Keywordless Targeting, IndexNow, What SEOs Should Learn From PPCs w/Navah Hopkins of Microsoft

Speaker A: All right, everyone, welcome back to Experts in the Wire. I'm super excited to chat with Nava Hopkins about PPC and SEO and everything happening in AI search. I want to jump right in. What is keywordless targeting and why should SEOs know about this?

Speaker B: So this is, I think, the best jump right in to a question, uh, that we could talk about. So keyword list targeting, as most folks understand it, uh, is using anything other than a keyword to match the human to the brand. Now, that might be matching someone to a product, that might be matching someone to a, uh, service, whatever the case might be. Why I say how most people understand it is that the surfaces that are coming about in the AI era are evolving really quickly. Um, and how we think about even the idea of keywordless targeting is going to evolve really quickly. So as it stands today, it revolves quite a bit on understanding the landing page, understands quite a bit on understanding the human and their previous behaviors. It is very clued in, um, to also the goals that you may have set. So why this matters for SEOs to be aware of as a PPC sort of concept or sca, if you're in emea, um, the website and how easy it is to understand that website and how well it's matching is a really good signal to understand. Does Google, does Bing, does fill in the blank AI service like do. Are they able to understand you? So by plugging in your site into, say, a pmax campaign creator, you can get a really good sense of how easy is it for Google, for Microsoft to understand me and my brand and potentially even for competitive intelligence to plug your competitor brand in there to understand how is your competitor, uh, understood and are there opportunities there to improve even more on your own gains?

Speaker A: I love that. And that's one thing I want to get to, because I was working with a company and we had a meeting where I brought in the PPC person running all their advertising and they were like, oh, we got our two biggest leads last week from ads. What type of campaigns did they come from? And he said pmax, and that was the first time I ever heard that. And he explained that apparently pmax is a way of dynamically sort of going out and serving ads to somebody that's maybe already interacted with the brand. But his comment was that the quality of our SEO and organic content was helping to support that. So can you unpack the keywordless targeting a little bit more and maybe even the pmax a little bit more? Is keywordless targeting just like a concept or Is it actually a type of ad that you can run? So, question number one, and then number two, I'd love to talk a little bit more about how the website can support better performance of that advertising.

Speaker B: Sure. So, taking a step way back when a paid media person builds a campaign, historically they would have built a campaign using keywords, using location targeting, using scheduling, using audience targeting, you name it. Like, basically, who do you want to target, when do you want to target them, um, where do you want to target them? And then keywords were there too. Keywordless targeting is a concept, it's not an ad type. Um, it's a setting idea that instead of relying just on the semantic keywords and the syntax of the keywords to get that keyword to add landing page relevancy to match exactly to what someone is searching, instead you think, okay, these are the general ideas that someone who's useful is going to search. These are the ideas that I want to be in the sphere of influence for, but I also want to make sure that I'm available to come up when someone searches for me. And it's maybe not something I thought of. That's where the landing page is really critical. Because if your landing page easily describes your services and it's an easy page to index and to understand, you'll be in great shape. Where I think not enough people were prepared for the AI era is that because pmax and aimax and AI tools need to look at landing pages to understand the service. And for a really long time, paid media folks would have their landing pages as no index, nofollow. That would mean that you're basically depriving a really key, um, insight to that AI campaign type. So I would caution people on leaning too heavily on habits that served really well in the past. I mean, I was the first person to be in the make your paid landing pages on a no index, nofollow subdomain and like, keep it totally separate as brand and performance are converging quite a bit, um, as journeys are really shrinking, the value of having those standalone paid landing pages not only is shrinking from a strategic level, but it also could get in your way technically. So for the SEOs out there who are looking for additional insights of like, how to think about this, how to position it, you now have extra ammo in your arsenal to get more data sharing with the paid team. And even though they're not going to have full visibility with all platforms, with Microsoft, you will have that full visibility into search terms so long as there was a click not all platforms offer that. Um, but even though you don't have that full visibility for every platform, you at least can still begin to, um, get those learnings of what works well and what doesn't for, uh, PA traffic or organic. Whether you're using free tools like Microsoft Clarity to understand that, whether you're starting to lean into AI reporting and in the AI citations to understand what are the AI systems on the back end asking that your site is being found. We're now getting a little bit away from keywordless targeting. But like these are the sorts of considerations that, beyond just keywordless targeting for relevancy that like your landing page is easy to understand. Like there's other bigger questions that making sure that your landing page is really easy to understand, really great user experience, super high trust signals, like all of that's, that's in order.

Speaker A: Yeah. And I want to, I want to come back a little bit more to the landing pages and the web content. But first you mentioned Microsoft. So we should tell people who you're currently with and what your role is at Microsoft and also a little bit of your background because I know you've had mostly an experience in ppc. Uh, uh, you definitely have a lot of knowledge in SEO as well. So I'd love to understand just sort of your expertise, background, where you kind of came from and then what you're doing now at Microsoft. And then I want to continue talking a little bit about landing pages.

Speaker B: Sounds good. Um, so the biggest thing to think about when it comes to me is I am motivated by helping people. I wanted to be an English teacher when I was growing up. That was going to be what I was going to do. And I realized I would be both poor and ineffectual going to the education system. So I switched to marketing and I had the privilege of getting to work in the SEO industry to begin with. The problem was I got in right at the cusp of Panda Penguin and felt, I don't know how I feel about this. So I switched to, um, paid, uh, and it's been really, really great. And I've had the opportunity to work in software companies, have had the opportunity to work in house, uh, working agencies, um, but software has always been my happy place because you can, uh, have one conversation and help many, getting to the point of kind of this, this journey that we find ourselves in of like, AI versus like the world. Back in 2008 when I got into this, I think that there's a lot more value in strategic, creative, critical thinking that not Enough people today maybe get access to. So, for example, um, one of the things I would do with new hires is I'd get them to play board games with me. Because like you, you train someone to manage resources, to learn a set of rules and have to adapt to them really quickly. That's a really useful mental exercise for the brain. Um, when you think about channels today in digital marketing, like, it's effectively the same thing, except money's involved. So if you can get someone used to thinking about that, um, and if you get yourself used to thinking about that, uh, without having to risk money, it's kind of helpful. I don't know if I answered your

Speaker A: question, but that does set up some of your story and background and then take us to currently your role at Microsoft. Uh, just explain that a little bit because I think that relates a lot to what you're talking about and can also support like, you know, why should we listen to Nava? Like, you have so much expertise in your current role at Microsoft. I think makes you a really great person to speak to my audience, which is mostly SEOs, but I think a lot of people just in search and trying to make sense of AI search. So I'd love to hear about your, Your role.

Speaker B: So my, my role is product liaison for Microsoft Advertising, Microsoft AI. Um, I actually sit within the product organization, not the marketing organization. And what's so special about that is that it means that I have a seat directly in products and product marketing to ensure that voice of customer reaches folks. So when there's something that we're doing well, they get to know. When there's something that we could improve, they get to know that too. Um, but even more importantly, when there are things that are clearly myths or there are things that are clearly getting in people's way, I can help translate it. Or if we're launching something, we can, um, have that conversation. So my role is a mixture of what I like to call the loud work of getting up on stage, sharing things, having conversations with wonderful forces for good, like you and your audience, um, writing my monthly columns, um, for, uh, Search Engine Journal at Search Engine Land, working on the board of the PACERS association, um, doing kind of community outreach, like all those various things, social listening for Microsoft. And then there's the quiet work, um, of actually being in the room with product folks, advocating for what will serve customers best, being that voice of the customer, uh, and then also helping to think through how will this land, um, and really putting that hat on for agency, putting that hat on for in House, whether it's the practitioner, whether it's the team lead. So I get to do a lot of really cool fun things. Um, but it's all motivated by helping people.

Speaker A: That sounds fun. Is there a big myth that comes to mind that you've had to. If you're referring to a myth out there in the world that you're bringing back into the product team at Microsoft, is there a big one that comes to mind?

Speaker B: So one of my, there's two. Um, the first is that Microsoft is just for old people. Um, it's really funny. Uh, we have a beautiful population that ranges in ages. Uh, we have a wonderful audience that um, tends to be more affluent, tends to uh, have a higher education. Um, but what I don't think most people realize is that I think it's either 71 or 73% of Microsoft uh, users are 45 or younger.

Speaker A: Oh wow. Uh, I did not know that.

Speaker B: It's just one of those myths that for whatever reason we've been pigeonholed as this is where you go to reach the old people and it's ah, you can but there's definitely more people there.

Speaker A: When you say Microsoft user, is that of any Microsoft product or are you speaking about Bing?

Speaker B: What's who you can reach through Microsoft advertising? Bing has ah, over a billion monthly users. Um, people think about Microsoft as being the super teeny tiny, not terribly helpful and that's just not true. Um, the other thing, and that leads to the second thing uh, is that um, import is the easiest way to get started on Microsoft because there's so many incredibly useful Microsoft specific tools. Well Microsoft uh, import can be helpful and we have a Google import, uh, Facebook or Meta import, um, and a Pinterest import. It actually can be a lot easier to start from scratch because of the rules of engagement that are subtly different. So for example ad group level location targeting and ad scheduling, the fact that ads can serve in the time zone of the user or the account. Um, our impression based remarketing where you can actually build targeting exclusion or bidding lists off of just an impression. Um, our LinkedIn profile targeting. There's a number of things that are just really, really interesting and one thing that was actually kind of funny um, during uh, GML this year Google announced a pilot for disclaimers. Meanwhile we've had disclaimers since 2021. Uh, Google is unveiling the ad serving and AI overviews and AI mode. Like we've had ad serving in Copilot since 2024. So just helping to shatter those Myths that like people, like what are people not quite seeing or hearing.

Speaker A: Right. So when you're talking about import, you're talking about a mechanism to start advertising.

Speaker B: Yes.

Speaker A: Using literally a feature to import data or your information or. Yeah, interesting. Um, so that the background's great and I want to come back to you, just a couple things about the landing pages and the PBAX campaigns. Uh, is there still a case for having a, uh, noindex nofollow no crawl landing page for advertising? Is there still place for that?

Speaker B: Sorry? Um, uh, is there a place for

Speaker A: landing pages for noindex nofollow landing pages? And I'm impressed with the double tasking if people can hear the typing.

Speaker B: So the noindex nofollow landing pages is really interesting. So long as there are uh, humans who will never use, um, AI anything. There probably is.

Speaker A: Okay.

Speaker B: Where I'm concerned, candidly, is that we're going to a place where landing pages and websites might not be the main mechanism that people are accessing content. So if we're getting kind of held up on this idea of no index, no follow, um, and like we have to keep traffic separate for paid and organic, we're gonna lose the forest through the trees a little bit. Um, so it's not that there's no place, but I would encourage people to be cautious about clinging to that as a best practice as AI surfaces become that much more important, as AI in campaigns become that much more important. Because ultimately going back to that keywordless targeting, landing pages are a critical part of that. And if you tell the system that it can't do anything, um, specifically with the robot txt, um, your landing page, it's like taking a Ferrari and driving it through the mud, taking a four wheeler and asking it to do like a, a NASCAR race or something or something like that.

Speaker A: I love the analogy. Um, now when you spoke about the quality of the landing page and my brain went to an assumption and I got excited. So I just want to make sure that I'm not over assuming something. I like to think that the work that I do, let's say, uh, in blog articles or in resource pages or in content marketing on a website is part of helping the advertising side of things understand what this company is all about, what the topics are all about, what their products and services are all about, who their customers are, is on the ad side. Are they crawling and evaluating the entire website to get a better idea about how to target and run the PMAX ads? Am I thinking of that correctly?

Speaker B: Kind of uh, so if you allow for what's called file URL expansion, it will look, uh, at the whole site and it will find those new pages to target. The thing is, you may not want it to do that. If you have ads going to the blog, for example, or to a sitemap, you probably don't want that to happen. I see.

Speaker A: To allow them to.

Speaker B: Can I just finish the thought and then you can chime in? So in order to prevent that, there are things that you can do called URL exclusions. Yeah, you can say, I don't want any URL from the blog, or I don't want the sitemap, or I don't want the homepage included so that you're then able to get that benefit.

Speaker A: Interesting. So if you include like a blog article, then you're opening it up for them to run an ad to that blog article?

Speaker B: Yeah, if you make it eligible to be pulled, yes. Uh, and especially if you outright target it. What I'm getting at is the accidental inclusion of blogs or pick your poison on the site. If you turn on final URL, um, expansion and you don't include rules which are the equivalent of dynamic search ads of like, whatever rules you want to, um, set for what, uh, URLs can't be included, then, sure, like, it's possible. But, like, most people don't want to send paid traffic to blog pages. Most people want to send traffic to service pages, to collection pages, to product pages. The reason to include a blog might be as an extension because you're giving someone more information. Um, it also can be helpful, like for your own knowledge, to plug the blog in and see how does the PMAX campaign generator process the blog. Like, are you being understood in the way that you intend to? So I would really encourage every SEO to actually partner with their paid friend and just borrow their, um, access to their account. Not run any ads, but, like, just see how does the content translate? Do you agree with the creative that comes up? Do you agree with the positioning that comes up? And if the answer is no, like, maybe spend some time there. Um, and then on the paid side, the reason to send traffic to any part of your business is because you want that business. So, like, it's, it's really worth having those conversations. How well does all does the whole part of your site do at helping people understand why you're, why you're helpful?

Speaker A: And that dovetails. Perfect. To what I was going to ask, which is you mentioned on a podcast interview that I listened to that or article you wrote that I read that we should be gathering intelligence for SEO from pmax campaigns. And so you mentioned one which is I. And see this is one reason I wanted to talk to you because I think more SEOs and PPC should talk to one another. I have very little knowledge of what you're talking about when it comes to a pmax campaign or setting it up and looking at the data you get.

Speaker B: Let's take that step back. So a pmax campaign is an aggregate of all possible ad types. So on the Google side that means search, that means video, that means local, uh, that means Gmail. It's like the lot on Microsoft it means search, it means display and native and video through our audience ads, uh, not video actually on Microsoft we specifically don't a force video whereas Google they uh do. There was a little um, misstatement. Um, there's also um, what's it called uh, on the search side uh we have something called multimedia ads that also serve on the right hand rail or now in the center of search result page. So there's a number of different ad types. The premise of pmax is, is that humans are very clever. Humans have really good strategy but humans also have biases. And so a pmax campaign is really good at allocating your budget based off of your goals, based off of whatever creative. Um, and then if that is all correct then that buy is free piece comes in and can get you those incremental conversions. If however you're still in the process of figuring out who your ideal customer is, like what sort of creative like needs to match or if you need to have very official creative, you might lock into a standalone search campaign or standalone display. Uh campaign now on the Google site is just um, a mangen on the Microsoft side is audience ads, um, things like that. So when we're talking about PMAX and its relation for SEO, PMAX's whole like the whole hypothesis of PMAX is what would it look like if the best parts of SEO, which is empathy for the human technically set up correct experiences, feeds, clean hygiene and feeds came together and you took human oriented strategy like guardrails for brands and said go. Yeah, most people are pretty happy with the results even if they have questions about how the results happened. Um, but like that's, that's the general premise of pmax.

Speaker A: So what do I, let's say I'm working with a, with a company and there's a PPC team, there's an ads team, uh, how do I go to them or what would be the mechanisms in which I would actually get some of this intelligence and data from PMAX campaigns that are, do I go to them and say give me access to your ads account and then if so what do I look for? Uh, do I just ask them for a report? What kinds of things should I be looking for in those campaigns for insight on how good like my pages are and the SEO is

Speaker B: how good the SEO is.

Speaker A: So the question is, what do I need to do to make use of this intelligence that's in pmax? You know, do I go to the ads team? Do I get access to the account? If I do what do I go in there and look for? Uh, do I ask for a report? What sorts of intelligence should I be looking for?

Speaker B: Got it. So there's a couple of reports. Uh, the first is make sure that your Google, uh, account has Search Console linked into, um, the Google Ads account. Because then you can get um, SEO versus paid ratio, uh, of impressions to clicks, like how often uh, the SEO term would have come up versus paid. Um, this particularly helpful in branded campaigns. Um, I also would really recommend looking at um, search terms that are available, um, or search term insights in the case of pmax. Um, the reason why that's really helpful is that you can then see what actual humans searched. The thing is, and this is important, as AI takes over more and more of those conversations, the ability to get the full, full query is going to be really, really hard.

Speaker A: Um, and you mentioned, oh God, connecting Search Console in Google Ads on a thing I listen to and I said why have I never heard of that before? And uh, maybe I'm just been under a rock this whole time, but talk about that a little bit more because I, uh, to me that feels like an SEO gold mine.

Speaker B: It is an SEO goldmine. The main thing, uh, or the main reason why it's not talked about is it's a hidden report in custom reports and you have to know to connect Search Console. And historically it's been really hard to build those relationships between paid folks and organic folks that like you get that linking unless it's either paid in organic in the same agency or paid in organic teams in house. But like for whatever reason we as humans have had like this weird relationship. So the first thing is just like making that connection and having that um, in service of the brand mindset. The other thing just to be aware of is that there are tools in both Google and Microsoft ads. Um, in Microsoft it's copilot, um, in Google it's going to be either Ask Advisor, uh, Or uh, Gemini, where you can ask what's available, what reporting is available, how can I get this done? I don't think enough people actually ask the platform enough or with the AI assistance enough what's available. Because even if it gets it wrong, which it can, because AI will get things wrong, you still at least have something to go back to for support and whatnot. Say like, hey, this is where this came from. And then they'll be able to not only help you, um, with your individual question, but they'll also be able to say, hey, how did this happen? What's the root cause documentation that may have informed this?

Speaker A: That's a great tip. I very often don't think to just go ask the tool nowadays in the age of AI, you know, what can I do with your platform? Uh, so you're saying when we run a PMAX campaign, it will give you information about what it thinks about your page, such as the audience or like,

Speaker B: um, it'll give you um, potential creative. Um, so whether it's display creative, whether it's text creative, um, and depending on how useful or not useful it is, that's a really good sign of whether you're understood. So for example, I plugged in um, a lawn care company and I plugged in just their homepage and their homepage actually had accurate um, images on it that were talking about season specific lawn care more. Um, it was like for flood prevention and things like that. And the only images that got pulled and the only services that got pulled were snow removal. And it was just fascinating because it's like those images, those pages, like that content was just so much easier for the system to understand that it pulled that instead of what was actually like right there. So it's just a good reminder, make sure that your alt, uh, text is there for not only images but for videos. Um, JavaScript is gonna be a little bit harder AI systems to process and understand. So like just be mindful. Um, but then the other thing too is just how easy is it for a fifth grader, a third grader to understand what it is that you're doing? So when you ask like you were this targeting, I began with what I tried to make a, uh, non technical, non jargony, like simple explanation and then went into the technical. Because most people can't process a lot of technical information if it's not their specific field.

Speaker A: That's right.

Speaker B: You also want to be mindful that not everyone learns best by reading. Um, some people learn best by seeing. I happen to learn best by hearing which is why I love the sound of my own voice. Yes. So, yeah, you just want to just be careful, um, how you think about your various creative types of.

Speaker A: So I want to shift a little. So one of the reasons why. Well, the reason why you came on my radar was Miriam Jessier on LinkedIn. They shared a really great, uh, takeaway from a presentation you gave recently. I think it was about a month ago. I'm going to read what they said, uh, in the LinkedIn post and then ask for your take on it and interpretation, because this really caught my eye. So they said, brand performance converge in AI answers. This is the framing we need in the SEO community if we hope to make any headway getting through to the branding department. And this was a takeaway from a snippet of your presentation, so I'd love to talk about that a little bit more. So brand and performance converge in AI answers. Yeah. Can you unpack that a little?

Speaker B: Sure. So when I was, uh, looking to get my, um, sweater, which was the opening story in that talk, I specified a very specific, um, brand, and I had my request. AI, um, in this case, Copilot could not provide me the specific product I wanted. But what it did do is, is it provided me two other examples of potential products. That is, in a nutshell, a brand discovery moment. If I was open to a different brand, as well as the opportunity to buy right in that moment. So the conversion path is shrinking so quickly because people are going from ideas and conversations and just exploring things to, in a matter of two or three, uh, beats into that conversation. Okay, I'm ready to go by. I'm ready to give you my money. Uh, Mr. Or Mrs. Brand. The thing that I think really clicked, uh, for Miriam, who's amazing, and everyone should go follow them. They're wonderful. Um, is that the content and the social piece is more about the human trust. So the trust signals that we confer, um, by recommending somebody by empowering someone and inciting them and giving them their flowers. Those are signals just as much as the technical signals and getting all of your tech SEO correct. Um, so in the other example that I showed, I showed, um, uh, a maxi dress that copilot was ready to send me straight to one brand. But because the, the SKU that I actually was ready to, to buy and I thought was, was a better fit for me, had a better price or in, like, better reviews on this other brand. That was another discovery moment of like, okay, I'm going to go with this brand. I'm going to go over here and buy it here. So being able to keep up with uh, uh, a uh, conversion path that ideates to transacts and sometimes as little as 30 minutes, sometimes even faster, like five or 10 minutes. You can't use conventional funnels and tracking that are used to thinking in 24 hour cycles.

Speaker A: In that case, I assume in the first example it was two different brands entirely of two different products that were recommended to you other than the brand you already knew about and you were ready to purchase from that brand. What do you think that brand was doing wrong? If you could hypothesize, why did that bring?

Speaker B: I don't have to hypothesize. Co Pilot told me what they did wrong because I asked it. I was like why? Why what? What's going on Co Pilot? Where's the brand I asked for? And the specific answer is they didn't have a crawlable product catalog to use. So because they didn't have a catalog they tried, they genuinely tried to provide me the brand they couldn't. So they on um, the back end did their various queries that would have been those citation queries. And then I had two other brands with two different products come up. So uh, the feed hygiene and the product catalog point, I cannot overstate it. It is easily the most important thing. And what's really interesting is that you don't need to spend money to serve here. Like you can get your ucp, um, Universal Commerce protocol set up for free, um, in various merchant centers. If you're a Shopify customer, it's going to definitely be there already. Um, but the thing that I don't think enough people are really thinking about is also the review component, the human component, because price is important for sure. Um, but the reviews that people leave really do add weight. Um, because AI ultimately wants to be helpful or it's programmed to seek the most helpful option. It's not sentient. Um, so if you position yourself as the most helpful option, you'll be in a better spot.

Speaker A: Yeah, that's very fascinating. That Copilot just told you what the problem was. So when you're referring to a feed, if I'm thinking about like Google Merchant center, right. There's a product feed in that scenario which companies can take use of for free. Right. You can have free merchant listings in Google. So that's one type of feed that comes down. Are there specific other types of feeds that come to mind? Like that's what you're referring to when you're saying Copilot didn't have A crawlable shopping catalog.

Speaker B: Sorry, what was the, uh, I missed the question in there?

Speaker A: So the question is you mentioned. Copilot said they didn't have a crawlable product. Yeah, catalog. And we're talking about feeds. So one feed that comes to mind is Merchant center, the feed for Google Merchant Center.

Speaker B: So Merchant center isn't actually a feed. Um, Merchant center houses the feed and Microsoft has one as well. Um, Merchant Centers are just the way to process that information.

Speaker A: That's what I'm saying. You supply a feed to Google Merchant Center.

Speaker B: Yes.

Speaker A: Right.

Speaker B: Or Microsoft Merchant Center. Yeah.

Speaker A: Is that the same feed? Like literally the same type of feed? Okay, so. And what is that type of feed called?

Speaker B: Get.

Speaker A: Uh, for anybody that. Does it have a name?

Speaker B: Just a shopping feed.

Speaker A: Just a shopping feed. Okay. Are there other types of feeds?

Speaker B: Yeah, so. So there's like, um, there's feeds, uh, for auto, like the auto catalogs, there's travel, there's real estate. Um, you can actually build your own feed. Um, I did an article talking about Legion in AI and that like you can actually build out with an Excel file or if you really want the Google sheets, um, as long as it's the URL and then fill in the blank with various categories, as long as you keep that maintained in a standardized way, you can upload that and give that additional context to your campaigns. So, um, this is not necessarily something that everyone is doing because it requires work and thinking. Like, the more, the more important piece is, is your landing page easy to understand? Like, I would not worry about this like the A work if you're in lead genes. Like, I'd be much more concerned about CRM integration and the cleanliness of that data. But if you have all that set and your landing pages are immaculate, then yes, like, feeds is an additional area. But for those of you in E Commerce or those of you in travel, um, Auto. Absolutely. Yeah, exactly like absolutely. You want to take care of that?

Speaker A: Yeah. And that was kind of the question because, uh, my brain starts to think about the LLMs TXT file, which is debatable. People don't seem to agree on if you should have one or not have one. And so the idea of having a feed for lead gen or something that's not, uh, like ecom in my mind. That goes to the question of how do we know that our time is worthwhile to develop a feed or to use something that falls a little bit more in the speculative edge of feedback.

Speaker B: So if you want to be found in the AI era, where conversion rates, uh, at Least Microsoft's data shows that it's um, 25% greater relevancy um, for uh, AI surfaces versus a search result page. And then I think it's like 194% better conversion rate um, on an AI surface than search result page. So like the Surfaces clearly convert. The thing is the volume on the paid side is actually going to be lower than a traditional search result page because there is such a high bar to be relevant for that AI placement. Organically it's a little bit different um, because organically if you are a good answer, like you're going to be there, um, or ideally you would be there, I would put a stack ranking on tasks to consider or things to think about in the following order. The first is are you being crawled and indexed? Like are you even available to participate? If the answer is no, um, there's a free tool called Index now through Bing Webmaster Tools and Bing Webmaster Tools also has the AI Citation Report. There's also the Clarity one. Uh, folks aren't familiar with those. I can go in depth in a bit. Um, but basically the reason why this is so important is that you can ask Bing, hey, check on this for me. Like make sure that this is up to date. Because if it's not up to date, that's where you get issues of incorrect information or just not even being found. The next pecking order, uh, is that you have said yes to AI commerce. Um, so ucp, um, is a protocol that allows you to do business plus you directly. Um, um, or allows you to do uh, this is um, agentically or through the AI system. But if the human still wants to come to your site, they can, um, so just making sure that you've said yes to those things you should do. The final thing I'll throw out there that is really important, um, is that you have set yourself up from an accessibility standpoint. Accessibility isn't just high contrast landing pages. Buttons are buttons. Like we're not using giant in the way formatting. Accessibility also comes down to how you're communicating. Um, so if your language is difficult to read and m. I'm not going to get into the serif, sans serif debate, but like if it's really difficult to understand what it is that you do. If humans would struggle like AI is going to for sure struggle.

Speaker A: So you're talking about ADA compliance, Accessibility.

Speaker B: Yeah.

Speaker A: How does an LLM. So I've been, I've heard a myth, maybe not a myth. Uh, this is. I'd love your take on this LLMs just look plain text. So if you strip away everything and you just have the plain text that, that's what they look at. Are LLMs now also looking at or do they look at the UX of a website when it comes to fonts and ADA readability?

Speaker B: So I'm not going to comment on like specifics because like it's always changing. Yes, I would encourage everyone to check Microsoft, uh, put out a really good um, AEOGEO playbook. Um, Maryam's actually one of uh, the guests in there. So again, shout outs to Maryam. Uh, but the other thing I would definitely encourage folks to check out is the AI principles of your LLM of choice. Um, because most of us at this point have set out what our AI principles are. So Microsoft makes a point about talking about accessibility, inclusivity, accountability, security, like transparency. Um, and if you care about that, like that's something just to be aware of that, like the AI systems are going to have uh, that mindset. I'm not going to say that we're unique in having AI principles like pretty much most platforms do, but I would check against it and then you can stack rank how much you prioritize based off of your AI of choices focusing or not focusing on those things.

Speaker A: Yeah, and I'm a believer that we should care about accessibility and ADA compliance for humans. Like it's always interesting to me how a lot of people just have like a robot's first lens or like a crawler first lens. But to me it always seems like if it's something that humans would benefit from and allows you to reach more humans, that that's going to benefit you algorithmically whether it's a crawler or an LLM. So that just seems like something that you should do no matter what. Uh, and I was going to ask about IndexNow, is that something you can just sign up for for free like through Bing Webmaster Tools? Yeah, I don't, that's another one of those tools. I don't think enough people talk about it, you know.

Speaker B: Well, well not only is it like mind boggling that people don't talk about it enough, um, because the Bing Grounding API, uh, powers most AI assistants. When you do a great job with Bing being able to find you and understand you, you've done the work for most other AI assistants. So again, leveraging Index now is not a guarantee that you're suddenly going to get vast improvements in crawling and findability and all that. But it is a way for you to say, hey Bing Microsoft, um, I Made a change. Please investigate. Um, as a reminder, uh, what we consider a change is a net new content, a meaningful change, uh, to an existing bit of content, or a deletion. Please don't delete good content and get us to come say hi and investigate you. It's much better to either make updates or add net new. Um, but if you have any questions, uh, there's a number of really, really good resources I'll happy to send along to go with this.

Speaker A: That'd be great. I can put them in the show notes as well. Uh, I had a question about where did that. Oh, so you mentioned in another interview that video is no longer an option. And so I'd love to talk. So it was in reference to, uh, I think it was an interview you gave last year and basically you were talking about how for companies, for brands not doing video.

Speaker B: Oh, I was like doing video is not an option. I'm like, what? When did I say that doing video was not. Was not option? No. Okay, so I, yes, video is important. Um, 100% still agree. Um, the reason though that video is so useful is because people consume information in the way that they best process it. And there have been a number of studies that have shown that 65% of humans learn best by viewing something. Again, I'm the minority by learning, by listening. And if you are asking people to read a big wall of text, or if you're asking people to kind of infer things from static images when you could just give them a video that's really quick, super engaging, explains exactly who you are, it's much more useful. It's actually the same reason why text is where tone goes to die. Um, like it's better to just hop on a five minute call and just have that conversation with the person and get that clarity. Uh, videos do a lot of that good. All of that said though, not all videos are equal. So when you're thinking about how to allocate time, the best time spent is where your humans are. So if you know that your humans are on LinkedIn, for example, LinkedIn has been getting really into the short form video content game. Um, it is worth investing a little bit of time just to do some short form videos. On the flip side, if you know that your viewers are into podcasts, uh, video blogs or vlogs as it were, you, um, might invest that time in the longer form content and then repurpose, um, the bits for snips for whether it's shorts or on LinkedIn. Um, I think substack also Allows for videos. Yep. There's a number of ways to make

Speaker A: it work and that's all just really good marketing. Like, like I was talking about earlier, if your audience is there or consuming something in a certain way, you should be there as well. I'm wondering if there is a deeper reason in the world of AI to be investing in video. Uh, one thing I heard, I don't know if you have any clarification around this. I had heard from a few sources that Google Gemini may look at YouTube short videos for training, for LLM training data or to pull in information.

Speaker B: So I need to be really clear. I, I don't work for Google so I can't comment on. I just know on whether they do or they don't.

Speaker A: Yeah, uh, um, I don't know as a, as sort of a, you know, industry peer colleague, I don't know if you had heard anything, uh, about that as well. Just, you know, out in the public.

Speaker B: So what, what I will say is there is a reason that you are seeing that you might be seeing more and more video content and sources in AI surfaces. It is great and a great way to get found doing YouTube content, doing LinkedIn content, doing uh, Reddit content. Um, but in terms of like one type of content being weighted over another, like I just, I genuinely can't say what Google does.

Speaker A: Yeah.

Speaker B: What I will say on the copilot side is that we have different models that like rely on different sources. So, so if someone wants a more thoughtful like indexed approach, you'll typically actually find more sources that are longer form content than if someone's looking for quick, speedy answers. Like you find shorter form content. But it's not that like you're going to get ranked higher or lower because you did short form content or long form content. It's just like what is the type of answer that will best meet that need?

Speaker A: Right. Are there other things that you think SEO, other ways you think SEOs and PPCs can collaborate that we have not talked about? Because there's an unknown, unknown territory for all of us. Like I don't know what, I don't know. I don't know what's going on in the world of PPC and advertising that I could be learning about or collaborating with. When it comes to doing SEO and combining that with ppc, is there anything else that comes to mind that we should be doing?

Speaker B: The biggest thing is have a shared Clarity account. I know that might seem like really shameless self plugging. Clarity is free. We charge exactly $0 to leverage Microsoft Clarity, um, and what it does is it allows you to see uh, how the humans who engage on your site, what they do do they get stuck do AB tests on landing pages. Uh, so it's the behavioral analytics and then also the AI reporting. So AI citations, um, your crawl budget that's going to different AI bots, things like that. The reason I think that is the most important for PA teams and organic teams to work on together is that ultimately the landing page experience is a thing that both have a responsibility to ensure is correct. Because ultimately if the landing page can't close the deal, it doesn't matter how great you were at ranking, it doesn't matter how clever that targeting was in the paid campaign you lost the person. So we want to make sure that it's easy to do the thing that you wanted them to do, but you also want to make sure that the people you're getting are correct. So Clarity can also be really helpful to uncover if non helpful bots are coming to your site.

Speaker A: Mhm.

Speaker B: Um, so that's just a useful thing to be aware of.

Speaker A: Are you uh, aware of anywhere to get real prompt data? I know, I believe I just saw Clarity announced that you're providing more fan out query data in that or in Bing Webmaster tools. And that's great, I'll take that data. But is there anywhere to get prompt data?

Speaker B: So prompt data is a privacy issue.

Speaker A: Right.

Speaker B: Just like you wouldn't want people looking into every little thing that you conveyed from like a highly personalized assistant.

Speaker A: I don't want somebody in my cloud account.

Speaker B: Exactly. So if you don't want someone in your cloud account.

Speaker A: Yeah.

Speaker B: So we're actively thinking about how can we reasonably and responsibly provide as much data as we can so that we're showing our work. But I, I don't know that we ever want to get to a point where we are showing full prompts. Because if we show full prompts that that's a really big violation of privacy. Um, because the types of things that people put into prompts, they're not just simple little keywords. Like they, they are really thoughtful, sometimes highly personalized information. Like I'll give even an example, um, with m. My, my husband and I, um, we were going through like a really rough period like a year and a half ago. And like I was asking like, would I have been better off not marrying my husband? Now obviously the answer is no. Like my husband is one of the best humans in the world and like life with him is amazing and like I'm very grateful that I have an amazing, lovely in him. Um, but like that sort of personal. Yeah, you don't, you don't want that being visible.

Speaker A: Yeah. And I agree. I think privacy, privacy is very important. So obviously that's a big concern. What can companies, if you're a company that wants to understand how people are prompting and using LLMs, you know, I thought of like, do you survey people? Like, how do you get an idea? Because again, I, there's a positive intent here, which is to understand how to help people that are using these new surfaces. And I don't know of a better way to do that than knowing how are people interacting and prompting and speaking to these things.

Speaker B: So here's the thing. You don't need to know what they prompted if you know what the AI searched to help them address that prompt. Because if you know what the AI is searching in the background, you can get a sense of, okay, do I actually match that? Like, is coming up for this sort of citation useful to me and my brand? So I would argue that you actually don't need to know the human prompt at all. You need to know what, what is going into that search, uh, on the back end that's going to ultimately inform the answer given to the human. But in terms of like actual attribution is effectively what I think you're getting at. Yes. Ask your humans how they're engaging with you because most people actually don't do this. They don't take the time to ask, yeah, did you, how do you hear about me? Where, where do you find me? Uh, and even if the answer is not the last click attribution source, that's still helpful to know because then you know, okay, this thing that I might have pulled budget from is actually still helpful.

Speaker A: Where do you think the future? Uh, the big existential question, okay, so be prepared. Where do you think the future of the web websites? I know in another interview you talked about, or even now you talked about, let's stop thinking about landing pages so much and think about the agentix scenario. You gave an example in another interview where you had an issue with your dog and you could have gone to like an agent and that, uh, agent would have helped you figure out what to do medically about the issue your dog was having. Uh, so that's one part of, to ask the question, like, do you see a future where there's no websites, there's no web as we know it, it's all agents. How are you thinking about that as we're moving into the future.

Speaker B: So to put everyone's minds at ease, I don't think there will ever be a world where websites fully go away. I just think that most people won't engage with them. Um, and the reason most people won't engage with them is that websites as a tool are really clunky. Um, so a really good way to think about it is you can read Wuthering Heights. You can go through the several hundreds of pages of melodramatic Heathcliff screaming into the moors and Catherine totally being a horrible person to him. This is only funny if you've actually read Wuthering Heights, or you can read the Cliff Notes, or you could ask AI, summarize Wuthering Heights for me in a minute or less. Different people process information differently. And so the people that will still want to sit and read Wuthering Heights or, or uh, like even thinking about, like Stephen Colbert and Lord of the Rings, like he goes back and rereads Lord of the Rings, there will be people that watch the movies and book purists will say, how dare you, you're violating my sacred text. And then there'll be people who do even more than that. Websites are going to be a constant layer because that's also how like the agentic layer on top of websites will engage. And humans are showing a desire to want to engage in that way, but there will still be humans who choose the website. So, um, making this more practical and less book related, um, people who say yes to one click checkout have already said yes to the first step on the path to agents. People who do not do one click checkout or who do not do one click passwords and don't want to have that information saved. There's never going to be a world where the agentic future will be something that they'll say yes to. So for them there still needs to be websites. It's just I wouldn't count on websites being how the lion's share of your humans engage with you. It's, it's how you protect the ones that still want a website, but then also how you communicate with agents, what your value is, the accuracy, things like that.

Speaker A: It's interesting because I see so much of the other sentiment where for example, you probably know Google I/O. Just happened and they made some big announcements about how they're completely changing, or there was like an article on TechCrunch or something about uh, the web is over as we know it, or Google searches over, like what, it was a very click baity headline. Uh, but somebody shared this news in TikTok. And then I was reading all the comments, every single comment, hundreds of them were like, we don't want this enough of AI. Like they, we want the old Google. People want the old web, the Internet back where you can click around and discover websites. And it had a charm to it, uh, and a fun to it. So it's interesting how I'm seeing such strong sentiment in one direction. But then you're right, there are people I find AI, uh, useful, you know, in many applications. I went to an article you wrote and summarized it with Gemini, so I could just take in what you talked about two years ago in that article much more quickly. So it's, it's interesting because I'm torn. I don't love it or hate it. I'm somewhere in the middle and I think there's a lot of people that just refuse they don't want it. They're like, go to DuckDuckGo, you know,

Speaker B: uh, which ironically DuckDuckGo is a partner with, with Microsoft. So like we're, we've, we're, we're all for it. Like, absolutely. Duck, duck, go forever.

Speaker A: Yeah. So to kind of wrap things up, I want to make sure. Are there any other AI based, uh, I don't know if I asked this question correctly, but are there any other ad features within Big Bing or Microsoft, um, that people should know about something that's exciting to you that you think is a big opportunity?

Speaker B: So one thing that I definitely think not enough people realize within Microsoft, like a. Has been serving in copilot since 2024. I think I said that already. But the thing that I personally find very exciting is impression based remarketing, which is that you have.

Speaker A: What is that again?

Speaker B: Impression based remarketing.

Speaker A: Okay.

Speaker B: You can target, exclude, bid up or bid down based off of someone just seeing your ad. Now why that's incredibly exciting is that if someone sees your ad on Copilot, that counts for impression based remarketing, that you could then target them on search. So why this is interesting. People who maybe have thoughts and feelings about, um, I want to just run brand to people who haven't ever seen my ad before. You can have impression based remarketing list that then excludes everyone who's seen your ad before. Um, you can also vice versa. Like if you want to make sure that you're targeting exclusively lower in the funnel off of someone who has already seen some messaging and get a more targeted offer, you can do that as a targeting mechanism. So long as there's at least one audience ad in the up to 20, um, seed source lists. So it's either an ad group or a campaign. Any campaign type can speak to any campaign type. So that's something that Google doesn't have. Um, it's something that Microsoft built specifically because we heard people that you need a privacy compliant way to engage and sometimes the click is a really expensive way to preclude someone like you shouldn't have to wait for the click to build the remarketing. Uh, so that impression based remarketing is really, really, really cool.

Speaker A: That's great. Impression based remarketing. One of the top PPC people that I work with and we collaborate on clients, he swears by Bing, we had a client shut off their Microsoft ads and he was on them every day until they turned him back on because they, they performed so well. Uh, so I think that's just good for anybody, even if you're an SEO. Like I often when I'm working with clients, I try to bring things to their radar that are outside of SEO because if they could benefit the client, you know, I, that's just helping everybody. So I think if you're an SEO listening to this and you're like, well, I don't run ads. Why do I care about that? I would mention this impression based retargeting to a client or your company or your site you work with because I think anything that, you know, helps in the advertising side is beneficial in one way or another to the SEO side. Uh, so as we wrap things up, it's been great chatting with you. Uh, I knew from the beginning I didn't provide so for people listening, I did not provide Nava any notes or questions ahead of time because I listened to you speak in other podcasts and I saw your material and I know that you can just have a conversation. And that's what I love about people coming on Experts in the wire. So as we wrap things up, uh, any place you'd like to direct people where they can find you? Any call to action? This is your chance to mention anything and everything here.

Speaker B: Yeah. So first, thank you so much for the conversation. This was, this was really, really fun. Um, people can find me on LinkedIn. I do a monthly column on Search Engine, Search Engine Journal on Search Engine Land. Uh, as I mentioned, I'm on the board of the Paid Search Association. We have a Slack group. I'm also very active in the ppc chat, women, tech, SEO, Slack groups. Um, and I would encourage everybody, if you are at all in the Boston area. I think, uh, Dan and I are going to be there next week for SMX maybe.

Speaker A: Yes, yes. And I'm going to try to post this episode on Monday. So if you're listening the day it comes out right, it'll be tomorrow. So we'll see uh, maybe hopefully some people in Boston tomorrow. Um, but uh, no that's, that's great. And so uh, what I'll do Nava, if you send me some of those links or I'll, we'll get them in the show notes for people as well. So uh, the best place to find Experts in the Wire is Spotify with full video and that's where all the shownotes are and everything. But you can also check us out on Apple, YouTube, et cetera, all the places. So Nava, it's been great having you on Experts in the Wire and we'll talk to everyone next time.

Speaker B: Cheers. Dan.