Episode Transcript
[00:00:01] Speaker A: Hello, I'm Jeremy Rivera, your unscripted SEO podcast host. I'm here with Grant Simmons, who's a legend in his own right. But for those who don't know what you have done, where you started, give us a sense of your history and tell people why they should trust you. As an SEO, that's very good.
[00:00:19] Speaker B: I'm a legend in my own lunchtime, Jeremy.
Yeah, so it's funny, you know, I have a kind of winding road. I, I was in marketing. I was a programmer developer in the 80s. So I'm an old guy.
Then got into some marketing when I was in New Zealand, Australia. I'm originally from the UK and then when I moved to the US in the early 90s, I worked at a marketing agency.
You know, essentially people were asking for websites and how to show up.
So I built websites for people from their agency and then when they wanted to show up, our first real client was the city of Santa Monica who wanted to show up for Santa Monica.com, you know, hotels and restaurants. So I learned everything I could. I joined the LA SEO community at that point. I jumped onto the message boards back in that point. I mean this is mid-90s.
Even built a pre YouTube video sharing website and learned how to rank that and stuff. So I've been in there a long time.
Agency side, brand side, back to agency consultant, fractional.
So yeah, you can trust me because I have a really cool accent. No one knows actually where I'm from. And number two, uh, I think I'll misquote farmers. I, I've seen it all, so I know it all. Uh, maybe I've seen a lot of it, so I know a lot of it.
[00:01:57] Speaker A: That's kind of the, the, the dichotomy because I started 2007 so I'm not quite that aged or that mature in it, but I have seen a lot. I'm curious, what have you seen recently where you're just kind of floored? What's new? Like genuinely new?
[00:02:15] Speaker B: You know, there's new and old and I hate to hog back on. You know, GEO is just great SEO. I'm not a Google shill or anything, but I still see like I just saw a site that essentially had a site map that was full of escaped URLs. I know. And so people look at these things and they say I want to show up for this, but they haven't often done the fundamentals. So the things that often surprise me is people shooting for the moon without a rocket ship. You know, they're basically in Their Hyundai, you know, trying to driving off the
[00:02:52] Speaker A: cliff as fast as they can. I'm going to the moon.
[00:02:57] Speaker B: It's, it's like that classic scene from Toy Story where he leaps off there to infinity and beyond and just crashes to the floor. I think a lot of people are shooting for great citations or visibility in Geo and really not getting there. What's excited me recently, one of my many hats I wear is as an ambassador for Inlinks and Waycay.
And so Waycat just launched unofficially through Dixon Jones and Fred Laurel Entity Map as hopefully generating a standard. And that's where I've kind of focused my last five, six years probably around entity SEO.
[00:03:38] Speaker A: Yeah.
[00:03:38] Speaker B: And so that's exciting to me is that people are actually realizing that clarifying and, and teaching, you know, or treating Google and Lams like small kids. We have to actually explain who you are and what you do is really where I'm excited about these days.
[00:03:56] Speaker A: It's like blind 5 year old is AJ Cohn's blog, you know, very seminal. If anyone hasn't read his articles, he comes out with a banger every year or two that's just like it's goog enough, you know, like he'll, he'll hit them.
I want to go back to the entity optimization and have you dig a little bit more into some of your process around that, some of the fundamentals, because I talked with Bernard of Calicube about entity optimization before it came up with Alexander Mayerhan's Getme links as we kind of dug into it. But I'm curious, what are some of your fundamentals in that process that you do yourself, that you assign to somebody, or maybe you're even, you know, building stuff in Claude or in LLM itself to try to use that. And tell me a little bit, tell the folk a little bit about that entity map concept and why that stands out to you.
[00:04:56] Speaker B: Wow, many questions there. First off, Jason Bernard, smart guy, Kalicube. It's really cool.
His fundamental thought is create an entity home for whatever entity you want to show up for and make sure that that's reinforced. And that really leads into the whole idea of, you know, what is entity SEO? For those folks that might not have just dabbled or have heard the term, it really comes down to how do you clarify the content on your site?
Leveraging known information points in this large graph we call the web.
And by information points, it could be something definitively stated within a catalog like a Wikipedia page.
It could also be the NHS drug list or it could be something even LinkedIn as an entity catalog of individuals and companies.
So these are all ways of finding these divinity sources that the entity has I'll say a point in the catalog. Generally it has a number in the catalog and really connecting your site concepts to those known concepts, those well defined concepts. So the first thing is when you look at a page is is it tight around distinct entity topics? So is there what I call query drift?
And yeah, I did create a tool with my friend Richard Wong called QueryDrift which essentially analyzes a page fundamentally at the entity level and says based on the queries that map to that page. So looking at Google outcome as a good indicator of what they see your page is about and really seeing whether a page has outlier topics that probably should deserve their own page or a subpage or whether it's purely focused in on the main key topics that page is about. And topics entities not interchangeable but I use them interchangeably.
That idea of clarifying a page, connecting it to known graph through same as schema to Wikipedia, things like that and then really making sure there's focus on a page.
So one entity per page, ideally with you know, mentions. So clarifying disambiguation entities within a page that support that main entity. If that sound very geeky, I'm sure it wasn't that geeky. But you know, I try and simplify these concepts. It goes deeper than that. So I answered I think three of your questions there. Yeah, the questions did you have?
[00:07:48] Speaker A: Well, I'm curious how you're seeing the connection to the entity map and that kind of ties into, you know, I'll throw it answer that and I'll throw in the bonus question of can you dig into a little bit of how you think that that knowledge layer in LLMs is the same or different from how Google approaches that knowledge layer.
[00:08:11] Speaker B: Yeah, each one is different and it's generally based on their grounding. So what they use for grounding as far as Google Gemini obviously uses Google Index and Google Search and same way ChatGPT uses Bing and some of the other large LLMs are obviously using different grounding things. So where there is an existing graph for Bing and Google, those are fundamentally different potential results.
I'll tie this back to why entity graph as a standard makes sense. And there's some other articles that are going to be pushed out through AK, but entitymap.org is the actual spec.
And so what is that for? Why is it important? And I think the key thing here is what do we see in LLMs that is problematic now the first thing we see is hallucinations. And hallucinations where they will make stuff up is generally because we haven't filled the gaps in on our site itself to be able to explain fully, clearly and ambiguously who we are, what we do, who we serve, areas we serve. These are things that traditionally you create a page about, but that doesn't mean it's well optimized for an entity. So understand even connecting to a graph. So what entity map does is essentially creates from key pages you can define a kind of summary. So it's not a summary as in this is what the page is about. It's a summary with associations, connections, disambiguation and that clarification layer, or I call it a meaning layer as opposed to just saying, you know, it's an entity layer, but it's really about meaning and understanding and those connections. So that's where I see it as valuable. The proof is going to be in the pudding. You know, there's limited deployment so far, but it's good to see some people test. It was only announced last week, but I do think it's. It fills that gap of LLMs and treating them like blind 5 year olds because they are at that nascent stage where they do rely on entities and obviously Google says they rely on it from an entity standpoint.
You know, Dwayne Forrester, when he was at Bing and beyond is talking about an entity layer, a meaning layer, an understanding layer. So it makes logical sense.
[00:10:50] Speaker A: Yeah.
[00:10:51] Speaker B: Whether we can prove the point that it makes a difference in what you show up for, not whether you show up, but what you show up for and how your brand is represented? I think that's really key.
[00:11:03] Speaker A: How is this a distinction from the step forward, step back of Google, you know, pulling the rug out on fact, SERP results and people taking that to the next step of well, schema. That means schema doesn't matter, you shouldn't do it.
I obviously am throwing doubt on that case. I think it does matter. I think there's plenty of.
There's varying levels of evidence, but we do know it is a data point entry. So how do you view schema in this moment? Is it something to add in? Not as more of a backing semantic step to help align and add those same as references on your about page to help with that entity optimization as much as topically direct, you know, and clarify on your service page, like what are the type of prompts that we should be feeding into our cloud to come up with you know, schema for an X page that we just as a good habit versus like oh, I'm going like instead of the spec of hey client, we want a thousand dollars to add fact schema everywhere.
[00:12:20] Speaker B: Yeah, that's a good question and a question that will obviously divide everyone listening to us. So that's good. I do like that.
I think it is unambiguous to say that people believe generally across the SEO community that schema is good for rich results. Is that fair?
[00:12:48] Speaker A: I'd say that's an accurate representation of the general take on it at the moment.
[00:12:53] Speaker B: Great.
And then there are folks that believe that having amazingly nested schema down to the small concept within a page is really necessary for showing up better in search results. I don't agree with that.
I think there is a limit to return on investment in schema. But I do think that as I mentioned, there are grounding indexes that LLMs use and whether they strip out schema within a result before they can put it into rag resurfacing. Yeah, I think that schema is leveraged by Google in ontology, categorization, insects, charge and everything else. And so it is important when a grounded source is using schema for categorization and disambiguation, I think it's important to use schema. And so Google AI views are still the most prominent AI overviews compared to any other thing. So why wouldn't you, if you're not doing the large lift of absolutely finding every single concept you can do there, but if you're doing it at that mid surface, why not?
[00:14:24] Speaker A: Yeah, I think at the minimum it is a why not of like Google hasn't said, oh, I have no appetite for knowing more semantic information. Breakdown of what this is because. Because it's difficult to spam schema markup because you're just saying in very detailed specific markup language exactly what's there. So you know, if you have the content and you're saying what it is, it isn't something that I could see that they wouldn't have some sort of appetite for on the back end.
So as far as workflow, yeah, let's do it.
[00:15:03] Speaker B: And if not today, if not today, there's going to be something that LLMs are going to leverage to understand page content better.
I mean, if not today.
Mark Cook and I got into this, you know, not in a nice way is schema useful for LLMs? And we're saying no, but it's good for Google, so why not?
[00:15:27] Speaker A: Well, if it's good for Google, then it is good for LLMs because ignoring the fact that for the large part, rag in LLMs is a wrapper for Google results.
[00:15:40] Speaker B: That's right. I think. Yeah. Reinforce my point and put it much nicer, more eloquently. Yeah, but my thing is, if not today, then what is going to be an ontology that LLMs are going to use for clarification, classification, everything else. So yeah, today, no, tomorrow. Why not?
And if FAQ schema has now gone away, does that mean you shouldn't use FAQ schema?
I would. Because it's a very low lift. Most tools automatically add it. You're not going to go through and take it out.
[00:16:17] Speaker A: Yeah. You're not going to remove it. And I don't see the harm. I don't see any negative indicators. So on the back end, because I think of, you know, the training data set and then the Google results in the fan out, you know, if it's not getting it here, it's going to try to get it from there. So strength in your hands. Makes sense to me.
[00:16:40] Speaker B: Yeah, I mean we can get onto pagination and whether you should self canonicalize paginated pages. We can go into H1s, whether they're important or not, whether multiple H1s are a problem. We can just go across the whole gamut of these.
You know, let's divide the search industry 100%.
[00:16:57] Speaker A: Yeah. Let's go back to the subdomain argument. Do L LLMs parse and treat subdomains equally? Let's rehash all of the questions and arguments that we've had in Google and just put it through the LLM question factor.
[00:17:11] Speaker B: Yeah, we should do LLM.domain.com because that's how you.
[00:17:18] Speaker A: Well, I mean we had our moment.
I saw it was a year and a half ago. We had white text on a white background to rank in LLMs.
[00:17:27] Speaker B: So yeah, what goes around comes around.
I must admit that when I created my first website, I think it was 1995. 94.
Yes. I was so happy when I showed up in a Yahoo directory, you know, when I showed up in Ask Jeeves at some point. I mean those, those were the days.
Stuffing and hacks. And now we've come back around that we're looking at stuffing and hacks. So just different surfaces.
[00:18:00] Speaker A: Yeah. I was on the site the other day and I got a pop up and then I had a banner bar at the top. I'm like, this is 2000 all over again though it is old as new, but directionally as an industry, I was seeing somebody saying that while it is useful to clarify that many of the things that are in the stack of optimizing for these LLM agents, are we shooting ourselves in the foot by trying to say, oh, we'll handle it in the same budget as an SEO and saying, hey, it's not new, it's just SEO and like, oh, well, I was just giving you 2,000amonth for SEO anyway, so you got it handled.
Are we shooting ourselves in our foot by disavowing the unique challenges of LLM and citation and brand and visibility?
By just claiming dominion over it and not giving ourselves the chance to say, yeah, it is, I probably need some more budget to also conquer this.
[00:19:12] Speaker B: Well, I would just say that sounds more like a client relationship problem than an industry problem. Okay, so, yeah, I think that if I'm doing SEO for someone, then they ask me to do YouTube optimization, then that's an additional budget if I'm taking up my budget to do the SEO. So, I mean, if I've got to that point, as most SEO engagements get to after three, four, five years, you know, what are you doing for them now that you can demonstrate success? So sometimes that agency relationship has to, has to change a little bit over time. I think as you introduce new services, there should be an additional cost, but I don't think you have to create a whole new, hey, I'm doing paid search now. You know, in the same way that paid and paid social are generally under one umbrella, but two different budgets. I think it can be set up like that by saying to a client, look, we need to still create links for our SEO campaigns and the same way we need to create links that are more focused on citable sources or key sources. So yeah, we're going to need more budget for that. And I'm just using link building because generally that's an easy way of people saying there's a cost to it.
But when it comes to creating content, you should be creating content for your audience, not redefining the audience as a bot or. Because that's not what we've done as SEOs, if we've done it right and not redefining now for a different bot. And we're writing for that. So I do think, and once again, I'm not unoutspoken. I just generally keep some opinions to myself. But I think the difference between people that say it's a completely new discipline that needs its own budget line, everything else, and people that sign it's part of SEO, but it still needs a budget different for the people that saying it's SEO and we're going to do it from the same budget. I mean, I said client problem.
But that being said, I think the definition of good SEO has been refined by people that say I wasn't doing that and now I am for LLMs.
So I do think entity SEO, which I wrote my first article about strings. Strings, not things or things not strings in 2013 when Hummingbird was being pushed out. And so these are things that aren't new concepts.
[00:21:49] Speaker A: Yes.
[00:21:49] Speaker B: But suddenly people are focusing on again. I think it's the same with some of this. The idea of chunking or self contained paragraphs. You should be writing like passage rank in.
[00:22:00] Speaker A: We were talking about that.
[00:22:03] Speaker B: I was just going off on a tirade there.
[00:22:05] Speaker A: No, no, definitely. I just mean like in the industry it's like, you know, like, oh, it's new, it's chunking. No, that's just a different phrase. We talked about Passage Rankin when Bert came out. Right.
[00:22:19] Speaker B: Well, it's the same thing. I mean, you know, the idea of saying self contained paragraph is just saying. Trying to avoid query drift within a page or having distinct topics within the page that are semantically, hierarchically set correctly.
[00:22:34] Speaker A: H1, H2, H3.
How old is H1 structure?
[00:22:40] Speaker B: I know, yeah. I mean I do think, look, there are differences. I think that digital PR should be. Those guys should be laughing all the way to the bank.
[00:22:51] Speaker A: Yeah, no, absolutely. They have. Absolutely.
[00:22:55] Speaker B: And so that's where a lot of people, you know, you can ask for additional budget.
[00:23:00] Speaker A: I think, I think it, I think it's because we got, I think we got. I call it the golden age of Google where for SEO where from 2010 to 2020 there was so much value that you could get just by doing some of the fundamentals and doing it would send you so much traffic with so few caveats of as long as you weren't a blatant really bad spammer for most organizations you could get your ROI for your SEO budget fairly easily.
But now there's more landmines.
What they obfuscated of the ABC of content links and click behavior on the SERPs, which they denied all through 2010-2020.
That's not a ranking factor. Oh, okay. It's not a ranking factor. It's part of a ranking system. But the documentation in the COR course proves that it's the C. It's. It's 30% of how you are deciding this is what queries are clicking, how long and that whether they're PO going back and anybody that had said that in the in 20 teens was derided and, and mocked and it was true. The whole thing was true.
So I feel like a vindicated tin hat theorist, but at the same time shows that too much credulity of the voices at Google is not a great thing for our industry. But ignoring them entirely also has its own peril. So where do we stand now with what Google is saying, what Google isn't saying and how they're saying the same thing from both sides of their mouth?
[00:24:44] Speaker B: Well, that's so difficult because I do think that folks like John Mueller, who I know, he's a great guy, I think it's like any company, you're not giving away the milk, you have the cow, you have the cash cow, you're not going to say exactly what it is. And I think that there are caveats to what Google and through the mouth of Danny, Gary and John can actually say, you know, and I think, I think that's, that's challenging for SEOs because you know, there is, listen to what Google says and just do that. There is. What I've been doing for 15, more than 15 years is test and validate, you know, and kind of come up with your own ideas. And also 2013, 2014, I was writing about engagement as being a key component to this engagement in the SERP engagement on site because I couldn't see any other way that there were some things that manifested that couldn't be based. I had a crappy site that was ranking brilliantly and the only way I saw that justification was I was getting 20,000 people a month to the site who were taking very similar paths through the site.
And I just said there was consistent engagement within certain paths. You know, I call these paths to satisfaction that I'm sure that Google through Chrome can see this idea of connecting, well connected pages, connected paths that people are actually taking and then there are consistent signals that either the pogo sticking or dwell time or whatever you want to call it within the SERP that have to be taken into consideration. How, how else would they do, you know, a hundred SERP tests a day and not be able to glean data from that, from interaction? I mean, what, what are they going to do?
So yeah, I think the same thing will happen and sorry, I don't mean to, to pontificate the same thing will happen. What we see now today, I'm sure Google is not telling us the whole truth, but that doesn't mean, well, one, they actually know the whole Truth. I've got this in a deck that I'm doing this week at SMX in Boston which is really around Google doesn't know how it actually makes decisions. The algorithm, they know how it works fundamentally.
But the results are a matter of machine learning and you know, and predictive analytics and everything else.
So I think we will learn more by seeing outcomes than we can by listening to Google.
[00:27:42] Speaker A: I think tied into that is something I was about to bring up which is that rest in peace, Bill Slosky, he did a lot of analysis of the patents that are fundamental in the systems behind Google and to use it they must have a patent. But I'm seeing something on LLMs that isn't the same. I don't think they know what they're doing at Anthropic and they're not running the same patent game because it's to break fast, to move fast, break, break stuff. I'm not seeing the same level of visibility or I don't know anybody who's following patents from these LLM tools that's able to publish anything of use that I've seen from the patent analysis slide.
Like we have been able to glean and derive from patent analysis of what Google is doing. Is am I missing someone? Is there someone that is analyzing patents to try to help bring any clarity to how these LLM tools are working or am I right? That drifted the like. It also follows my other theory, and you can comment on this of like we have become a beta testing culture and we have accepted the launch of complete and total garbage failure launches of software that will literally kill somebody. You know, if you check diving pressure depths on through LLMs 3 out of 4 if you, if you follow those, they would kill you. You know, obviously putting gasoline in your Mac and cheese is also not healthy. But that's more like an amusing anecdote. But they're, they're more like Donald story. That's right. The things that, you know, I remember there's a huge hubbub in, I think it was 2011 of you know, there was something about the Jewish people being labeled a particular way in a search result and there's been, you know, these fewer moments of these certain results showing up this way. But now we have, it's a beta test culture. We, we just accept it from top to bottom that hey, software releases are now, you know, you can't criticize them because it's just going to play out.
[00:29:59] Speaker B: There's probably two ways of looking at that. First off, all of us that are having fun today are creating tools, are creating products, are doing that. So yes, there is a massive beta tester culture because most of the stuff that's being put out is shite. All right, so that's number one. And most of us know that when we push something out, we're really just after feedback.
I think Dan over at Dajun SEO is the closest you're going to get to someone that's kicking the algorithms and understanding patents.
And Bill, great guy. He and I had some very long conversations about search and yeah, I think Dan is probably the closest to that where he's questioning and obviously pushing out a lot of very cool stuff around LLMs and showing up in LLMs fighting. So I think that's close.
I think Chris Long is tickling a little bit at the edges.
I'm not saying he's at any kind of same level as Dan or any kind of same level as Bill, but.
[00:31:19] Speaker A: They never will be.
[00:31:20] Speaker B: So I think, I think that that's something.
I think it's interesting when you say the beta culture.
Yeah, we are definitely willing to put up as the SEO industry with LLMs that give us subpar answers as insights rather than really working hard at getting the right answers. And there's obviously different levels of SEOs.
I think that creating content in LLMs, definitive content, can be problematic, but there's obviously ways of asking LLMs for more unique rag results.
I mean, so, so I think, I think we are beta testing the beta testers to make it work better based on our experience and our ability to confront LLMs with what we know as obvious ambiguity or obvious hallucination. So I do think there's different levels and you know, I mean we're all in this industry and we see things like yes, we can get you 50 citations in LLM source sites and it only costs you xyz as we saw from a large guys last week. And also we see one click SEO optimize your site. We see this.
And so this is just beta testing through beta testing platforms that aren't ready for primetime. Obviously for me, I leverage tools now. I can do more work now leveraging LLMs. But I don't what I call outsource uncertainty to LLMs. I don't take what they say as the answer I want.
I take it as the answer they give me.
And so yes, we are just in a big petri dish right now. A petri dish, depends what you call American English petri dish.
And people are dropping little droplets on us and going does this work? And we can see after a few days whether culture has grown or not.
[00:33:44] Speaker A: There's a tale of technology, and sometimes the enthusiasm for new tech far outpaces either the general understanding and acceptance of it. I remember Scoble jumping into the shower with his Google glasses, and it's like it was, you know, new technology, augmented reality everywhere.
You know, it's real. And you know, how many billions were wasted on building the Metaverse, which collapsed in on itself. I think we will see a resurgence at some point of some sort of, you know, system like that. But it's not snow crash. You know, we're not all sitting around. It's not ready player one. We're not sitting around with our headphones, although, you know, we do have these in front of our eyes.
[00:34:37] Speaker B: I was just going to mention that
[00:34:39] Speaker A: all of the time. So it's funny because, like, on the one hand, you know, no, we're not constantly staring at screens.
Let me, As I stare at the screen and talk to you and, oh, check this other screen over here. It's not as convenient. It's, It's. I think it's the jetpack system. You know, it's like, oh, you said that we would have jetpacks. Well, you know what we do? In fact, I saw a really cool one of like, they have speeders, like a land speeder, like Star wars in a zip.
This is not AI but we're not zip. You know, we're not zipping around in a jetpack, mostly because we're not suicidal. And I think there's. And there's. I think there is a.
A.
This is too much aspect to the general culture. And I saw it in my gym class the other day.
My wife and I do go to the gym with my wife and she and another girl were talking about, oh, I saw this post from.
From so and so from their shop. So AI that's such AI slop. Like, they obviously made that in GPT. Like, come on. We know. And they're like, you know, her friend was like, basically like a Luddite, you know, like, my wife's much more technically savvy, but they both equated AI not to futurism, not to the enterprise and data and androids and high technology, but to crap.
And it's. The insidification factor is definitely at play of. That's the downside of the beta technology, but it's also the long tail and acceptance because we have, you know, quite frankly, the number of asphalt delivery guys and people that you can call to actually make you an asphalt driveway. Very few of them understand even websites, they barely understand phones.
Try to get the contractor.
[00:36:37] Speaker B: Yeah, we, well we forget we're at the bleeding edge and people like Robert Scoble who you mentioned and once again I'm not name dropping here, I actually bought him in to speak to our folks when I was at Homestead Calm, you know, he was a rating about things he had seen within the Apple campus.
So these are all things that don't even get to us, you know. And when it does it's a Google Glass which, you know, which is, you know, it's kind of more embarrassing than fits in into your pocket or fits on your face, you know. But I do have to have lots of caveats here.
Often companies like Apple produce, you know, devices that we never knew we needed until they come out. So things like flying cars, things like, you know, even self driving cars at scale, these are things that we can, you know, I guess I was going to say prognosticate but that's a very long word. We can think they're going to be here sooner than they are but generally that's an adoption issue, not a technology issue.
Right. And so I wrote an article maybe 7 oh my small now 12, 13 years ago was have you talked to your car today?
This is before Google, Google Play, Apple Play and Google Auto, whatever it is.
But it really was the next screen would be the inside of your car because self driving cars, especially in places like la and I don't know where you live but where there's traffic, yeah, you are stuck in there. You listen right now to talk radio but when you have entertainment that can be projected on the screen in front of you, people are going to do that, hands off driving. You're going to be watching something and then that's the chance for you know what, what I could showing up in the right place, you know, the right message, the right medium, where there is opportunity of attention, you know, SEO is going to move them, might be called, I don't know, Auto Optima, I don't know. But it's still going to be SEO fundamentally which is showing up at the right time, right place, right message in an organic way. That is you're not paying for it.
[00:39:00] Speaker A: I am curious what Your connection to homes.com is because I have a connection too. I was part of Advance Access, a real estate hosting company that was bought up by Dominion along with Homes.com and Agent Advantage when they cornered the market 2012, 25.
So I'm curious when you were part of Homes.
[00:39:20] Speaker B: No Problem? I'll give you the story. How's this?
Homes.com suffered a major devaluation of the site in around 2012, 2013.
So for the first time in their history, they looked outside for a head of SEO.
And I was fortunate enough to be the person selected.
So in 2014, I was hired.
I had about a month before I could jump in, and I was working with Search Metrics was the kind of agency they had built in to work on recovery.
They basically lost 60% of their traffic, organic traffic. And so immediately I came in and these were the days where you could say, hey, that looks like a real algorithmic shift. And I think it's Panda.
All right, so what should we do? We should clean up a lot of crap, a lot of thin pages that existed purely for just existing, you know, places where there was no listings, things like that. So I worked with them seven and a half years.
I was the VP of search, then became VP of all search, then became VP of SEO and then VP of performance marketing when we got acquired by Costar.
And I've been out of that since 2022, where I've been by myself.
That's my tie in there. And the end of the story is I was already scheduled to speak at SMX London.
I went to SMX London, where I was on stage with, I think, if you remember, Marty early from Google, we were talking about at that time, Google Profiles as being, you know, individual profiles, authority, authority, authoritatorship, all that stuff.
[00:41:20] Speaker A: Anyway, yeah, yeah.
[00:41:22] Speaker B: So the first time I met John Mueller, he was there in the audience.
He was an engineer, wasn't really the face of Google. I just said to him, hey, I'm from homes.com. i think we've been hit by. By Panda. This was before I was actually physically working there. I said, I think we're here by Panda.
This is what we're doing to mitigate and kind of remediate, you know, we've got rid of 5 million pages from the index. We got rid of some really SAS internal linking, We've got some fallbacks, We've done this. We've got da, da, da. And he seriously, he got on his laptop, he typed in, and 20 seconds later he goes, yeah, you're under Panda penalty.
And I go, all right, well, they know that stuff. This was back in, I mean, a long time ago, 2014.
Yeah, we made those changes and when Panda was pushed out at that time, the rolling updates, we got back all our traffic and then some and had a brilliant year. But it was a Matter of getting rid of millions of pages from a 110 million page site, taking some really big risks in chopping and search metrics who got some of the way there. They were not thinking to be aggressive. And I've seen in those days recovery had to be aggressive. So long story. To answer your question, that is where I came from before that, homes.com was a client of the search agency which I led the SEO and social product team there. So they were our clients before with Realtor and that and stuff.
[00:43:01] Speaker A: So it's funny because just a random connection because I work.
[00:43:07] Speaker B: We're all connected.
[00:43:08] Speaker A: We're all connected.
[00:43:09] Speaker B: This is our graph. There you go.
[00:43:11] Speaker A: That's an arguable graph.
I'm curious if there's any question, what's in your crop? What are you getting stuck on right now that you're trying to solve? And if I could ask the next SEO, hey, Grant was working on this. He wants a second opinion.
So what are you turning over? What test are you running? What are you looking for more data in? What are you trying to solve right now for yourself, your clients, your projects?
[00:43:45] Speaker B: So I created a few tools that leverage Google search console data and there's a lot of parallel paths on this, you know, like QueryDrift was one of them. And then I've created some cell phone tools that might launch the public at some point. It's a appslicer.com for want of a better thing. But so I'm solving how do we get more use out of Google search query data? And one of those is for example, estimating impact of AI overviews. All right, so taking year over year data, seeing click through rates, seeing length of queries, and coming up with an algorithmic way of saying this has been affected based on traffic position, click through rate and some other metrics, query count for pages, things like that. So doing that is one thing I think is a massive opportunity to keep diving into there where I still think we have massive problems. Last year I started to create a tool with some savvy SEO people that essentially was SEO attribution.
So how do we clarify the value in what we do? And this is equally important for LLM as well, where there's massive over investment, I think in LM traffic based on what you get back. So I think the ROI of SEO based on actual activities, tactics and expectations and getting into that predictive way of saying we know that these kind of updates will make these kind of changes.
Obviously Market Brew is doing some brilliant stuff around essentially recreating search engines and the algorithms to do some of that stuff. But for the average Joe, what, what can we do? And then the next thing is really once again, I think the biggest challenge that most agencies have and most individual consultants have is reporting and raven tools who you add the connection to. You know, started that idea of customized modules, you know, building dashboards that make sense.
I still think it's a primary agency issue where there's no consistency in tying efforts, outcomes back to KPIs for the business. So I think that's where we should be focusing.
Spending time, energy, money and breaking out of a data studio or whatever they want to call it next week looking at stuff.
But that would be where I'm, I'm looking. And then I think the other thing is don't forget Google is driving Gemini and driving AI overviews.
And although there is limited and fairly good studies on the correlation between showing up in organic results and showing up in LLM mentions within a overviews, I'm more worried about when we can't make those correlations with AI mode. And I'm more concerned about people forgetting the fundamentals of information retrieval, information architecture, information organization and starting to focus on the next shiny object.
[00:47:12] Speaker A: I absolutely agree and I will pick the branch of the next person and bring up, you know, reporting and find out what they're doing, what, what their approach is. Because I agree, I think Google's greatest punishment for the SEO industry was GA4 and their second punishment for the industry was Google Data Studio. Like oh here it's better than our worst thing. But have you, have you checked out SEO gets?
Matt Mellinger has done some pretty good stuff.
[00:47:45] Speaker B: I use that all my clients.
[00:47:48] Speaker A: Brilliant. Brilliant.
[00:47:50] Speaker B: Yeah, I mean it's great. And tying into mpcp, tying into being able to query that data is great. I mean those are the kind of talks, but it's still not reporting it's insights. Yeah, I've created lots of narratives that work on top of that and I've worked on a lot of things that try to explain the tasks that need to come out of this stuff. But yeah, it's brilliant. And I think where that brilliance comes in as well for sunlight SEO gets is looking at organic performance in GA4 in a way that's actually useful.
Comparing against key events actually useful.
And layering in the GA4, sorry, GSC data on top of that.
[00:48:47] Speaker A: You know what I would like to see as far as a tool, if I were going to build another one myself, would be a degapinator because I have, ironically I have a Post that says Google Search Console's data isn't accurate and it's ranking and it's getting like you know, 100 clicks a month. Do you know how many queries it tells me detail for and how many clicks?
Five.
It has told me about everything else. Yeah, even through the API or sorry, going to the API, it went from 5 to 12 out of all of those clicks. I know what 12 of them are now. I know what the page is all about.
So I would love a tool that would look at a third party ranking tool, Google Analytics and Google Search Console to rough out, hey, these are probably the, this is probably the missing picture of the keywords that you are actually getting clicks for. And maybe it comes down to like even you know, click timing because now we have the 24 hour period. So if you on the Google Analytics side, you know, you had the search query and it came in, you had third party rank tracking, maybe we can with some confidence say probably, you know, these are the keywords that are actually driving these clicks.
Is it. Because at this point with what, 3% of the data, this is a more accurate representation of the direction that it's going. That's so little data that anything that I'm guessing is not scientifically based. So I'd love a tool that would give, give some sort of secondary measure of. Okay, Ahrefs says you're ranking for these phrases. This is the estimated CTR within this timeframe in Google Analytics. You got a verified organic search that aligns with this click and these are the phrases that get have query detail that Google Earth that GSC says that you got clicks for.
We think this is your actual click map of what you likely got that traffic from. In terms of queries. It's, it's the, it's removing not provided. I would love to.
[00:51:16] Speaker B: Well, we're getting back to the whole idea of, you know, how much guessing are we willing to put up with? You know, most of my tools I add in a confidence score because the math is just math and reasoning is just reasoning. But I think that's, I mean look, I don't think that is a feasible product to get better because you're still dealing with garbage in, garbage out. Now the way I would look at it, and once again, we're not building products here, but the way I would look at it is generally you get a lot more information when you look at the 16 months of Google Search Console data to have an understanding of if the article has been around for that long, all the queries that have mapped not just the most recent queries. I think that's often forgotten.
And there are tools like SEO Crawl, I don't know if you've heard of those guys. It was kind of a precursor to SEO gets for me. I was using them for everything and they were warehousing the data and servicing a lot more queries. Every query show up for every query based on even zero impressions it would show up based on historical data. So I think it's possible when you look at larger time span but with everything's changing to your point earlier, everything's changing so quickly.
I think that's problematic and I think relying on third party scrape data is. Or we have 10 billion queries we're looking at. It's the same way of saying about prompts. Do we really want to get back into the idea of let's track a billion prompts so we can have an understanding of ineffective. My main thing is what happens once someone gets to that page? Are they engaging and taking the next step? Is there a consistent dwell time within that page? That to me would be more effective if we could work out dwell time per query and layer that over search intent and layer that over maybe some reasoning around if I had this search intent, what queries might I use and then looking at whether that expands out to additional content or additional elements within that one piece of content. So I'm more of the nature of it's still outcome based.
[00:53:41] Speaker A: That's fair. And maybe it's the wrong question. Maybe a more useful functional tool would be is where you're funneling in traffic actually going anywhere or are you not optimizing organic.
You're getting traffic in on a page in Google but it's not also ranking in Google search console. Like are there funnels that are coming in that you can optimize for or there's is there an organic miss where you're getting people in like a true analysis of second, you know, third page behavior after they come to that that page. Maybe that's the more useful play rather than just wishing wishing for not provided to be dead.
[00:54:31] Speaker B: No, no. We're spitballing it. And the way I solve this on the tool I created last year, organic roi.com or organic ROI was having a first party pixel and trying then to leverage log file analysis and first party pixel to connect actual data, not third party data. And it is challenging, not impossible challenging. And we ran out of steam on that product just because the attribution side of it proved math worthy.
Like we could solve it but Actual implementation was a lot more outliers than what I wanted.
I think there's the possibility long file analysis maybe being the next frontier that SEO is really starting to focus on and having better tools for that. And I built tools that analyze, you know, LLM traffic easily from log files and you know, build stuff that looks at distinct paths that you know, as I say this, that crawlers are following. Obviously they're not necessarily following, but they're looking at whether they're coming back consistently to certain pages, certain paths, aggregating those. I think that gives us more insight into addressing user intent and ultimate satisfaction than maybe looking at third parties and seeing gaps. And once again there are no bad ideas. So not knocking the idea, but I think that would be a good frontier for us to have simple tools for doing complex things that have a high degree of confidence and first party data is really probably where that exists.
[00:56:25] Speaker A: Give a shout out to where you're at now. Can people where's your. What's your social media platform of choice? If people want to ping you with questions or work with you, let us know. You know, I think you said you were in a personal We've gone back and forth between agency and consultancy. Where you're at at the moment. Are you looking for to take anybody on or you're working on anything that you're looking to promote or release?
What's new on that side?
[00:56:52] Speaker B: I'm always after beta testers so go to appslicer.com and find if I even put a link in there that says beta test. But I've got some stuff that I just have to add you in so you can add your Google search console data. Bless you, number one.
Number two, my day job is Fiat Growth.
We're an aggressive fintech based, fintech focused growth agency out of San Francisco.
My day job, I run the SEO and geo platforms there.
I'm always building stuff.
I'm also a U.S. ambassador for Waycay and Inlinks, a better way of tracking and improving how you show up in LLM Search and Inlinks, the grandfather of entity SEO tools. It's been around for seven years now really about how do you optimize your site for anti SEO. Apart from that, I'm really happy I don't have to take on any new clients but if there's any interesting projects and I mean these are problems that no one else has been able to solve that gets me up in the morning. I love that stuff. So yeah, always open for chats with interesting people. Find me on LinkedIn and that you can do a search for Grant Simmons SEO on Google, LinkedIn or anywhere else.
And if we want to alienate people, I can say I'm not on bluesky, I'm on Twitter or X or whatever you want to call it. I am on bluesky, but I don't post there.
Twitter is still a great platform for SEOs, even though there's a lot of other crap going on there. There's also really good SEO community.
Yeah. LinkedIn.
And also find me at the next search conference near you. I'm SMX this week in Boston. I'm going to be SEO FOMO on a panel with Alada and some of friends there in a couple of weeks. I'm at Digital Summit in Denver and then I'll be running SEO in San Diego in September.
Yes. So otherwise find me traveling the world, enjoying time with my family.
It's a good life. SEO has been very good to me. And if one person gets one good thing out of this interview today, it's all about giving back.
[00:59:14] Speaker A: I love that. And thanks for stopping in and spitting tax. I appreciate it.
[00:59:19] Speaker B: No worries. Great. Brilliant. Thank you.
[00:59:22] Speaker A: See ya.
