Join Nate Matherson as he sits down with Matt Lenhard for the sixty-first episode of the Optimize podcast. Matt is the co-founder and CTO of Positional, a company building an all-in-one tool set for content marketing and SEO teams.
In our episode today, Nate and Matt explore topics like AI overviews, SEO optimization, and Google's SERP strategy. Matt also speaks to the mechanics of AI overviews, expressing insights from his recent study on their workings. Listen to hear why he suggests using AI overviews as a prompt for content improvement.
In this week’s deep dive, listen to Matt share everything you need to know about optimizing for AI overviews in 2024. Matt gets tactical, sharing multiple case studies on how vector similarity impacts the likelihood of being referenced in AI overviews. Rounding out the episode, Matt and Nate cover topics like the correlation between SERP ranking and AI overview inclusion, and the future of backlinks in SEO. Closing the episode is our popular lightning round of questions!
So there are really two parts to an AI overview. There's the text that appears and then there is the source links that appear below that text typically. Do I have it right that like, in theory, we think that Google is generating that text before they determine what are those sources or source links that they want to pull into?
Yeah, and the reason why that might have been contrary to some initial belief is like, you might think that Google's finding the source and then summarizing that source and pulling the content out. Like how feature snippets work. Sort of work right where they like find the source to pull the data out, but it kind of seemed flipped at least based on the patent that the kind of opposite was happening where they're generating like the snippet and then finding, okay, who verified, who can we source for this snippet?
Hi, and welcome to the Optimize Podcast. My name is Nate Matherson Matherson and I'm your host. On this weekly podcast, we sit down with some of the smartest minds in content marketing and SEO. Our goal is to give you perspective and insights on what's moving the needle in organic search. Today, I'm thrilled to sit down with Matt Lenhard Lenhard.
Matt Lenhard is the co founder and CTO of Positional, a company building the all in one tool set for content marketing and SEO teams. You might've heard of them. Now we've talked about AI overviews on this podcast before, but on today's episode, Matt Lenhard and I go in depth on AI overviews and Matt Lenhard shares some really interesting data and insights from a recent study that he's conducted.
Thanks for listening to this week's episode of the Optimize Podcast. Podcast it's brought to you by positional. If you don't know by now, my name's Nate Matherson and I'm one of the co founders of positional and we've built what I think is a pretty awesome tool set for content marketing and SEO teams. We've got tools for keyword research, internal linking, content optimization, and even a couple of tools for analytics.
We'd love for you to check it out at positional. com. Matt Lenhard, thank you so much for coming on this week's episode of the optimized podcast.
Hey, Nate Matherson. Thanks for having me. Long time, no see, about five minutes ago, excited to finally. I'll be on the podcast and chat. Um, I also, I love the podcast voice. It's like slightly different. It's kind of funny.
For all of our listeners that don’t know, Matt Lenhard is like right over there. Um, I can see him, but we're recording from two different locations, uh, in the same room. Um, Matt Lenhard, the first question I ask all of our listeners is how did you get into the world of content marketing and SEO?
Well, I'm sure a lot of them or a lot of your listeners have probably heard a slight variation of the story because I got into SEO the same way you did, where we started a company together all the way back in college.
Um, it was in the financial lead gen space, extremely difficult to acquire customers profitably in that space. We had tried, you know, AdWords, PPC marketing on like all the social networks. Taboola Display Advertising, and it was really a struggle to make the customer acquisition costs work within the bounds we needed it to.
And luckily we ended up finding SEO. We had some like early success there and kind of, you know, decided that like this Was the most efficient ways to acquire customers in that vertical, um, dove at first and, and kind of been doing it with you ever since. So it's been like 10 years, I guess, at this point.
Yeah, we've been working together a long time since, uh, I think like 2014. So it's been a, it's been a decade, a couple of different companies, a lot of SEO, a lot of blog posts, and it's been a lot of fun. Um, and I've been looking forward to this conversation all week. One, because it's. The first time I've ever interviewed you for anything.
Um, but also to, uh, you know, everyone's been talking about Google's new AI overviews, you know, chat GPT just launched search GPT, or at least announced it this week. So there's a, there's always change in a SEO land. Um, but from my understanding, you've recently done a pretty interesting and new analysis on, uh, Google's AI overviews to help, uh, our listeners and other SEOs better understand them, how they work, what goes into them, um, and I'm excited.
I'm excited to unpack some of those learnings with you on today's episode. Um, but Matt Lenhard, uh, first, could you tell our listeners what is an AI overview?
Yeah. So I'm sure a lot of people probably saw the chatter when they initially released and everybody was kind of making fun of Google, but they're the. A snippet of text that will appear above the normal 10 blue links or the normal search results.
And it basically will give a generalized overview of the query, kind of in more of like a, you know, chat type interface. Sometimes there will be reference links. There's normally a see more button where it will continue generating. Oftentimes there's more reference links there as well, but it was kind of, to put it succinctly, like Google's answer to chat GPT, um, and just a new way of organizing the SERPs.
How often do AI overviews appear, or at least, you know, broadly, do AI overviews appear for every search query or for only some search queries?
The first thing to point out is they don't normally appear for, uh, incognito users. So users who aren't logged in, uh, the vast majority of the time aren't going to see the AI air reviews.
So that's like a, a big kind of like threshold that needs to be crossed. Um, and then alongside when they are shown, I don't have exact numbers. Um, I think the SEMrush data was like a few percentage points, but even that, that's all the crawlers are done through logged out users. So I don't actually know that there's great.
Insight into the exact proportion of SERPs that have them. The reason being nobody is crawling the SERPs right now, or the vast majority of people aren't crawling SERPs at that type of scale, uh, with, um, logged in Google accounts. So it kind of adds this like whole extra layer of complexity to the, uh, scraping that needs to be done.
And that's why there's not a lot of super. Accurate, large scale kind of datasets.
Yeah. totally agree, uh, and that's good context for our listeners, just to play it back to you. It sounds like AI overviews only appear for logged in users, and so knowing how often they appear right now is actually quite challenging because most SEO tools don't.
Crawl or scrape the SERPs, uh, from logged in accounts, they do it from unlogged in accounts. And so it's hard to know how often they appear. Um, before the recording, you and I chatted a little bit about the mechanics of an AI overview, and you mentioned a recent Google patent to me that might offer some insights into how Google is generating AI overviews.
Um, it's a question I always get asked by our customers. How, how does Google determine what gets included in an AI overview? Uh, and so I'm going to ask you for our listeners, how do you think Google is determining what it pulls into or displays in, uh, in AI overview?
I saw a little bit of chatter about this on Twitter, which is kind of like why I initially brought it up to you.
Um, and I guess. Before the like AI interviews have rolled out, or like this is actually before even SGE, Google's precursor to AI interviews had fully rolled out, there was a paper outlining how you could essentially generate kind of like LLM type responses to search queries and then find references for them with the API actual generation of content being pretty similar to, um, how typical LLMs work. But what I kind of found like most interesting is, you know, what in regards to like what we do, how do you, how are they determining what gets pulled in as a reference alongside the text that's being, um, generated? And what I found was that, um, there was a few things that kind of were contrary to my initial thoughts of how I would have envisioned it working.
The first being that I guess like the most important almost being that they were generating the text first and then looking for sources based on the text that was generated. And the path that kind of like that led me down and why ultimately I brought it up to you was, um, if there The text is generated first, then they must be finding what to reference based on some type of, like, vector similarity calculation, uh, alongside, like, the signals that they typically look at.
And so, it's cool to have, like, a starting point for, okay, this Maybe a possibility of how we could start optimizing.
So there are really two parts to an AI overview. There are, there's the texts that appears, and then there is the, uh, the source links that appear, um, below that text typically, um, do I have it right?
That like, in theory, we think that Google is generating that text before they determine what are those sources or source links that they want to pull into.
Yeah, the reason why that might have been contrary to some initial belief is like, you might think that Google's finding the source and then summarizing that source, um, and pulling the content out, like how feature snippets.
Sort of work right where they like find the source and pull the data out, but it kind of seemed flipped at least based on the patent that the kind of opposite was happening where they were generating like the snippet and then finding, okay, who verified, can we source for this snippet?
Yeah, that sounds very different from a featured snippet, uh, and also from even a SERP feature like, you know, people also ask where Google often generates those also ask questions from.
The text of a source, uh, directly. And, uh, yeah, I agree with you. It feels like it would be the other way around. But maybe that is an opportunity for us SEOs. Uh, and, uh, I want to talk a little bit about that in a minute. Um, but I know last month you worked with a member of our team to collect a large number of AI overviews and to do a little bit of research on them.
Uh, can you tell us about the data you collected and the reason you decided to study AI overviews?
Yeah, so I, you know, they were obviously kind of brand new, super exciting. And after reading this paper, I kind of had a hunch for how we could start to optimize for them. So I know I came to you with a lot of enthusiasm, like, Hey, can we rewrite these blog posts?
I've got an idea. But yeah, so since, like I said before, that there was this issue where you needed a logged in user. It can be kind of difficult. So we actually like paid freelancers to just manually go as they're logged in and collect a bunch of these like AI overviews, all the reference sources across a number of different verticals.
So we did like 500 AI overviews for 500. They were all specifically targeted mainly to generate reviews. We actually figured out that questions are much more likely to generate reviews. So we targeted those and then paid a freelancer to go and collect all of them. And then the pieces of information that we collected were All of the URLs, the SERPs, which we collected with a script.
By SERPs, the top ten blue links. The text that was matched in the AI overview for the source. So, as some people might have noticed, when you click on a AI overview source, It will link you over to that source, and it will highlight a piece of text on that page. That's often confirming or, like, referencing the AI overview itself.
And so, um, I collected that alongside the actual text in the article. And, uh, and then finally the position of, like, all of these sources. And, um, I thought the findings were pretty interesting. There's kind of like a few key things there. The most interesting thing that we found was kind of like confirmed my initial theory was that there's a high correlation between the cosine similarity of the AI overview alongside the matched text from the reference sources. So I'll kind of like break that down.
Cosine similarity is a way to compare two vectors against each other to make that even simpler, take two pieces of text and tell me how similar they are to each other. Um, and what we found is that, uh, there's a high correlation, almost like exactly linear between the vector similarity, the reference, uh, source. Alongside, uh, how likely it was to have that, like, top position.
So there's definitely, Google's probably using something more sophisticated than Curse Unsimilarity, namely because Curse Unsimilarity doesn't look at if it's accurate, or like, the magnitude of the emotion. So, Curse Unsimilarity would say, You know, it was all the husband's fault and it was all the wife's fault.
They're pretty similar, even though they're kind of saying the exact opposite thing. I'm sure Google has like, uh, another layer where they're kind of verifying things like that as well. But what, what we essentially found was that, uh, there's a high, very, very high similarity between the reference sex and what Google is pulling in.
And so from there, I was able to You know, go and update a few of our pages, optimize that cursor line similarity in the given sections, and, um, in a few cases, get the, uh, become one of the reference sources.
So by that first spot, you meant reference sources. And just to play it back, so it sounds like Google is generating an AI overview.
They're then providing these reference sources. If you click on a reference source, it takes you to a piece of text within that web page that You've collected and like in your analysis, you found that like the text that Google is referencing has a very high cosine similarity to the text that Google is generating for the AI overview.
And if we buy that theory that Google is generating these AI overviews first and then looking for sources that support what they already want to say, then in theory, it could be possible to optimize for placement in an AI overview by, uh, Creating text or including text that's high in cosine similarity to what's already existing within an AI overview.
Do I have that right?
That's a hundred percent right. And you've seen less sophisticated kind of chatter of this on Twitter. People basically found, look at AI reviews, taking the text, rewriting it. And publish it and been able to become one of the reference sources, like all this being said, you know, I don't think that'll last forever.
It's probably going to get more sophisticated over time, but it did seem that like in the rush for Google to release this, there was some like cracks maybe in how the data was getting pulled in.
Yeah. Was it Cyrus Shepherd? Am I getting this wrong? Was it Cyrus that had done a similar kind of analysis and, uh, and rewrite to his piece of content to then get it pulled into the AI overview?
Yeah, it was Cyrus. Um, I think he did end up losing them, but it's kind of the same thing that we found that like, if you take the similarity of the two pieces of text, Mind them up, and you're already ranking decently well, so you're doing other things correctly, there is a good chance that you can get pulled in.
Yeah, you know, your point on does this work forever or does this last forever, I would think probably not, because like, as soon as like, uh, Google senses that people are trying to manipulate The SERP page, uh, they generally don't like that, but, uh, and I wouldn't advise like the listeners of this podcast to go and modify all of their pages in some unhelpful way, just to get included in an AI overview, but, um, Well, maybe it's something for us to think about, like, as we're, uh, as we're creating a new piece of content, you know, we, we already go through that checklist of, you know, do we have an H1 and maybe, maybe there's another part of that checklist, uh, where we try to see if we can, in theory, optimize for an AI overview, um, assuming it makes our content better.
Yeah, I think you definitely want to focus on like, does it actually improve the content? Um, Google, anything that can be scaled in a kind of spammy like way, Google always cracks down on. Like those things never last. Um, so I think tackling this from like, how do we use this information to make our content better?
It's probably the better long term way to look at this.
Well, I think that's a really interesting point because you and I kind of tag teamed the edits together, um, where like you had gone back to some of the blog posts that I'd written and tried to figure out like if we could modify or add some texts to possibly get our pages included into an AI overview.
And I found that like in a lot of situations, the texts that we were adding actually made the article better. More clear, uh, or, uh, provided something that maybe we hadn't been previously. That was a helpful context to a reader. And so through this exercise, you know, yes, we were able to get our, our articles and some of these AI overviews, but.
In any case, I think with many of these pieces of content that we edited, we actually just made our content more helpful. So maybe that's another way to look at this too. Like if an AI overview is saying something, it's probably pretty important. Um, and, and maybe that could be something we'd want to touch on in our piece of content, regardless of, of if it gets included in the AI overview or not.
Do you think that's a fair way to think about it for all the listeners out there?
Yeah, I actually think an interesting approach is that you can take sections of your article and use the AI reviews as a way to like prompt, like, what am I missing from this section? Like, maybe you're, how do I improve my bounce rate?
Right? And, you know, But you have an article on bounce rate and there's a section. How do I improve it? You can look at the AI review and see if Google came up with any like interesting ideas that you didn't and then add them to make the piece of content better. So I think like looking at adjacent AI reviews and using them to see any of the subsections of your article are missing anything, let's say.
An interesting strategy and just a way to like make the content overall more helpful.
And now just a quick word from one of our sponsors, actually our only sponsor that is positional. com. Positional has what I think is an awesome tool set for content marketers. We've got everything you need to take you from start to finish and help you scale this channel as I've done over the last 11 years of my career.
We'd love for you to check it out at positional. com. And, you know, another question that I always get asked, uh, is, uh, is there any sort of correlation between, like, uh, you know, your position in the SERP, like, one of those ten blue links, and the likelihood that, like, you'd, you'd get pulled into an AI overview?
And I think you did a little bit of analysis here as well. So, just to recap, do you think there is any sort of correlation between a SERP's current ranking and the likelihood that it'll get pulled in as a source in an AI overview?
Yeah, there's definitely a correlation. I'm sure we can include like the exact numbers and like, um, of reference when we post like the show notes.
But yeah, I mean, the higher you read the first spot, you know, double the amount of references as like the next. So generally The things that work well for regular SEO or SERP optimization are going to work well for AI reviews, so I think stick to what's already working and there's a good chance that you're going to show up in the reference.
Yeah, for our listeners, Matt Lenhard did send me some of this data before this recording, so I have an advantage here. And I think, Matt Lenhard, from the data you sent me, it looked like the number one search result was 55 percent more likely to be pulled in as a. AI overview source, uh, versus the number two search result.
Uh, and as another example, uh, the number two search result was 90 percent more likely to be pulled in as an AI overview source, uh, versus the ninth search result. So it does appear that there is a correlation between current SERP ranking and, um, the likelihood that a URL will get pulled in to the sources within an AI overview.
Matt Lenhard, I know, you know, this is just one analysis and it is fairly limited and there's probably many more AI overviews that we could go and analyze. But at least from our analysis, uh, were you able to determine how many sources generally appeared in an AI overview?
Yeah, so we found that on average, there were 479 sources per AI overview. Um, without clicking show more.
479. It's my lucky number. Um, you know, speaking of being included as a source, um, you know, in an interview with Barry Schwartz at search engine land, uh, Google's Hema Buddha, Jaru, Google senior director of product search generative experience.
I told Barry that the link cards within an AI overview generate a higher click through rate than traditional web search results. Uh, she went on to compare the click through rate increases from AI overviews to the CTR increases that a website owner will often see from featured snippet placements. Uh, you know, Like, I know it's hard to listen to anything that Google has to say these days, uh, but I'm going to ask you, do you think that AI overviews are actually going to send more traffic to publishers?
Maybe in the beginning. It'll be interesting to see how it works over time. I think in the beginning, people are always. more willing to engage with it. Um, but I think long term it might be less. I also don't know if they're gonna stick around because I don't think the experience is as good as Google's like current experience.
Um, I also don't know that answer from them was so vague and guarded. It's hard to take anything from it. Does it mean click through rate like In terms of like any of the sources, the top source, like versus the feature snippet, um, how does like the difference in positions where the overall click through rate.
So it's hard to take anything from that because it was a very vague answer, but overall, I don't know, it'd be interesting to see how. The more I think the top results would probably get more clicks overall, but I could see it becoming very winner take all like you want to be, you know, top three, top four.
Yeah, you know, Kevin Indig, uh, a guest of this pod, uh, did, did a little analysis here too. And, uh, it was back in may and it was also a very limited. Sample size, but he found that when AI overviews were present, it caused an 8. 9 percent drop in traffic, uh, from those queries. And that was, you know, about what I was expecting.
I think it's within the range of what I was expecting. Uh, you know, had it been like a 40 percent drop off, I think we all might be questioning, uh, is SEO dead? That's a question I'm going to ask you later. But, um, you know, Kevin. I recently wrote that, uh, he sees AI overviews dropping in prevalence. Uh, he did an analysis from June, um, in June, I should say, uh, versus the analysis he did in May.
And he found that, uh, the prevalence of AI overviews, it's dropped by about 50 to 60%, uh, in that time or, or how often they were appearing in search results. Um, I, I know you mentioned just. Earlier that you think Google might roll back AI overviews or, or change them dramatically from here. Do you think that like AI overviews are going to be more or less prevalent in the SERPs and let's say six months from now?
Yeah. I mean, this is all a hunch, but I just don't think the user experience is as good. That's kind of what it boils down to for me. I don't think they're that helpful most of the time. That could change. But I don't think the current iteration is going to stick. I think if they're going to do something, it's, it's, it'll have to be dramatically different if they want to actually improve, um, like the core product.
Brian Dean on this podcast speculated that he, well, I should say he wasn't that worried about AI overviews because he, um, he likened it to mobile to say that, like, you know, everyone was very worried about mobile, like in, in. Dropping click through rate, uh, on a mobile phone versus a desktop search, but ultimately like mobile was great for search because a lot more people started searching and even though click through rate dropped slightly, like there were many more searches.
And so traffic actually went up as a result. And he, he likened the rollout of SGA overviews to that release in mobile, basically saying, um, that. That AI overviews are going to be so great that people are going to search so much more. And so even if there's a drop in traffic, or I should say click through rate, there's going to be actually an increase in traffic.
Would you buy that theory? Do you think these, uh, AI overviews, maybe not today, but in the future are going to be so good that it'll actually lead to an increase in the number of searches happening?
Yeah, I can see that. I think if they're, depending on how Google rolls them out, they should be more interactive in a sense where it's easier to continually generate.
I also just think like Google still said the vast majority of internet traffic referrals are still from Google like it's and it's not close. So, you know, there's always going to be opportunity and information retrieval. Um, that's kind of what it boils down to for me. Like there's people are always going to be looking to find things on the internet.
Um, and. That's probably always going to be the best way to get referrals to your website. And like, obviously it's kind of information retrieval has changed dramatically over the years, but there's always going to be opportunities.
Well, Matt Lenhard, information retrieval is a wonderful name, uh, for a racehorse.
Uh, and I just checked as we were talking and, uh, it's available. So for any of our listeners who are into horse racing, uh, like me, uh, you can name your next information retrieval. That is. That is a triple crown winning horse race name right there. Now, earlier on this podcast, you mentioned that it's very hard for SEO tools to get data on AI overviews because.
They're not appearing for non logged in users. Uh, do you think that, like, at some point, Google is going to start displaying AI overviews for non logged in users? Is it only a Matt Lenharder of time?
Yeah, I mean, that's the hunch. I guess The counter argument that maybe they won't is because it's too expensive to generate them.
That's like a theory I've seen people toss around, but Google's printing so much cash, I don't really buy that, and they've never been worried about losing money. And I think it'll be interesting to see how it interplays with ads. I wonder if that's like another reasoning, like does this affect ad click through rates?
How long till we see ads on top of, or like in an AI overview? So I think there's probably like a couple of factors at play there, but my hunch is not anytime soon.
Yeah, that's a really good point. Like if Google, and they've said they will start rolling out, um, you AI overviews. They, I don't think they've exactly said how or what it's going to look like, but if that happens, then yeah, I think they'd want to.
Uh, roll it out to non logged in users if, if there is like an uplift in, uh, how they monetize search. But yeah, Google has a, uh, has a war chest, so they have plenty of time to get this figured out. Um, I want to transition a little bit, you know, it's been a pretty hectic year. Dick 12 months and search, uh, you know, you and I have got friends whose websites haven't done as well over this last year, we've got some friends whose sites seem to be doing quite well.
So highs and lows for a lot of folks in SEO world over this last year, there's been a lot of change. Uh, I just have a few, uh, things I want to ask you, uh, backlinks. What do you think? backlinks are more or less important than they were when we first started working on SEO, let's say 10 years ago,
I knew you were going to ask this.
All right, backlinks are so important that they don't Matt Lenharder in the sense that, and this is something I actually got from you, you, I mean, you can look at the search and see for any competitive term in specific verticals. I mean, the same five sites rank like U. S. News, Forbes, and like you talked to people who work at some of these bigger sites and SEO for them is, oh, we just write an article, publish it, and we're first.
They don't do SEO, they just publish an article and it's first. So I do think backlinks Matt Lenharder. They might be a proxy for other things. So sites that have more RDS backlinks generally are going to be more popular sites. So maybe backlinks are just a proxy for how popular your website is. Um, but the reason why they don't Matt Lenharder is because if you're in a competitive vertical, you're never going to be able to build backlinks fast enough for them to map.
Like everybody else already has too much of a head start. On it, like compared to you, like if you're in finance, you're never going to catch up to NerdWallet, you're never going to catch up to U. S. News. So, in that sense, like, they don't Matt Lenharder, but in a bigger sense, they do. I am sort of curious, like, how much are backlinks just a proxy for other site popularity metrics?
Or just like how well known that site is, which impacts things like click through rate. Um, so I kind of gave no answer there. They do Matt Lenharder, but they Matt Lenharder so much they don't Matt Lenharder.
Yeah. We'll say you were to start like the next generation nerd wallet. And the content that you create is twice as good as nerd wallets, and it would dramatically improve the quality of the experience for a searcher if that content were to appear highly, but in a category like personal finance, like a brand new website, even if their content's like twice as good as a nerd wallet, uh, could never rank if backlinks were like the only ranking factor.
Um, and as a result, the quality for Google searchers wouldn't necessarily be as good. So it feels like backlinks Matt Lenharder, but then if they aren't. If it was the only thing that Matt Lenhardered, then search would get stale and there would be like, then no new opportunity for an improved quality of experience for a searcher.
Is that kind of what you were getting at a little bit?
Yeah, I think that that plays on like, they're almost like a proxy for how popular a site, how like, much people like a site as well. I like it. You know, people click on their wall because they know it. People click on U. S. News. Like my mom types in best credit cards and she sees US News. She's going to go, Oh, I know U. S. News. Like she trusts them, right? Because they've been around for a while. They've built up this credibility. And so. They also have a ton of backlinks, but I think there's like something else to be said just about like the overall popularity and trust that that's like their sites have built over time.
Um, you know, people see the nerd wallet TV ads and when they later go and search, they're probably more likely to click on that because they've seen them before. And so I think like just, it's, it's almost like correlation at this point, we're like, The sites that have a ton of backlinks are just also more authoritative, but not necessarily because of the backlinks.
They're more authoritative in general, like NerdWallet's public and traded company, like, they're just more well known than, like, maybe the newer up and coming sites. And so, I think backlinks Matt Lenharder. They're not all that Matt Lenharders, but I think they're, they're also like a good proxy for the things that I'm on.
Right. Right. You know, thinking back, like back when you and I started our first company together, you know, we were ranking in the Serbs. We were, you know, doing just fine. And then all of a sudden, like one of the new kind of competitors, uh, you know, jumped us in the Serbs, despite having, you know, much lower funding.
36:14 Page authority to any given page. But the thing was, is they were running a lot of TV ads and radio ads and like overnight, like their brand became a very popular one. And so even though they had like lower domain authority or page authority to a specific page, like, like to your point of when someone is searching for best credit cards and they've heard of nerd wallet before, like they're probably going to be more likely to click on that, that SERP as a result.
Yeah. And, uh, Can I add on to my previous answer? I don't think page authority or any type of like backlinks to a page Matt Lenharders much at all. I do think that there's some type of overall domain popularity, um, metric that must be being used just looking at SERPs, you know, anytime I Google a popular keyword, it's all 10 or sites that I'm very well familiar with.
Very rarely see a site for the first time in Google search. And on your second point, There was actually, I believe I read something about this from the head of SEO at Wayfair where they ran like a mini experiment where they bought TV ads encouraging people to search for like a specific term and then click them and they saw positive impacts and ranking them.
Because of that, I might be butchering a little bit, but that's kind of like a gist of it is that there is an impact like a initial impact between click theory and, uh, your rankings.
Right. I remember you mentioning that to me. And if we can find that URL, uh, breaking down that, that, that way for study analysis, we'll, we'll include it in the show notes.
Um, but you know, You, I want to double click on something. You said backlinks to a specific page doesn't Matt Lenharder. So you don't believe that we need to be building backlinks to any given URL on our site. It doesn't Matt Lenharder.
No, I think it's, there's better uses of your time.
So speaking of uses of time, what's one thing that SEO teams spend time on that is just a total waste of time.
I think sometimes people can get over caught up in the details in regards to technical SEO, you know, they'll, they'll over optimize these super small things that don't really make a difference. So like, you'll see, um, we actually took this into account when we built our audit tools, we tried to pick things that like really Matt Lenhardered, like this patient is in the title tag, that's not good, he should be in the title tag.
But I think oftentimes. People over optimize like super, super minute technical SEO details, like they'll spend, you know, way too much time maybe perfectly writing this like alt text and like that isn't going to make a huge difference at the end of the day. If you would have taken that time and thought, how do I, make this article better?
Is there like a widget I can add to it? Am I missing something that the other articles do cover? You know, is there any piece of data I could introduce to it that like nobody else has? That stuff's going to make a way bigger difference than your image alt text.
Totally. As, uh, as JH said, uh, very early on in this podcast, you got to let the, uh, little fires burn, uh, and focus on the stuff that actually Matt Lenharders.
And, uh, we'll move the needle. Um, you know, Matt Lenhard, this has been fun. I really enjoyed getting to know you on this podcast episode. Uh, if it's okay with you, uh, we I've got like, I don't know, five or six different lightning round questions. I'd love to ask you, does that sound good? Yeah, let's do it. All right.
So we're going to do an over under, um, according to bank of America, um, Google's global market share of search is 90. 8%. Um, as of May, 2024, do you think Google's global market share of search is going to be? Over or under 90. 8 percent in May of 2025
Under, but not by much
Search GPT. I recently heard that search GPT is, is going to be rolling out.
Do you think that your mom is going to use search GPT anytime in the next two years? There's no, there's no chance. What's your favorite restaurant in Charleston? Little Jack's, that's our dirty spot. Yeah, so we will include a link to Little Jack's in the show notes. It's a fantastic spot for burgers for all our listeners next time you're in Charleston.
And also let Matt Lenhard and I know next time you're in Charleston. Matt Lenhard, I'm going to give you a would you rather. We've done an over under, let's do a would you rather. Let's say you've got a hundred grand and you have to either invest it in Perplexity AI, At a 3 billion valuation. I think that's the most recent reported valuation from some new fundraising round they're doing.
So 3 billion in perplexity or, uh, in Google at a, uh, 2. 06 trillion dollar valuation, and you can't touch it for 10 years. Would you rather put that a hundred grand in perplexity or Google at today's market prices?
I mean, if Perplexity raised it to three billion, that's insane to me. Um, so yeah, Google, and it's not close.
Sorry to the Perplexity team. Just, I don't know. They've got a lot of competition from a lot of really, really smart players who, I thought, get a lot more money than they ever will.
Yeah, you know, if anyone from Perplexity is listening, you know, come on the pod. Let's talk about it. Um, Matt Lenhard, is SEO dead?
There's always going to be information retrieval optimization. IRO. Yeah, I mean, Google still sends the vast majority of referral traffic. Search engines. GPT, whatever engines, whatever you want to call them. There's always going to be a way to generate referral traffic for that. I mean, people are looking for stuff on the internet.
No, there's going to be a way to optimize for that.
For sure. Well, Matt Lenhard, thanks for coming on this week's episode of the optimized podcast. This has been so much fun. Is there anything else you want to say to our listeners?
Go check out positional. com. I mean, if you're. Listening to this podcast, you've, uh, probably checked it out before, but, um, and I'm super excited about what we've built there.
And we've got some cool tools coming out in the future, but I'm really excited to share with everybody.
Yeah, we'll include a link to Positional, uh, on our website where these show notes will exist. Um, and, uh, and we'll also include a link to Matt Lenhard's LinkedIn profile, uh, if you want to connect with him. Um, and, uh, yeah, it's a wrap.
Thanks, Matt Lenhard, for coming on.
And that's a wrap. And I just want to thank our sponsor positional. They've got what is a pretty awesome tool set for content marketers and SEOs. They've even got a couple of tools for social media. Uh, we'd love for you to check out positional and the tool set that we've created, uh, over the last 15 months at positional.
And you can always reach out to me. If you have any questions, my email is Nate Matherson at positional. com. Whether you've got questions about our tool set or comments and complaints, uh, or even positive feedback about this podcast, I'd love to hear from you. Um, and don't forget to hit the like and subscribe button, uh, wherever you are listening to this podcast.
Uh, thanks so much for tuning in.
Join hundreds of leading SEO and content marketing teams who trust Positional to enhance their content strategy, provide comprehensive analytics, and highlight data-driven optimizations for their channel.
Start Free TrialPositional's tools are an essential supplement to any search-driven content effort. They help us save time and produce better content for both our company blog and our clients.
We used to create outlines for our posts, either by paying a consultant $75+ each, or by spending 1-2 hours researching and creating each one ourselves. With Positional, we can create the best outlines for our target keyword clusters and get alternatives within a couple clicks.
Positional has proven indispensable in our SEO strategy. Its rapid optimization capabilities for our blogs led to noticeable improvements in search rankings within a month. From planning to making our content better, it’s like having a teammate. Our team loves it!
Positional is a must-use tool for any growing startup that cares about SEO. It's simple and easy to use but as powerful as anything out there. Plus their customer support is next level.
Positional's tools are an essential supplement to any search-driven content effort. They help us save time and produce better content for both our company blog and our clients.
We used to create outlines for our posts, either by paying a consultant $75+ each, or by spending 1-2 hours researching and creating each one ourselves. With Positional, we can create the best outlines for our target keyword clusters and get alternatives within a couple clicks.
Positional has proven indispensable in our SEO strategy. Its rapid optimization capabilities for our blogs led to noticeable improvements in search rankings within a month. From planning to making our content better, it’s like having a teammate. Our team loves it!
Positional is a must-use tool for any growing startup that cares about SEO. It's simple and easy to use but as powerful as anything out there. Plus their customer support is next level.
Positional has been an amazing addition to our SEO and Content team's workflows, enhancing our overall efficiency. We particularly love using AutoDetect and Internal Linking on a daily basis!
Nate and the positional team are the best of the best in SEO, content marketing, and helping you grow your organic traffic. The combination of their expertise and the SEO and content tool they’ve built has allowed us to build a scalable content engine. Reach out to me anytime for a testimony. They are truly phenomenal.
As an SEO novice, Positional makes it easy. I can quickly go from keyword research, to clustering, to content outlines, then go focus on just making good content. I felt like it helped bridge the gaps between what would’ve taken 3 or more tools in the past.
The first time we used Positional's toolset was to revamp an older but important piece of content. We used Optimize for optimization, and Internals for internal linking suggestions. We went from position #6 to #1 with the changes and increased our organic search traffic to the page by 400%. Today, Positional is an integral part of our blogging strategy, from topic generation to blog renovation.
Positional has been an amazing addition to our SEO and Content team's workflows, enhancing our overall efficiency. We particularly love using AutoDetect and Internal Linking on a daily basis!
Nate and the positional team are the best of the best in SEO, content marketing, and helping you grow your organic traffic. The combination of their expertise and the SEO and content tool they’ve built has allowed us to build a scalable content engine. Reach out to me anytime for a testimony. They are truly phenomenal.
As an SEO novice, Positional makes it easy. I can quickly go from keyword research, to clustering, to content outlines, then go focus on just making good content. I felt like it helped bridge the gaps between what would’ve taken 3 or more tools in the past.
The first time we used Positional's toolset was to revamp an older but important piece of content. We used Optimize for optimization, and Internals for internal linking suggestions. We went from position #6 to #1 with the changes and increased our organic search traffic to the page by 400%. Today, Positional is an integral part of our blogging strategy, from topic generation to blog renovation.
“We’ve been moving up the search rankings. When we first started using Positional, we had about 1,000 visitors from organic search per month, and today, we have over 12,000 visitors from organic search per month. And obviously, Positional has played a large role in our growth.”
Positional takes the guessing game out of our content and SEO strategy. It allows me to do extremely quick keyword research which I can then turn into detailed instructions for our content writers through their Optimize tool. I love the speed new capabilities are being added!
I've been using Positional since its closed beta, and it boosted our SEO results so far! We've published over 80 articles with Positional and it has gained traction very well. The "Optimize" tool is my favorite — it ensures we use the right keywords for better rankings. The "Content Analytics" tool is also great for showing us exactly where we should improve our content.
Positional's tools are an essential supplement to any search-driven content effort. They help us save time and produce better content for both our company blog and our clients.
“We’ve been moving up the search rankings. When we first started using Positional, we had about 1,000 visitors from organic search per month, and today, we have over 12,000 visitors from organic search per month. And obviously, Positional has played a large role in our growth.”
Positional takes the guessing game out of our content and SEO strategy. It allows me to do extremely quick keyword research which I can then turn into detailed instructions for our content writers through their Optimize tool. I love the speed new capabilities are being added!
I've been using Positional since its closed beta, and it boosted our SEO results so far! We've published over 80 articles with Positional and it has gained traction very well. The "Optimize" tool is my favorite — it ensures we use the right keywords for better rankings. The "Content Analytics" tool is also great for showing us exactly where we should improve our content.
Positional's tools are an essential supplement to any search-driven content effort. They help us save time and produce better content for both our company blog and our clients.
