she just finished her first year of soccer. Go out and buy something, and my budget is $50."
[00:14:32] Preston Boling: And so I definitely see that, that world coming to that, and that's what these, frontier models are definitely building towards.
[00:14:39] Preston Boling: And so I think that was something that we'll probably start to see in the next few years. But as of right now, like where we are, is it's still having that human loop at the very end of it. And so there's just a middle part of that discovery factor that the large language models are handling.
[00:14:53] Preston Boling: And then that's where we need to focus so we're actually discoverable.
LLMs txt and Agent Instructions
[00:14:56] Crystal: Okay, so it gives me two follow-up questions. The first one [00:15:00] is regarding the LLMs.txt file.
[00:15:03] Crystal: Because I've heard conversations about whether or not it's useful .Or it's not. So I'll go ahead and throw that question to you first, and then I'll ask my next question
[00:15:11] Preston Boling: Yeah. The robots.txt file, to my knowledge, it actually even just will allow like the Google bot. I might be wrong about that. But I would say it is absolutely necessary just because you'll see if you are allowing ChatGPT or Claude or other large language models to access the content on your website, so it can be a blocker.
[00:15:29] Preston Boling: And we have seen that websites that are not having that have a poor audit score around accessibility on their website. And then from previous history of working with either agencies or their customers, when they put that on and make that individual change, we are seeing more agentic visitors on their website and then seeing more referral traffic from those agents as well.
[00:15:53] Crystal: Are you talking about robots.txt or the LLMs.txt?
[00:15:57] Preston Boling: Did you ask LLMs.txt? Oh, [00:16:00] so sorry.
[00:16:00] Crystal: That's okay.
[00:16:00] Crystal: That's still helpful. But yeah, I was more interested in like the LLMs.txt because I was wondering if that could be a part of giving agentic instructions.
[00:16:12] Preston Boling: I would say honestly, I've heard that it does not matter. The feedback that I've heard from a lot of like people that I trust is that it's just kind of noise. It's not showing any difference at all. So it's not anything that I would like rush to put on your website or hurry about.
[00:16:25] Preston Boling: And so we might see that change. But as of today, I have not seen that make a big improvement at all.
[00:16:32] Crystal: Because from a business perspective, and I'm thinking about what can I do to make it easier for my customer to use an agent to purchase from me? That's like what, how my brain is working right now...
[00:16:41] Preston Boling: Mmhmm.
[00:16:42] Crystal: I wonder if there's a way that I could communicate with an agent to say.
[00:16:48] Crystal: If you want to purchase from this page on behalf of the human that you're representing, is there a way to set up a file, almost like a Google Form type of doc, but just like [00:17:00] a reference for the agent to know, these are the pieces of information that I need you to have in order to make this purchase. So it's not okay, what color would you like for your lettering?
[00:17:12] Crystal: And then have to go back and wait for a response. 'Cause I
[00:17:14] Preston Boling: Yeah
[00:17:15] Crystal: when Claude is waiting for a response from me, sometimes I don't even notice. Even though it
[00:17:18] Preston Boling: Yep.
[00:17:19] Crystal: me a notification on my phone.
[00:17:20] Preston Boling: Yeah
[00:17:20] Crystal: It said I'm working on something else, so I don't want to hold up the purchase process. And get in my own way causing friction. But I'm just like, I wonder if that's a possibility? If there's a way that we can set up communication within our site that an agent can read.
[00:17:34] Preston Boling: Yeah.
Enterprise Agent Experience Platforms
[00:17:35] Preston Boling: It is a very tricky question. At my-- The last company that I worked for was a company called Scrunch, and they were trying to solve for this around, they called it the agent experience platform.
[00:17:45] Preston Boling: And basically trying to serve up an optimized version of a website specifically for large language models.
[00:17:52] Preston Boling: And this is something that's only been tested at a very high enterprise level. And so the methodology behind it is very [00:18:00] fascinating. And and ultimately I think we're still gonna have to see over the next couple years if that's an approach that a lot of companies are gonna be able to start to take, especially when you start to look at agencies and other, like, small business owners.
[00:18:14] Preston Boling: It's not gonna be something that's necessarily probably going to be financially available just based off of the actual groundwork that it takes to upload that.
[00:18:23] Preston Boling: Until that is found out or find a way to kinda streamline that. So ultimately, what I would be focusing on is creating more just helpful content around FAQs.
[00:18:33] Preston Boling: Like for your example, Crystal, of anyone actually searching for those individual gifts. Having content around those FAQs and that it's easily accessible and having the schema markup on the website too, that large language models can can scrape directly from that. And pull that.
[00:18:47] Preston Boling: Is like a miniature version of that, so they're able to actually access that information.
[00:18:52] Preston Boling: And then that way, when people ask those individual questions, you are showing up in those individual channels. And so it kinda takes a [00:19:00] big leg out of the technical side of that. The agent experience platform does. But as I mentioned that's only been built around enterprises today.
[00:19:08] Preston Boling: And so a lot of for everyone else, ultimately it is just having good content.
[00:19:12] Preston Boling: Making sure that, yeah, we have the robots.txt file. We're allowing those large language models to access our website.
[00:19:17] Preston Boling: And we're thinking about the agents and the way that when they come to our website, they're looking for an answer.
[00:19:23] Preston Boling: And our content needs to mimic that. And give them an answer. 'Cause if they're not able to get an answer, they're gonna go somewhere else.
[00:19:29] Preston Boling: And that's why, we see a ton of third-party citations coming from Reddit and Quora and different forums. And that's just a part of that individual process from the large language model.
[00:19:39] Crystal: Think my neurodivergent, tendencies are kicking in right now because I've, I'm still going back to my website thinking, "Okay, how does what Preston just said apply to my Shopify store?"
[00:19:51] Crystal: And one of the things you said earlier was, like, an agent might look for something under $50 or whatever.
[00:19:57] Crystal: I happen to sell very [00:20:00] high-end keepsakes. So let's just say, most of my products are priced from 120 to six or $900.
[00:20:07] Preston Boling: Mmhmm.
[00:20:09] Crystal: Thought to myself, I was like, "One thing I could take away right away is that maybe I need to create some collections that match up with those parameters.
[00:20:17] Crystal: Of how people search. Of under 500, under 100, under 200." Do you think that's on the right thought pattern or thought path?
[00:20:25] Preston Boling: Absolutely.
Searchable Method and Prompt Testing
[00:20:26] Preston Boling: 'Cause what we're doing at Searchable to make this easier for folks is we are looking at ba-basically like connecting to your Google Search Console. So we're tracking like the traditional search and the traditional visitors on your website.
[00:20:40] Preston Boling: And then we're mixing in something called panel data, which is basically like clickstream data, and it's validating what people are clicking, the user behavior.
[00:20:49] Preston Boling: And then we're also bringing into factor the query fan-outs from the large language models. And based off of like those three points, there's a few more, but based off of mainly those three points, we're [00:21:00] gonna help you basically create prompts.
[00:21:02] Preston Boling: That are trying to mimic what the large language models are gonna be searching for.
[00:21:06] Preston Boling: And then from there, we'll run them against the large language models that you care most about.
[00:21:10] Preston Boling: And that way we can start to see, okay, once we run them against the large language model, and it's from a logged out state of the model as well.
[00:21:17] Preston Boling: So there's no personalization taking a factor here. We're looking at just the raw model itself.
[00:21:23] Preston Boling: And from there we can start to see, okay, when these individual questions are asked around keepsakes or gifts, here is what the models gravitate towards.
[00:21:31] Preston Boling: Here's what they're picking up on. And then if we look at like the content types, maybe that's where they're focusing on comparison pages. Or, maybe it is Reddit forums or LinkedIn posts. Or Facebook or just other third-party citations.
[00:21:48] Preston Boling: And so we can start to see basically from those individual sources that it's trusting within that initial search. What's driving the most influence there?
[00:21:57] Preston Boling: To help you then say, "Okay, [00:22:00] if it is comparison pages, then this is what I should go out and create.
[00:22:03] Preston Boling: And then that way I'm appropriately showing up where they're searching, and then that way my brand can start to show up in those individual responses."
[00:22:12] Preston Boling: And then you can start to identify any of those gaps. If people are searching for, yeah, a keepsake for $50, then maybe that's a new product that you could create.
[00:22:19] Preston Boling: That's like a miniature version of it, or it gets them in the door and then they want, the, the big number three that they can hang on like their child's wall.
[00:22:27] Preston Boling: And so there's a lot that kinda goes into that, but that's like the methodology behind what we're actively doing at Searchable to help make that process a little bit easier. And maybe help answer those, those questions.
[00:22:38] Crystal: Yeah. It's so neat to hear the examples and that, and to hear that there's a tool that kinda helps you see the SERP for what it is and what it's what it's giving you back for certain searches.
[00:22:48] Crystal: Because for one of my clients I run her blog.
[00:22:51] Crystal: And As I do research for every article that we're creating, I go to the SERP and I manually look at it.
[00:22:57] Crystal: And I look to see which [00:23:00] SERP element is showing up for that particular keyword, you know, and variations. And so I'm like, "Okay. We might have a chance here if there's some Reddit threads that are showing up or some outdated information here."
[00:23:11] Crystal: Just looking for where to position, her brand, her story, and the thing that she sells.
[00:23:17] Crystal: It's so interesting that there's so many different places now where you can show up in search, and on that SERP page you might show up, like you said, in a Reddit thread or a mention somewhere else, and it's not even taken into account like the traditional 10 blue links.
[00:23:33] Preston Boling: Not at all
[00:23:34] Crystal: But you could just insert yourself in the conversation
[00:23:36] Preston Boling: Yeah. And that's exactly with that story with the mom that created that lovable-- that's how she wasn't showing up on Google. If someone were to Google it, she was not gonna show up.
[00:23:46] Preston Boling: But because the website was basically AI ready, the large language models were able to find that when a very specific niche question was asked.
[00:23:54] Preston Boling: And so then she had customers lining up before she even knew that she had a business. And so yeah, it is very [00:24:00] helpful. And on your example with like the blog one piece of like feedback is when you are looking at those individual SERPs, I would try to do it from like an incognito or a logged out state.
[00:24:11] Preston Boling: So that you're getting just the raw methodology behind it, that there's no personalization coming into play there that might be able to skew any of those individual results.
[00:24:19] Crystal: Yeah, that's a really good point because even just yesterday when I was working on this, I was not incognito. But I did the search twice and the SERP changed completely.
[00:24:29] Crystal: I would say 8 out of 10 of the links were actually taken up by the main brand.
[00:24:33] Crystal: But what I was looking for was videos and, different SERP elements that we could target instead.
[00:24:38] Crystal: And the videos didn't even show up in the second search. Which I thought was really interesting. I was like, they were just completely gone.
[00:24:44] Crystal: So I thought,
[00:24:44] Preston Boling: Yeah.
[00:24:45] Crystal: "Huh. Yeah."
[00:24:46] Crystal Waddell: Next week Preston and I get practical: YouTube, structured data, and how to update your site without losing the rankings you already have. Thanks for joining us today and we'll see you next week for part 2!