YT YouTube2Text

YouTube2Text → Transcripts → Kyle Roof Ranked Gibberish #1 on Google. Then Google Deindexed 20 of His Sites.

Kyle Roof Ranked Gibberish #1 on Google. Then Google Deindexed 20 of His Sites. — Transcript

by Edward Sturm · 25,922 words · 3,728 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00Kyle Roof, you are a legend in the
0:02industry. You've done hundreds of SEO
0:04tests. Your work is consistently shared,
0:07constantly shared all the time by
0:09everybody. I've shared you. Joy Hawkins
0:12has shared you. Everybody has shared you
0:13and I'm just so hyped to have you on. Uh
0:17so yeah, thank you for coming on, Kyle.
0:19>> Well, thanks for having me. I'm very
0:21happy to be here.
0:21>> And for for people who don't know, like
0:24if you're in digital marketing and you
0:26don't know Kyle, um you absolutely
0:28should. And uh I think by the end of
0:30this episode, you're gonna be like, "Oh
0:31my god, this guy's amazing." If you
0:32don't know who he is, um because I mean
0:35you to say,
0:35>> well, dude, all of the tests like in
0:372018, you famously you ranked a page
0:41containing mostly gibberish for a
0:45rhinoplasty keyword. We're going to talk
0:46about that. U but all your tests are so
0:49interesting. Um we were DMing about um
0:52how no how you found out that no follow
0:55links actually count. Um yeah. All
0:57right. This is going to be fun. So, I'll
0:59start with this question. What is the
1:01most surprising thing that you've
1:02learned about that you've learned from
1:05testing Google's algorithms that most
1:07SEOs still don't know?
1:10>> Uh, it really doesn't change that much.
1:13Uh, you know, like everyone's like, "Oh,
1:14SEO is always changing." From my
1:16perspective, not really. Uh, the
1:18foundational stuff is always the same.
1:20They do tweak the knobs a little bit.
1:21This is a little more important or
1:22that's a little more important, but the
1:24core part of what comes down to rank is
1:26always the is almost always the same. It
1:28is the same. It's always the same.
1:30>> I'm so so happy you actually said that
1:33cuz like I mean maybe you hear this all
1:35the time. I certainly do. It's like um
1:38what are the what are the 2026 methods?
1:41Like you got to be I I've heard people
1:43say you have to be constantly adjusting
1:45in this industry. I'm like no you don't.
1:48>> Now the the beauty of it is that the
1:50secret is hiding in plain sight. Google
1:52doesn't hide the ball. You know they
1:54show you the sites that they like. And
1:56the idea is, you know, things do change
1:58a bit keyword to keyword or niche to
2:00niche, but I can tell you in the last 11
2:03years of running consistent tests every
2:05year. I can tell you that the core
2:06foundational part of it doesn't change.
2:09Little bits and pieces do change. Uh
2:11especially when they're being
2:12reactionary towards maybe some new
2:14method or some new thing has come out or
2:15some new form of spam. Of course,
2:17they're going to adjust a little bit,
2:18but they're adjusting within a framework
2:20that already exists. They can't take
2:22away the core parts of the algorithm
2:23because it would just completely change.
2:25And at the end of the day, the average
2:27user is very happy with the results.
2:28That's why they continue to come back.
2:31The only people that really obsess about
2:33it are nerds like you and me. Like,
2:35we're the we're the only ones that are
2:36like, "What's going on here?" Like, but
2:38the average consumer doesn't care.
2:40They're getting the answers they want on
2:42a consistent basis, and that's why they
2:43continue to come back. So, there's no
2:45economic incentive for Google to change
2:48that. it is their incentive is to police
2:51it, you know, a bit is really what
2:53they're they're shooting for, but to
2:55change the algorithm or it's always
2:56changing is not on their agenda
2:58whatsoever.
2:59>> So, you mentioned the the core parts of
3:01the algorithm don't really change. What
3:03are the core parts of the algorithm or
3:04algorithms?
3:06>> Well, I mean, you know, you think about
3:10uh find, crawl, index, and rank. Those
3:13are the the four pillars of the whole
3:15thing. So Google needs to to be able to
3:17find your site and there are elements
3:19within that. Google needs to be able to
3:21crawl your site get all the way through
3:22there and there are elements to that.
3:24You know then Google index which means
3:26they need to understand the page. They
3:28need to understand what that page is
3:29about and then they rank it. And those
3:31four things are always consistent.
3:33Little bits and pieces change here and
3:35there within there and things become a
3:37little more important or less important
3:38but those have been the four elements of
3:40of what we're trying to do in terms of
3:42getting visibility within Google or any
3:44other search engine. And now with like
3:46AI search, it's still the same with
3:48those as well. You know, it's it's it's
3:50really it's all the same. It's all just
3:53kind of different versions of the same
3:54animal.
3:55>> This method of marketing is so
3:56effective, I had to make sure it wasn't
3:58against Google's rules before I kept
3:59using it. It's a form of SEO I call
4:02compact keywords. Whereas most SEO
4:04focuses on putting up articles to answer
4:06questions, how, what, when, compact
4:08keywords focuses on putting up dozens of
4:10pages that sell to searchers who are
4:13actually looking to buy. These pages
4:15rank on Google and convert so much
4:17better than normal that when I
4:18discovered this years ago, I couldn't
4:20believe this was allowed. It's less
4:22work, too. The average compact keywords
4:24landing page is only $415 words. Compact
4:27Keywords is a 13-hour deep course on
4:29getting sales with SEO. A customer said,
4:32"Compact Keywords contributed to a
4:34$4,000 sale within the first 6 weeks."
4:37Another customer said, "Give it to a
4:39junior employee. Have them follow it
4:41exactly as Edwards laid out. You don't
4:43have to do anything and you're going to
4:45gain a six-figure SEO level employee
4:48just by having them go through this
4:49course. Compact Keywords is about
4:51setting up an SEO funnel that brings you
4:53sales for years and years and years. It
4:56works with AI. It's less work than
4:58traditional SEO and it makes way more
5:01money. You can get it now at
5:03compactkeywords.com.
5:04Back to the podcast. Yeah, the honestly
5:06like some tactics change, but the
5:09tactics that change are for the people
5:11who are doing more churn and burn type
5:14SEO where you're trying to abuse a
5:17loophole. So like uh an easy, you know,
5:19an example might be what's about to
5:21happen with Discord. So I don't know if
5:23you heard, but Discord is is making um I
5:27think it's public servers and they're
5:28now going to be index indexable. Um and
5:31so you know Discord Yeah. So, it's going
5:34to be maybe what happened with Dig or
5:35Dig is like at the beginning of this
5:37year, easy example, everyone tries to
5:39spam Dig relaunches. Everybody tries to
5:42spam Dig for SEO. It actually works for
5:45a little bit.
5:46>> DIG shuts down all these these Digs, all
5:49these communities, uh, bans all these
5:52people, doesn't work, and then DIG just
5:54completely wipes the slate clean again
5:56and relaunches as like an AI only
6:00community or something. And um and so
6:02it's like yeah these are tactics that
6:05come and go but the but like the core
6:08parts like relevance and authority
6:10>> if you know if you understand that well
6:13>> you're going to be able to do SEO I
6:15don't I mean like for honestly forever
6:16because what's how how are search
6:18engines going to
6:21>> D you know you you told you were telling
6:22me before this you're going to be on
6:23stage with David Quaid friend of the
6:25podcast he was on yesterday
6:27>> and David David always says like Google
6:30doesn't understand the content. They
6:33they make it seem like they do and
6:35Google has said that themselves in
6:36internal presentations.
6:38>> Yeah. I mean it's a the one thing that
6:42you know Google is so it's powerful and
6:44it's amazing and and it's it's
6:46revolutionary and what it's done you
6:48know to the world is is phenomenal. At
6:50the end of the day, it's not a human.
6:52You know, it can't read as humans can.
6:54Like Google doesn't read a piece of
6:55content and be like, I am a much better
6:57computer now forever read that piece of
7:00thank you. You know, like that's not any
7:02part of there's no value judgment.
7:04That's the other two. They're not doing
7:05any kind of value judgment whatsoever on
7:07on the content. It's just if it
7:09satisfies the math and that's always
7:11what it comes down to is you have to
7:12give a algorithm the math that it wants.
7:15Now again things do tweak here and there
7:18but that's the core part of the whole
7:19thing is and that the only way it can
7:21read is through math that's it that
7:23that's the only way it can handle and
7:24now you think like oh well AI can
7:26understand you realize that AI is also
7:28an algorithm and it's also just doing
7:30math you know and so it it from a
7:33perception point it's like oh wow that's
7:35doing it can really understand what no
7:37it's it's doing just slightly different
7:39math that can comprehend things in in a
7:41in a slightly different way and it feels
7:43human in a lot of aspects but it's not
7:45human and it will never be.
7:47>> And honestly, you don't you don't want
7:48it to be. It's like a good example is
7:50like you might have like some video that
7:52you love and you send it to a friend
7:53thinking they're going to love it too
7:55and they're just kind of like meh. And
7:57so like and then they send something to
7:59you that they love and you're also like
8:01meh. And and so so like um for a search
8:05engine to to show you the best content
8:08like what's it going to be looking at?
8:10just like thinking like first principles
8:13it's going to be you want the content to
8:14be relevant to the thing people search
8:16you want the site that it's on or the
8:19page that it's on to be somewhat
8:20authoritative um and then you want it to
8:23have good aggregate user signals which
8:25Google can measure because it has all
8:27the surfaces in the world all right
8:29you've done um so you've done 400 more
8:32than 400 SEO experiments right
8:35>> that's the marketing line but we've been
8:37using that line since like 2018 I No,
8:40it's been it's a few more than that, but
8:41it was just like that seemed like a
8:42number.
8:43>> Let's say like I don't know 800,000s. A
8:45lot. You've done a lot.
8:46>> A lot. A lot. Yeah.
8:48>> Is there anything that you are
8:49absolutely convinced was a ranking
8:51factor until your own testing proved you
8:53wrong?
8:55>> Well, one of the big things that I got
8:57started with like way back in the day
8:58was like um uh readability. Like how
9:02well can this be read? like um uh like
9:04the grade level or like a flesh concaid
9:07reading score or grammar for example or
9:10using bad words like all that kind of
9:12thing like those are all ranking factors
9:13and and everyone was convinced at the
9:15time that they were cuz that was just
9:16the common you know common thought is
9:20like you need to be at this level and
9:21Google is checking and then none of that
9:23was true like it just it wasn't true at
9:25all but if you think about it again
9:26think about from a value judgment
9:28perspective let's say English is your
9:30third language but you have something
9:31really good to Okay, is Google going to
9:33start grading your grammar like for any
9:37sort of relevancy or or importance or,
9:40you know, authority? And there's simply
9:41not. And also, that's going to be really
9:43expensive to do. And you have to realize
9:45the bottom line is they need to do this
9:47as cheaply as possible. And and and
9:49doing a seventh grade grammar test isn't
9:52anything that they want to get into
9:53whatsoever.
9:55>> You remember what the test looked like?
9:58>> Yeah. Um, so it was setting up uh
10:02identical pages and then um swapping out
10:06um words so we could like sub out a word
10:08for like the same kind of word but then
10:10subbing out the word uh incorrectly like
10:12misspelling it for example. So now one
10:16test page has, you know, 800 misspelled
10:18words within, you know, 1,500 like half
10:20of the page is misspelled and it doesn't
10:22drop in rank, you know, like that kind
10:24of thing. And running tests in similar
10:27fashion just like that with all of those
10:29kind of things like with uh reading
10:31level or, you know, page speed. I had
10:33the most fun um we had pages that were
10:36loading at like 25 seconds and we
10:37couldn't get them to drop, you know, but
10:40page speed is so important.
10:42It's so true, man.
10:44>> Yeah, those are all those common things
10:46and they make sense though, you know.
10:47They make sense when you're putting of
10:49course you should have a page that that
10:51reads well. Of course you should have a
10:52page that isn't swearing at at the
10:55consumer. You know, of course you should
10:57have a page that loads quickly. Those
10:58are all things that go towards
10:59conversion, but they aren't things that
11:01go towards rank.
11:02>> Yeah. It's it's user signals. It's we're
11:04you're not saying that try to make a
11:06page that's slow because like also if
11:08it's a if it's a competitive SER and
11:11you're slow then they might actually go
11:13back but like if people believe if
11:15people believe they're going to get what
11:17they want
11:18>> from your page they'll they'll wait the
11:20extra seconds to get it.
11:21>> That's true. They will put up with a lot
11:22if if they think they're in the right
11:24spot. That that's entirely true. Uh you
11:26see some of those pages that just have
11:28the worst conversion metrics of all time
11:30and then you find out that they're just
11:32crushing it. It's like, damn it.
11:34[laughter]
11:34>> Yeah.
11:35>> How are they doing that?
11:36>> Yeah.
11:36>> But but also none of that had anything
11:38to do with rank.
11:40>> What do you mean?
11:42>> In terms of like uh if they have bad
11:44conversion metrics, you know, or like or
11:46like, you know, they have a bad call to
11:47action. Like none of that's going to
11:49matter towards the ranking of the page
11:50at all. And and people inherently trust
11:54uh sites that are on page one anyway.
11:56So, a site could be not quite as
11:58beautiful or not quite as, you know,
12:00converty looking as it could be, but
12:02because it's on page one, people are
12:03going to trust a lot more.
12:05>> How important have you seen behavioral
12:07signals to be?
12:08>> Those are the hardest thing to to to
12:10measure. Uh, the the clearest one and
12:13and Rand Fishkin did this way back in
12:15the day was showing that you could
12:16manipulate CTR. So, the the click is
12:20that that's one you can see and you can
12:22you can play with that and you can see
12:23that is is a thing. But when it comes to
12:25something like a bounce rate or like
12:27pogo concepts, you coming in and out of
12:29the SERs to different sites or dwell
12:32time or time on site or uh um like pages
12:37click through, those are impossible to
12:38to find anything clear there. Though you
12:41do feel that the sites that do better do
12:42have those things. That's the
12:45correlation. But it's hard to pin down
12:47exactly what might be being measured and
12:49something that you might be able to
12:50improve or optimize and say like
12:52definitively if you do this you will get
12:54that result. I've never been able to get
12:56that.
12:57>> You think dwell time matters?
13:00>> Feels like it, doesn't it? But then on
13:01the other hand, I can't tell you.
13:02>> Oh, I I don't think it does. I think
13:04pogo sticking matters, but I don't think
13:06dwell time does because somebody might
13:08get their answer right away and that's
13:09it. And then there's and that's
13:10>> that's always been my thing like on
13:12bounce rate. You know what? If somebody
13:13read something, they got their answer
13:15and they left.
13:17How's that a bad experience? Why would
13:19Google say that thinking is like
13:21somebody goes and then they're click on
13:23something else without doing another
13:24search, which would indicate that they
13:26were not satisfied with the first
13:28result?
13:29>> I could I could play devil's advocate on
13:31that.
13:32>> So, they read a page, they like it, and
13:35then they go back and as a sensible
13:36human being, they click on another one,
13:38get a little confirmation. They click on
13:39a third, they get a little confirmation,
13:40and they're gone.
13:41>> That's true. had a great experience.
13:43They had a great experience and then it
13:45would be terrible I think to punish the
13:47first page that actually gave the
13:48complete answer and reward than them the
13:50pages that just confirmed that complete
13:52answer. You know what I mean?
13:53>> I do actually and and but it might be
13:56looking at the aggregate scores of all
13:59the pages in a
14:00>> could again there could be a lot of
14:02other things going on. Um there's the
14:05idea of cocon coconurrent factors. So
14:08the idea is it's not a factor until
14:09something else pings as well. And that's
14:12probably a lot of what's going on with
14:13those types of um user signal factors is
14:17that other things need to appear. And so
14:19it becomes very hard to put into a test
14:20environment because it's hard to know
14:21what might trigger or activate the
14:24signal and then it's very hard then to
14:26to measure those things against each
14:27other. So, um, I've heard a lot of
14:29theories about that, like you need this
14:31like triggering factor and then these
14:32other ones get considered and that's
14:34what makes it really really hard to to
14:36nail down or to say definitively yes on
14:38this one, no on that one. A lot of
14:40people think that's what's happening
14:41with these mount AI patterns. So it
14:44might be like you have um all of a
14:46sudden you have like a spike in bad
14:48behavioral signals and then you'll look
14:50for um or it maybe it's a spike in like
14:54all these new pages that are getting
14:56crawled that are getting indexed um and
15:00and then simultaneously you see these
15:02poor user signals and then maybe it
15:03triggers some sort of like AI
15:06>> evaluation. Yeah, could be. You would
15:09think though for them to spend the money
15:12for a to go in it would have to be a
15:14really
15:15>> um
15:17a YML type niche, you know, your money
15:19or your life. So, finance, health, you
15:22know, loans, that kind of stuff. Like I
15:24could see maybe they'd want to spend a
15:26little more time there in some dangerous
15:28niches, but if you're selling pots and
15:30pans, I have a hard time believing that
15:33they're going to spend some AI time,
15:36you know, going through on all the
15:37things they could spend computing money
15:39on. That feels like the least important.
15:41>> They honestly might not need it anyway
15:43because like if you're throwing off if
15:45you just spiked behavior behavioral
15:48signals and you spiked pages, that could
15:51just that could be it. That could be all
15:53they need to know.
15:54Well, you got to figure they already
15:56have a framework that is pretty
15:57successful. If they find indicators like
16:00when this, it's usually that they might
16:03just dial it up on that one factor. But
16:05here's the the part that hurts is that
16:08um let's say there are 80 sites that
16:11Google has identified as uh doing bad,
16:14you know, outside of what they like. And
16:16then they tweak something and they
16:18realize they're going to kill a 100
16:19sites. 20 of those sites weren't doing
16:22bad, but they're going to get the 80.
16:24Google will take that as a win.
16:26>> Google will crush you. Google will take
16:28that collateral damage, which could be
16:30your livelihood, and they have no
16:32problem with that whatsoever.
16:35That's the scarier part is that you
16:38could fall into when they make these
16:39adjustments to find people that are
16:41spamming or do something terrible, but
16:43they will kill off a lot of sites. Now
16:45take that across the billions of
16:47websites and in in in the aggregate, you
16:50know, at 20% turns into a lot of mom and
16:53pop family businesses or, you know, a
16:55lot of like midsize businesses that have
16:5720 employees and now they've got some
16:59pretty hard decisions to make.
17:01>> Yeah.
17:01>> So that Google could catch those other
17:03bad actors.
17:03>> We we saw that we saw that happen in the
17:05past, right?
17:07>> Happens all the time.
17:09>> All the time.
17:09>> Happens all the time.
17:10>> Yeah. and and um a lot of people think
17:12that's coming right now with all the
17:14like what's Google's biggest problem
17:16right now. A lot of people think it's AI
17:17content and uh and so yeah, a lot of
17:20people think a major HCU
17:23level update is coming and there's going
17:25to be massive amounts of collateral
17:27damage.
17:29>> Not as concerned about that because I
17:30think they already solved that problem.
17:32Uh, and I think they solved it because
17:34LLMs, you know what's hilarious is the
17:37worst thing that LLMs can do is content.
17:40Um, in terms of what we need for SEO and
17:43in particular, what you might consider
17:44LSI terms or contextual terms, they do a
17:47terrible job putting those in. That's
17:49already built into Google's algorithm.
17:51So, all they have to do is crank it up a
17:53little bit on the importance of those
17:55things and that's going to sync all that
17:57spam content because people are not
18:00properly prompting to get thing uh get
18:02their content into the range that's
18:04needed for the number of contextual
18:06terms you need on a page. So, I I I
18:08think that's actually how they've been
18:09doing it over the last year and a half.
18:10They might crank that up and that might
18:13be what it is, but I don't think it's
18:14anything special. I think it's all
18:16completely within what they've already
18:17built.
18:18>> Can you go deeper into this? Sure. So,
18:21let's say in a thousandword article, um,
18:24you need 200 contextual terms. So, you
18:26need like your target term, you know,
18:28your main concept, you probably need
18:29variations of that term. And then you're
18:32going to need the terms around that that
18:34create meaning. You know, that and so
18:37like, you know, if you're the example I
18:38give all the time is if you're talking
18:39about a kitchen, you could use words
18:41like sink and stove, refrigerator.
18:44You're talking about an actual kitchen.
18:45Or if you're talking about a kitchen,
18:46you could say, you know, um, redesign,
18:50you know, uh, French country, you know,
18:53renaissance, whatever. I mean, now
18:54you're talking about, you know, redesign
18:56and remodeling a kitchen. Or you can
18:58talk about a kitchen and say love,
19:00holidays, family. Now you're talking
19:02about the kitchen is the heart of the
19:03home. And each of those things, you're
19:05talking about a kitchen as a main
19:06concept, but the contextual terms tell
19:08you what you're actually talking about.
19:10LLMs do a very bad job of getting those
19:13contextual terms in. And so let's say
19:15you need a hundred, let's say out of a
19:17thousand words, you need 200 contextual
19:19terms. So Google understand that's the
19:21concept you're talking about. Most LLMs
19:23are going to be one to two standard
19:25deviations below the range that you
19:27need. And what I found is if you're more
19:28than one standard deviation below that
19:30range, you're going to you you're in the
19:32red zone of getting hit.
19:35>> And
19:37do you use AI at all for for SEO
19:40content?
19:42If you have the right uh guide rails,
19:44it's fantastic. So, we have like we have
19:45our own AI writer, of course, but um
19:47using claude and like a good report that
19:49tells you know this is the what you
19:52need. That's a great way to do it too,
19:53>> but you have to give it though you have
19:54to give it that those parameters. These
19:56are the terms we need approximately this
19:58many times for each term to get an
19:59aggregate total of this many terms for
20:02the length of the article that you're
20:03trying to write so that Google and other
20:05search engines and other LLM search will
20:07understand what you're talking about.
20:08Yeah, we'll go deeper on this in a in a
20:10couple moments. I want to ask you about
20:12um in 2018 you so you ranked a page
20:16containing mostly gibberish for the
20:18rhinoplasty keyword. Um do you think you
20:21you could you could you share like what
20:23you did and then do you think that would
20:26be reproducible today or has Google
20:29fundamentally changed how it evaluates
20:31content?
20:33>> Sure. So, uh, the idea came up because
20:35there was a competition to to do
20:37something and, um, my team and I, we had
20:39actually already done this. You know,
20:40this was something that we were doing as
20:42we were testing things and trying it
20:43out. And I was like, you know, would be
20:44really fun is to show people that
20:46content doesn't matter. You know, that
20:49that was that was the impetus. You know,
20:51the content is king is a narrative that
20:53I really don't enjoy that much. Still
20:55don't. And I was like, let's just let's
20:57just kind of thumb our noses at those
20:59guys. And uh so it was a 30-day
21:02competition and the only requirement was
21:04you you had to have a fresh domain.
21:06After that you could do whatever you
21:08wanted. And so my team and I we uh we
21:10put up a site that was written in Laura
21:12Mipsum. And then we did the math. So
21:15let's you know we figured let's so for
21:17the target page we needed say 1500 words
21:18for this page right and then uh we did
21:21the math on how many times we need the
21:23target keyword its variations its
21:25contextual terms and then we very
21:27crassly copied and pasted that into the
21:29lauram
21:31all right so it was like Lauramipom
21:32lauripom laurum nose job so rhinoplasty
21:35is is a nose job so Laurum laurum
21:38patient Lauram Laurum you know Texas
21:40because it was also based in Texas that
21:42kind of thing and Um, after the end of
21:44the 30 days, we took fifth, but nobody
21:47was really lighting the world on fire
21:48cuz it was a a 30-day uh 27
21:51professionals entered, by the way, and
21:53only seven got a page to rank, which is
21:55kind of fun.
21:56>> Kind of fun to think about that.
21:58>> Um, but about two weeks later, we went
22:00to uh number one in the maps and number
22:04seven organic. About two weeks after
22:06that, we were number one organic and
22:08number one maps. And then a little while
22:10after that, we actually wiped out the
22:11maps and we were the knowledge panel,
22:12which is hilarious.
22:14>> What did the knowledge panel look like
22:15with your Laura Mipsum stuff.
22:17>> It It had our It had our stupid header
22:20[laughter] like our like stupid like
22:21thing and it was just like
22:23>> that's so funny.
22:25>> Our stupid It was It was very funny. Um
22:28and it was great because it was it made
22:30the point that we wanted to make, you
22:31know, and and everybody's having a good
22:33laugh. And then an article was written
22:35about it. [laughter]
22:37Google ranks site written in Latin
22:40and then um six hours later Google
22:42de-indexed the site.
22:43>> Yeah, [laughter]
22:44>> which is fair play. And then though um
22:47the scary part was that night from like
22:491:15 a.m. to 1:27 a.m. Google de-indexed
22:51um 20 of my test sites.
22:53>> Oh yeah.
22:54>> And that was a little uncool. [laughter]
22:56>> They went through your GSC. Did Did you
22:58have all of these connected to the same
23:00GSC?
23:00>> No. No, they were all completely not
23:03separate. Not not connected whatsoever.
23:05They weren't linking to the test site.
23:06They weren't linking to each other. They
23:07were on different hosts, different
23:08servers, different everything. Google
23:10just gave a phone call to the registar.
23:12>> Yeah.
23:13>> Which I'm sure of it.
23:15>> That's
23:16>> They're like They're like, "Hey, Greenho
23:18like who registered these things?" And
23:19they're like, "I'm sorry, we can't do
23:20that." And there's a pause and they
23:22like, "Here you go." [laughter]
23:25Just sent the I'm sure of it. There's
23:26the that was the only way. Um I can't
23:29remember who we registered through. It
23:30wasn't any of the big ones. I didn't
23:32mean to disparage any of them, but um uh
23:36yeah, so then they took those down and
23:38at first I was I was really concerned
23:40about the whole thing, but um then I
23:43realized they completely validated
23:44everything I was saying. You know, if it
23:46was nonsense or I was getting lucky,
23:48they would have rolled their eyes and
23:49moved on. But instead, they tried to
23:51take some punitive a uh action. And I
23:53think the biggest issue is because I
23:55showed there's a way where you can learn
23:56about the algorithm. you know, you can
23:58go through and I got into it about a
24:01year or so later with somebody or two
24:03years later with somebody and I was
24:04like, you know, we were encouraged to do
24:06this. You know, you can go through the
24:08Twitter feed of Matt Cuts and then, you
24:10know, John Mueller and all these other
24:11guys. They tell you test the algorithm.
24:13Test it. Test it. It was really funny. I
24:14went to make my point about that and
24:16they had scrubbed all their Twitter
24:19completely completely scrubbed the
24:21Twitter, which was kind of funny. But
24:23about 2 years later, I was doing a uh
24:26course on white hat SEO cuz I'm I'm
24:28actually a white hat SEO, which is
24:30people think I'm black hat cuz I think I
24:31I test, but whatever. Anyway, I was
24:34doing this um course. I was refreshing
24:36myself on the on the rules. And in the
24:39um automatically generated content, it
24:42said you can't have a web page that
24:44makes no sense to the reader but
24:46contains keywords. And I was like, well,
24:48when did that become a rule? And so I
24:50put in the wayback machine and before
24:52the competition rule doesn't exist. One
24:54week after one week after my site is
24:56exposed like everybody can see it and
24:57take a look at it and play around with
24:58it. The rule went in and then they they
25:00punished me for that rule like 3 months
25:02later.
25:03>> That's crazy that they took down your
25:05test your test sites.
25:07>> Yeah.
25:07>> Why do that? Why why why go why like
25:10okay I I get taken taken down the like
25:12Lauram the rhinoplasty site but like the
25:15test sites. So uh in the in the search
25:18console um uh the the the rhinoplasty
25:22site was pure spam was the um was the uh
25:26the penalty. All the others were just
25:27major spam.
25:30[laughter]
25:31You do see that sometimes where they say
25:33that was happening when uh a lot of
25:35people they were just it was like when
25:37Chachi PT came out they're like look
25:39what I can do and you would see that you
25:41would see the pure spam and it would
25:43happen a lot to the to the like
25:44influencers who are who were sharing
25:46their results on uh on X. Yeah, look I
25:49put out I put out 5,000 pages and
25:51they're all ranking and
25:53>> so to your question about does it work?
25:55So the next so I got a lot of hate mail
25:57for what I did which I wasn't
25:58anticipating and and people saying oh
26:01that doesn't work anymore there's a
26:02Lauripum update or this Google that so
26:05like the next year I spoke in um uh Bali
26:11in Hanoi in Mai in Milan and Munich and
26:14I ranked a page on page one in each city
26:17in Laura Mipsum for rhinoplasty for that
26:19city and and and and showed it off. So
26:22that was all those different, you know,
26:24regions, all those different SERs, all
26:27those different whatever. And and it and
26:28it worked every time. And uh I've kind
26:30of let it go a little bit, but two of
26:31the pages are still live. They're still
26:33on page one. And so they've been on page
26:36one since 2019. So we're talking seven
26:38years. And so when people want to tell
26:39me that it's all changed and Google's
26:41figured it out, I always then go look at
26:44my my Lauram sites and I'm like, nah, I
26:46don't think so. But then to your other
26:49question, does it still work? I do it
26:50all the time. I'm running tests right
26:51now that are completely Lauren Mipsum.
26:53It works. It gets up. They rank. They
26:56index. You can play with them how you
26:57want to play with them. It's all still
26:59there.
26:59>> You know what's funny, too? You You
27:01actually see that all the time from from
27:03just normal sites that rank. People
27:04leaving their Laura Mipsum blocks on
27:06Elementor.
27:08>> It happens all the time. Yeah. All the
27:09time.
27:10>> All the time. Or like um oops, this uh
27:13drop down box doesn't open anymore and
27:15there's a ton of stuff in there and that
27:16seems How does that work? That's so
27:18crazy. It's just cuz that's how it
27:20works.
27:23>> Also, it's for there are so many people
27:26out there who was like, "Google's never
27:28going to catch me and they just they
27:31they called your registar. They they
27:33just got like Yeah. I mean, that's Dude,
27:36that's crazy.
27:36>> That's my guess. That's my guess. But
27:39none of the sites none of the sites were
27:40connected.
27:41>> Yeah. Like Yeah. And and that's and
27:42that's that's amazing. Thank you for
27:44sharing that. Um, could for people at
27:46home who who want to do this are like
27:48they're like, "Man, I want to be like
27:49Kyle Roof when I grow up and and they
27:52want to Yeah. No, but they but they want
27:53to Yeah. They want to do this. They want
27:54to get these results." Like,
27:56>> what what are you doing um that the
27:59average SEO is not doing when you are
28:02ranking gibberish sites?
28:06>> Well, what I found the easiest thing is
28:08grab an expired domain. So, a site that
28:10already had traffic. You don't have to
28:11revive it, but just it was on the radar.
28:14So it it's already kind of past a state
28:16where it may or may not be crawled cuz
28:18the biggest issue is, you know, you
28:19launch 10 sites, real sites, even seven
28:22or eight will will do fine, but you can
28:24have two or three that just won't take
28:26off, right? And you can do everything
28:27right, everything 100% right, 100%
28:30within bounds, and you have two or three
28:31won't go. It's the same in the testing
28:32world, too. And that's frustrating
28:34because tests take a long time to set
28:36up, and they're they can be expensive
28:37and timeconuming. So a way to fast
28:40forward that is just grab an expired
28:42domain. you know, something that that
28:43existed already and had some traffic
28:45coming to it. Don't reput up the pages,
28:48but you can use that domain and and they
28:50they work faster and they hang out a
28:52little bit longer is kind of a a more
28:54recent tip on on how to get into it. So,
28:56grab a $10 domain that it's recently
28:58expired. Doesn't have to be anything
29:00special, but it just it was indexed and
29:02it was crawled and Google will crawl
29:04that one pretty quick and you can get
29:05some fast test results doing that. for a
29:07for a business that you were starting
29:09long term where you're like, I'm going
29:10to stick with with this thing for three
29:13years, five years, 10 years, would you
29:15start with an expired domain?
29:18>> 100%.
29:18>> Yeah.
29:19>> Um,
29:19>> start with an expired domain.
29:21>> Yeah. Even one that only has like a
29:23year's worth of you're fast forwarding
29:25eight months of your life by getting
29:27something. I would find something that
29:28is ranking for a couple of keywords
29:30within my niche that I'm going longterm.
29:32I would totally do that. Don't take the
29:34chance that that Google just hates the
29:36site. you know, or whatever. See
29:39something that has some traffic, has
29:41some has some previous staying power,
29:43and is also within the niche you're
29:44going for. And I would totally spend
29:46money on that. Cuz think about it, if
29:48you spend like a couple thousand dollars
29:49on a even on a domain, you're saving
29:51yourself probably 10K easy in SEO
29:54building up to that. It's a big money
29:56saver. It's a big time saver for sure.
29:58>> What do you mean by don't take the
29:59chance that Google hates the site?
30:02>> I mean, you you can put up a site and
30:04Google just doesn't love it.
30:06That's just the fact of the matter. Like
30:08you can put up something and Google just
30:10won't rank it for for anything and it
30:12doesn't matter what you do
30:15>> when you've seen that happen. Why does
30:16that happen?
30:18>> Well, I think there's just I think
30:19there's some randomization factor
30:21probably
30:22to it, but um it's hard to explain cuz
30:25uh you can put up very normal sites,
30:27very real sites, and very benign niches
30:31and every once in a while a couple of
30:33sites just won't take off like they
30:34should.
30:36you know, they might they'll see a
30:37little bit of love here and there, but
30:38nothing like they should do. Especially,
30:39you know, you know, you're doing it
30:40correctly. You know, you know, you're
30:42doing it in the way that Google likes
30:43this. This is good content. You know,
30:45this is within the parameters of what
30:46they want. This is set up that
30:47structurally that's all all the
30:49technical is on point. There's this this
30:51site should rank and it just won't. So,
30:54I would not take that risk on something
30:56that I'm doing longterm. I would I would
30:57I would fast forward the game a little
30:59bit and get something that's got a
31:00little bit of time under its belt. When
31:02you're buying expired domains, what do
31:04you look for?
31:06>> Traffic. Uh ranking for keywords and
31:08ahead traffic. That's those are the two
31:10things.
31:11>> So, you'll give it to maybe hrefs and
31:13you'll just see and yeah.
31:17>> Yeah. You have to trust third party
31:18tools a bit and and those are definitely
31:20a guess, but I just want to see that at
31:22least they tracked something related to
31:25what I'm trying to get to and then there
31:26was some traffic there. So, you know, if
31:28you take that over, it's a lot more of a
31:30fertile ground than just a domain you
31:32just plucked right out of nowhere and
31:34and are running with.
31:35>> And then how about keyword placement for
31:37for these gibberish sites?
31:40>> Sure. I mean, the top four places put a
31:43keyword are still the top four places.
31:44That's your title tag, your H1 in the P
31:47tags, and in the URL. If you did those
31:50four things, you've done 60% of SEO. How
31:52much how much uh placement in the P tags
31:57is too much is is keyword stuffing.
32:00>> That's not a number. That's that's
32:01that's not a number at all. Um it is,
32:04but also I can tell you um the
32:06variations of your keyword are going to
32:08do more heavy lifting than the actual
32:10keyword itself.
32:11>> Um for sure. So it's not putting that
32:15keyword in 10 times is going to win it
32:16or 20 times or or one time. You can do
32:19it no times and still do just fine. Um,
32:23you do probably want to do it one time
32:24though,
32:26just as a general rule, but uh the the
32:29variations of that keywords are
32:30different ways to say it. Uh, and then
32:31your contextual terms are going to do a
32:32lot more heavy lifting than trying to
32:33figure out how to spam uh key uh how
32:37many times put the target keyword in.
32:38>> Yeah, it's it's it's um it's also like
32:41dependent on the keyword. Like if you
32:42have a two-word keyword, it might just
32:44be unavoidable to use it on your page a
32:46lot of times. Whereas like if you have a
32:48longtail keyword and you're using it on
32:50your page a ton of times like that's a
32:52lot easier.
32:53>> It's a little awkward. It's a little
32:54awkward. Uh um a long time a long-term
32:58friend of mine Ted Kubitis talks about
32:59he uses the example if you want to rank
33:01for the word games just for some for
33:04some unknown reason you want to go over
33:05that like the number one term has it on
33:07like 10,000 times on the page or some
33:10some 8 thou some stupid number you know
33:12of just it's thousands of thousand times
33:14because it's a oneword keyword you're
33:16going to say the word games quite a bit
33:18>> that's just that's just how it's going
33:19to go within the content.
33:21>> Is there anything else that you were
33:22doing? So, okay. You got the expired
33:23domain. Were you building links to it or
33:25or you didn't have to?
33:27>> Oh, no. No. If it's for a test, I'm not
33:29going to do that at all. Um, unless
33:32that's part of the test, but uh but
33:34otherwise, that's just a way to kind of
33:36jumpstart the the testing field that you
33:38can put together. And that's a great way
33:40to do it.
33:40>> Mhm. Mhm. And and when you're um when
33:44you're evaluating expired domains, to
33:47what extent are are you looking for
33:48sites where there's no drop off in
33:50rankings, so rankings didn't start
33:52decreasing? Maybe maybe it was just
33:54expired.
33:55>> That would be better of course. So the
33:57more action you get, but also I'm pretty
33:59cheap when it comes to buying testing
34:01domains. So I'll I'll I've got a little
34:04bit of variance on what I'm trying to do
34:05rather than like something I'm trying to
34:07do for a business or something like
34:08that. Um, more than anything, I'm just
34:10trying to get a feel that maybe it
34:11wasn't penalized, like somebody was
34:12using it as a PBN or something or, you
34:15know, some porn site or something, you
34:17know, and so I'm trying to look at the
34:18way back machine and just make sure
34:19there isn't anything quite suspicious,
34:22uh, showing up there. But otherwise, um,
34:24just looking for a little activity is
34:26all I need. And then if you have that
34:27for for setting up a testing site,
34:29you're you're well ahead of, um, of
34:31other people get trying to start from
34:33scratch. If you were buying uh an an
34:36expired domain for a business that you
34:38wanted to start longterm, where would
34:41you buy it from?
34:45Where do I buy them from now? Uh
34:50I don't have any secret sauce on that.
34:53Uh I I subscribe to some of the larger
34:55auctions. Uh I don't I don't have
34:57anything special. I'm sorry on that.
35:00Well, that's I go to I go to the regular
35:02places. I don't know.
35:04>> Yeah. Well, some people would would be
35:05like, "Oh, I Edward didn't ask that and
35:07I don't know his I don't know his secret
35:08sauce." And
35:10>> I hate to break people's hearts, like I
35:12know exactly the place to go now. I
35:14don't have any special access to
35:15anything when it comes to that.
35:16>> Yeah. Um, all right. If you had to rank
35:20Google's on page signals by importance
35:22in 2026,
35:24we we've been talking about this. What
35:26would your top your top four or five be?
35:29Well, they're the same as they always
35:31have. It's the title tag and your H1.
35:33And um we ran a fun case study on that
35:36where we found that on LLM content,
35:39something around 60% of the content that
35:41comes out LMS does not have a target
35:43keyword in the title tag or the H1.
35:46Phenomenal. Like that's just a
35:49phenomenal like stat. And that's off
35:51based off of 11,000 pages that we ran.
35:54Um so those two are the most important.
35:57They have been. and they always will be
35:58as I said in the in the paragraph text
36:00and also in the URL after that you're
36:01looking at your H2s and H3s um Google
36:04doesn't look past H4 so like H5 H6 is
36:08not any part of the algorithm but H2s
36:09and H3s are are very important H2
36:11obviously is more important than H3 H3s
36:13are more important than H4 but if you
36:15had to if you had to choose do at least
36:16an H3
36:18you know if you're stuck some people get
36:20stuck in their template or whatever but
36:21make sure you've got H2s and H3s and
36:22you've got your subheadings after that
36:24you're looking at uh anchor text within
36:25the body So um like a a jump link going
36:30from the top to the bottom counts that
36:32sort of a thing. After that um things
36:36pretty much start to equalize in terms
36:38of importance. You kind of get into the
36:40just do all the things sort of a mode.
36:42Um you know is it going to move the
36:44needle hugely? No. Is there a chance
36:46that it could move a little? Yes. Is it
36:48that hard to do? Also no. So just just
36:51get that stuff done. But those are the
36:53big ones. So, I'm really that's what I'm
36:54most worried about getting down to the
36:56subheadings and then getting um anchor
36:59text. So, that's clickable text that
37:01contains your target keyword or
37:03contextual terms or that sort of thing.
37:05Um after that, it's all pretty even once
37:07you get below there.
37:08>> How much how so okay when you're saying
37:10like use a keyword in these places? Uh
37:14is it a minimum of every you want every
37:16page to have um your target keyword in
37:20the H in one H2 and one H3 or is it
37:23dependent on the keyword? How are you
37:25thinking about that?
37:26>> It it it depends on the keyword for
37:27sure. Um like if you couldn't run a tool
37:31to actually get some data then do it one
37:33time. If I were to if I were to
37:35templatize it, you know, just say like
37:36we're just going to do it this way and
37:38this is our 8020 rule, I would do it one
37:40time. Um, that's kind of the approach I
37:42would take if uh you want to get a
37:45little more fun about it. Uh, if there's
37:47a different way to say your term, you
37:49know, uh, lawyer, attorney, or maybe
37:52there's like a industry way to say it,
37:54but there's also then the the layman's
37:56way to say it. U, I would have that in
37:58my first H2 because those terms are
38:00probably very closely connected, but I
38:02want to make sure that I'm I'm showing
38:03that it's very important to me that I
38:05that this variation of that term is
38:06there. So, uh, if I've got my bases
38:08covered in terms I've got some good
38:09math, good math, I want then I want to
38:11get variations in that first A2 H2,
38:13maybe more so than the target keyword
38:15itself.
38:16>> Tool. What tool should people use to
38:18make sure their pages are optimized,
38:20Kyle? [laughter]
38:22>> Well, I like my tool, but it doesn't
38:24hurt my feelings if you don't use my
38:26tool. But, um, the thing is uh, you need
38:29>> you can say you can say your tool. You
38:30can give the pitch. You could say you
38:32could say like people people want
38:35something that's going to work. They
38:36don't they don't care if it's a pitch or
38:38not. They just want results. And um
38:40you're you're extremely knowledgeable.
38:41So
38:42>> I I am I'm my business partners hate the
38:45fact that I don't pitch enough.
38:47[laughter] But uh um my tools called
38:49page optimizer pro. And the bottom line
38:51is we we do our own math and we've
38:53written our own algorithms and uh what
38:56we're looking at is how many times to
38:58put different words in different places
38:59so that Google will love you more. And
39:00it turns out not only Google loves that,
39:02but LLM search loves that. But the tool
39:05does get mathy and is technical. But
39:06what we've done recently is we built an
39:08AI version of me. So uh we it's taken my
39:12courses. It's watched my you know
39:14YouTube shows and and seen me on stage
39:17and um it can do most of the things
39:19within the tool. So you can actually
39:20just have a natural conversation with it
39:22and it will tell you what to do which is
39:24actually we kind of bridge that gap
39:26between like being really a tech nerd
39:28and you know having no technical
39:30experience whatsoever but both want even
39:33if you're a tech nerd you can still use
39:34it because you can just say hey I've got
39:35this page this is what's important to me
39:37it'll run it for you it'll tell you what
39:39to you ask what to do and so um
39:42>> we're really proud we're proud of that
39:45>> uh page optimizer pro
39:46>> yeah yeah pilot yeah you're right yes
39:48pilot is the name of me.
39:50>> Yeah. And it's an AI agent and it it
39:53runs SEO reports and recommends fixes
39:56>> within the tool. So, you just log in and
39:58it sits there for you and you can just
39:59have a conversation with it. It'll set
40:01everything up and it'll tell you what to
40:02do.
40:02>> How much of an SEO's professional work
40:05do you think AI agents will eventually
40:07be able to automate?
40:09>> None. Not without the expert there to
40:13tell it what to do. Um, and we've been
40:16running a test, running a baseline, um,
40:21uh, I want a thousandword article on
40:22what to see in Rome. And we've been
40:23running this case study for the last
40:25year and a half, almost two years now.
40:27And the scores are getting worse.
40:30Like they're not getting any better with
40:32that in terms of what we need for
40:34ranking purposes.
40:35>> I think in fact, I think AI is getting
40:37worse at writing. Yeah, I've been saying
40:39that. I've been saying that
40:40>> for for many months now. I'm like,
40:42Chacha BT was making better copy a year
40:45and a half ago.
40:46>> It's gone downhill. Yeah,
40:47>> it has gone downhill. Um, but the bottom
40:49line is you need somebody that knows
40:50what they're doing. They need to
40:52understand what the output should look
40:53like in order to get it correct. And uh,
40:57LLMs can't do it. The counterintuitive
40:59part is if you ask an LLM, hey, can I
41:01have a checklist for on page SEO? It's
41:03going to give you a pretty decent
41:04checklist, but then they assume because
41:08you can ask it for that checklist, as
41:09soon as you ask it to do SEO, it knows
41:11what it's doing and it does not. You
41:14have
41:14>> it'll give you a good checklist cuz I
41:16like I've seen it give like a lot of
41:18like outofdate information. Oh yeah, use
41:21FAQ schema. Like
41:23>> I'm not saying it's the best, but I'm
41:25saying it's it's not the worst. But then
41:27compared to the output you get when you
41:28ask it just to SEO an article,
41:30>> that's garbage.
41:31>> Yeah. Oh, for sure. For sure. For sure.
41:34Um, what do you what is the most
41:38So, we we've talked about important
41:40ranking factors, but what do you think
41:41the SEO industry consistently
41:43underestimates?
41:47>> Relevance.
41:47>> Underestimates.
41:49>> Uh,
41:50see, I don't believe in things like
41:52relevance or intent. Maybe that's my
41:54unpopular opinion. I don't believe in
41:57them at all. And the reason I don't
41:58believe in them is because what can I do
42:00to teach Google relevance or intent? I
42:03can do nothing. I can do a search. I can
42:06see the pages that Google likes and
42:07Google shows me what
42:08>> you you just described relevance. You
42:10said have the target keyword in these in
42:12these places. That's relevance.
42:14>> I think that's per like um indexing or
42:17maybe understanding. Is this relevant?
42:19Or maybe like is this like um
42:23Yeah, maybe we're kind of describing
42:25slightly different things, but I I
42:26really think of like um I never worry
42:29about relevance. I never It's never
42:31something I'm worried about in terms of
42:33of what I'm writing. Is this relevant?
42:34Because what I'm doing is I'm looking at
42:35Google, what Google is rewarding.
42:38I'm taking that as my baseline and I'm
42:40building off of that. I but it's nothing
42:43where I feel like, you know, you have to
42:44sit down with a pen and paper and figure
42:46out intent and relevance because Google
42:48shows you what those things are
42:49immediately and and you can build off of
42:51what Google is showing them. But there's
42:53nothing that I can do to Google to make
42:54it change its idea of relevance or or or
42:59>> Google shows you those things
43:00immediately.
43:01>> You do the search, those 10 that you see
43:04right there are the most relevant and
43:05they have all the intent.
43:07>> Well, what if people haven't really made
43:08content for that keyword? So Google's
43:10just getting the closest matches that it
43:12can.
43:13>> At this day and age, I think that'd be a
43:15very special keyword and I would like to
43:16know what you found.
43:17>> Really, dude? There's a lot of keywords
43:18that people [laughter] are not targeting
43:20that people are not targeting directly.
43:22>> Yeah. Not too many though. Not enough I
43:24think where you need to really worry
43:25about that. But you can also see that
43:27from baselines like if um all of a
43:29sudden every page is like 500 words or
43:31every page is or or it's a real mixed
43:33SER, maybe you could see it that way
43:35where it's like it's not consistent
43:37across what these results are. then
43:39maybe you'd have one there. But um aside
43:42from the madeup phrases that I go after,
43:44I'm not seeing too many out there that
43:46uh I think are are in a situation where
43:48like you don't have to worry about
43:50>> phrases that I got.
43:53>> My phrases are like, you know, I'm I'm
43:55looking at my desk and I'm like, I don't
43:57know, book, microphone, cleaner, rug.
44:01>> Okay. So things that people don't
44:03search,
44:04>> right?
44:04>> Okay. like that that those are my SERPS
44:07>> for for the tests. Yeah. Yeah. So So
44:10okay, an interesting how about this?
44:11like an interesting um an interesting
44:14question might be because it's always
44:16it's always important I think a lot of
44:18SEOs get way too caught up in the
44:22technicalities which you know SEO is
44:24very technical but at the end of the day
44:25you want your SEO to make money and you
44:27want to you want to go after SEO that's
44:29actually going to get you visibility and
44:31you could get too caught up in the
44:32technicalities rather than actually
44:34doing SEO that gets the results for the
44:37purpose that we're doing SEO for. So, if
44:40you had a gibberish page and you're
44:43going after a keyword that lots of other
44:44people are also going after and they
44:46don't have gibberish pages, do you think
44:49of maybe you maybe you optimized the
44:52page perfectly? So, you actually did hit
44:54number one, but do you think you would
44:57you would you would fall out because
45:00these other pages actually maybe they're
45:03better satisfying intent or like
45:05>> No. See, I I would argue that page, if
45:07it hits number one, has the most intent
45:08and is the most relevant.
45:10>> Well, let's say, okay, how about this?
45:12How about this? How about this?
45:13>> You win you win the prize.
45:14>> How about this? How about this? Maybe
45:15you maybe you publish it on a really
45:17authoritative domain that always starts
45:19ranking number one. So, like you like
45:22New York Times puts out like, okay, give
45:24me a keyword that could be competitive.
45:29>> Uh, payday loans. Payday is most
45:31competitive.
45:32>> Payday loans. and New York Times just
45:33puts out a some article that says like
45:36payday loans has the keyword in the
45:38right places and the rest of their
45:40content is Lauraum.
45:42Do you think of of so like if it's the
45:45New York Times I mean it's that's such a
45:47competitive keyword it might not
45:48naturally rank number one but like
45:50>> but it's going to do well right it's I I
45:52could see it hitting at least for a
45:54little bit
45:55>> and the clickthrough rate honestly and
45:56the clickthrough rate on that would be
45:57good because people are going to click
45:58on the New York Times to see
45:59>> they're going to trust the New York
46:00Times.
46:00>> Yeah. But it if the bouncer so the the
46:03the way that I think about it maybe
46:05maybe you don't think about it this way
46:06that's very interesting which is like
46:10I'm going to say pogo sticking it seems
46:11like you don't like that word but if
46:12it's if the
46:13>> oh it's fine
46:14>> yeah but if there's a lot of pogo
46:15sticking for a piece of content like
46:16that it's like Google's going to be able
46:19to see that and then oh yeah maybe okay
46:21fine the New York Times has a good
46:23history but like this page isn't isn't
46:25doing it
46:27>> possibly Um
46:31I don't know what the conversion metric
46:34would be
46:36to say whe whether it is or is not
46:38satisfying the query from the from the
46:41user metrics as we've discussed. Um that
46:44seems rather tricky to me. Uh
46:47>> it's probably relative.
46:49Yeah, it's relative
46:52probably and how difficult the keyword
46:53is because I imagine again payday loans
46:56is going to be a SER that's monitored
46:58heavily might even have human
46:59interaction like that's the kind of SER
47:01that actually might have humans looking
47:03at where I can see it being knocked away
47:05whereas like rhinoplasty SERs for random
47:07cities across the globe probably fewer
47:10moderators yeah are looking at that
47:12hence I've had pages that have been on
47:14page one for seven years you know like
47:16that's
47:17>> and how much and how much traffic do
47:19these pages which is good.
47:21>> Oh, they're getting like 30 40 clicks a
47:23month easy.
47:24>> Yeah, really.
47:24>> And and and you can even see the Lauram
47:27in the in the description in the SERs,
47:29which is probably taking down the CTR.
47:33>> And
47:35do you are you giving like um is there
47:38like any call to action above the fold
47:40that maybe maybe people even aren't even
47:42reading the Laura Mipsum? Maybe like
47:43there's a reason why people are they're
47:45just not even looking at it and they're
47:47clicking and moving on.
47:49Initially all the forms worked and we
47:50did have a working phone number and um
47:52and that was for a couple years and we
47:54got approximately zero form submissions
47:56and zero calls. So then uh I don't think
47:59the forms work anymore. The the numbers
48:00are certainly dead uh at this point. But
48:03uh yeah, it got zero conversions. People
48:06clued in quickly. Actually, you know
48:08what's really funny is I got a call from
48:11Yelp. Was it Yelp? I think it was Yelp.
48:14And there was like, "Hey, this is Sally.
48:15I'm your Yelp rep for your region and
48:17I'm calling about uh rhinoplasty Plano
48:20and I was like, "Go on." And she goes,
48:22"So, I'm looking at your website here.
48:24Are you guys targeting a foreign
48:25audience?"
48:28[laughter]
48:28>> Cuz it doesn't look like this is
48:30English. And I was like, "Yep, we're
48:31going after the uh the foreign audience
48:33there in in in uh Plano, Texas that
48:36definitely only speaks ancient Latin."
48:40>> That's crazy.
48:42>> So, all right. How about so if
48:46Okay. So are basically are you are you
48:49saying that SEO is just on site and just
48:53like just external authority with no
48:56user signals aside from click-through
48:58rate?
48:59>> No, I'm sure user signals are part of
49:01it. It's just I I can't tell you exactly
49:02what they are or how they come into
49:04play. And and again, I think there are
49:07there it requires multiple factors to
49:09hit and it's probably not just binary.
49:12If A then B. I'm sure there are multiple
49:14things that have to hit and then they
49:16would consider those but it's just
49:18really hard to nail down. I I would I
49:20would argue for sure, you know, like and
49:22they are things that I would monitor and
49:24worry about and try to improve uh when I
49:27could, but um it's just from a from a
49:30from a perspective saying like I can
49:31tell you certainly that it's this I
49:33can't and and it just it's been
49:35impossible to it's been impossible to
49:38nail down in a testing environment.
49:39>> Yeah. Um, there are a lot of agencies
49:42who listen to this podcast. What's an
49:44SEO best practice that you believe is
49:46completely useless but agencies keep
49:48recommending anyway?
49:51>> Probably going after like um core web
49:53vitals if anybody's still getting on
49:55that.
49:56>> I've I've seen more sites tank going
49:58like trying to shave off you know 0001
50:01whatevers than sites that ever improve
50:04from that. you do need to be you need to
50:07be within ranges of course you know as
50:09we talked about before you probably want
50:10to site the loads quickly would be my
50:12guess but um if I'm ever clueing in on
50:15um actual metrics I'm looking at time to
50:18first bite as I think there's a there's
50:20a factor in there and um total blocking
50:23time uh total blocking time every time I
50:26see a site get hit I can look at their
50:28total blocking time and eight times out
50:30of 10 it's deep in the red on any site
50:33that ever gets taken out of Google's
50:34index or or tanks or falls under some
50:36sort of penalty. Their total blocking
50:38time is always bad. So I always look at
50:40those two as in terms of metrics, but
50:42the rest of the core vitals
50:44I think they can
50:46money could be spent on other things.
50:48You know, they might be a lot more
50:49productive
50:50>> and you and honestly like anybody can
50:51see this themselves like look at the top
50:53pages that are ranking and a lot of them
50:55or the the top sites that are ranking a
50:57lot of them have very poor core web
50:59vitals.
51:00>> It's true. Yeah. Uh but look at their
51:03time to first bite and look at their
51:04total blocking time. About those are all
51:06right.
51:08>> That makes sense.
51:10You um this is interesting. Okay. You
51:11you you recently partnered with GoDaddy
51:14on experiments suggesting that
51:16descriptive domain extensions can
51:18outperform traditional.com domains. Uh
51:22so yeah, could you could you share that?
51:24And if you were launching a new business
51:26today, would you actually choose
51:27something like florist over.com for an
51:30SEO advantage?
51:32The biggest thing is actually there's no
51:34disadvantage.
51:35I think that's I think that's the most
51:37important takeaway. So if you could get
51:39something branded where you couldn't get
51:42it with the.com, I would be very
51:44comfortable going with the descriptive
51:46domain to get my brand. Um we also found
51:50though that
51:51>> I might skip that one.
51:54>> No, I don't know. Because you remember
51:55like what 10 years ago when everybody
51:56spammed those
51:58>> into oblivion? Yeah. And um in fact,
52:01John Mueller literally said he said on
52:03the SEO subreddit, he's like, "Yeah,
52:05XYZ's might actually perform worse with
52:07SEO, but he didn't say he said it's
52:10because like some uh some TLDDS
52:16might not like they might be blocked on
52:18social media platforms or you might not
52:21be able to get an email address with
52:22them." Um, and uh, and so that can
52:25create very that can create business
52:27challenges that just make it harder to
52:28do SEO, that make it harder to do
52:30marketing. I I literally had that I I'll
52:32tell you, I had this happen to me before
52:35Elon bought uh, X. We had a club website
52:40>> and and Yeah. And so it was like the
52:42weirdest thing. If somebody and we we
52:44had like a share feature that lots of
52:46people were using. It was an it was an
52:48app. And if somebody shared our club
52:51domain and they they responded to a post
52:55on old Twitter, their response, no
52:58matter who they were or how
53:00authoritative the domain was, because we
53:02we looked at like other clubs that had
53:04that were like super well-known, super
53:06authoritative, it was always hidden. It
53:09was always shadowbanned on Twitter. any
53:12any link that mentioned a dot club in
53:16response to somebody on Twitter would
53:19get shadowbanned. And so we were doing
53:21this like big promotion with these huge
53:23organizations where part of the thing
53:25that you were supposed to do was share
53:27um your your link to to your challenge
53:30on our site and no one was seeing the
53:32links because it was all shadowban. And
53:34so old Twitter had just built a blanket
53:39shadowban for any dot club website. And
53:43uh yeah and and so I I saw that I saw it
53:46was the laziest develop like the laziest
53:48programming possible [laughter] like no
53:51no ifs like
53:53>> like literally just yeah just block any
53:56do club any club gets blocked.
53:59>> Oh it's crazy. Uh well I didn't test
54:02that within Twitter. uh but uh what we
54:06were doing though is uh we were looking
54:08at is there any potential say relevancy
54:12for for example like if you're uh have a
54:14site that's you know florist and you
54:17have a page about roses will that will
54:19that beat another page with a and and
54:22there's not too much there but what we
54:24did see is that um partial EMDs and EMDs
54:28that include the fun descriptive part
54:31are considered so like if that's part of
54:34your target keyword. You can get an EMD
54:36with the descriptive um uh um extension
54:41uh the TLD and and that will give you an
54:43edge.
54:44>> Um
54:46>> uh
54:48uh you you live in you know Plano Plano
54:52Florist is is the name of you know the
54:55term you're going after that will give
54:56you an edge that counts as an EMD.
54:58>> So wait Plano.florist florist.
55:01>> Correct. And then if the term you're
55:03going after is plano florist, you can
55:05get an edge with that.
55:06>> That's awesome.
55:08>> I'd also found that as a kind of um a
55:11sub benefit is a lot of the descriptive
55:13tldlds indexed faster than the dotcoms.
55:16Now, they might not have won in the end
55:17or they might have all tied in the end
55:18as kind of a you know, there really
55:20wasn't a great result, but what I did
55:21notice is the descriptive uh tldds will
55:23index faster. So, that can be an edge
55:25when you think about if you're pumping
55:26out content, you know, especially if
55:29you're very time-sensitive type things.
55:31Black Friday is coming up or, you know,
55:34like you're one of those you're just
55:35trying to stay ahead of different
55:37launches for products and stuff like
55:38that, you might have a chance to index
55:40faster and that that can mean a lot of
55:42money u based off of the uh of the
55:44descriptive TLDD that you have. Um, but
55:47the MDs and the PMDs were were were big
55:49factors. But the biggest thing across
55:51all the other tests is they didn't harm
55:52you. Like, you know, they they weren't
55:55taking anything away by not having the
55:57com. You know, we get we get an even
55:59split of results. You know, half the
56:01dotcoms win, half of the descriptive
56:03TLDDS win. It's just like, yeah, it's
56:05Google treats them the same. There other
56:07there are other factors they're looking
56:08at beyond just, you know, going saying
56:10that is the best.
56:11>> Why do you think um they were indexing
56:14faster?
56:15That's a great question and I I've
56:17thought about it for a while, [laughter]
56:19you know, like I've really thought about
56:20it and I I couldn't tell you. Uh is does
56:25the algorithm have something built in
56:26that nobody knew about? You know, one of
56:28those weird quirky things and it's just
56:30like it just prefers that. I don't I
56:32couldn't tell you, but um but it
56:34happened across the board repeatedly.
56:37You you al you you also found excuse me
56:39you also found that shorter domains and
56:41shorter extensions sometimes um was it
56:44what you were talking about where Yeah.
56:47So you you the you you saw that shorter
56:50domains and shorter extensions actually
56:52achieved rankings sooner.
56:54>> Yeah. Um that's almost always the case
56:57even if you're doing with too. Um, one
57:01thing about it though, there's a
57:02confusion that happens in the industry
57:03where they think um, like a shorter URL
57:08like so like.com/
57:10you know, Bob versus you know
57:15Bob whatever you're going to say. That's
57:18not that's not it on on that side of it.
57:21It's actually closer to the do the
57:23the.com is the shorter URL less folders
57:27to to get through. those will index
57:29faster and those have a better chance at
57:30ranking. Um but on the other side uh you
57:33can get a lot of action out of um
57:35smaller uh shorter domains for sure.
57:38>> Do you avoid uh subfolders when you're
57:40doing SEO or do you um do you keep and
57:43so you keep everything off the root do
57:45off the root domain?
57:47>> Yeah, that gives you the best chance to
57:48rank. Now going uh one folder deep is no
57:50problem. So having a category and then
57:53and that's that's totally fine. where I
57:54started to have a little bit of
57:55heartburn and where you start to see a
57:56difference in things is when you're
57:57going two folders deep. So, you know,
58:00marbles, pink marbles, that's pink
58:02marble. There's no reason to do that.
58:05Take it back one. You know, you you
58:07start to harm yourself when you go that
58:08deep.
58:08>> What about for very huge websites?
58:14>> So, the thing is is most people get
58:15locked into having to do that because of
58:17their menu. You know, it's kind of like
58:19built into their template. They have to
58:20have these categories and the menu is a
58:22fiction. The menu is there for humans.
58:25So do your menus however you want to do
58:27them and categorize them however you
58:28want to categorize them. But then when
58:29somebody gets to the page, get it as
58:30close to the do. The user doesn't care.
58:33Especially you've got breadcrumbs
58:35properly. You got a menu properly,
58:37you're fine. But um you'll do better
58:40like if if you're in a situation where
58:41you've got some massive e-commerce site,
58:42try to trim as many folders as you can.
58:46>> Yeah, that's great. Um and yeah, the way
58:48that you're described it is this is like
58:50this is for humans. So, how are you
58:52thinking about like with a huge site,
58:54how do you think about funneling
58:55authority to different pages or maybe
58:58different hub pages?
59:00>> Well, you usually have like category
59:02type pages or location pages or service
59:05pages that do list everything. That's
59:07the easiest way because that's going to
59:09spread through right there. And you need
59:10them for humans anyway. That's just good
59:12for humans to have. But, um, I think
59:13it's good for crawling purposes. And,
59:16um, you could send links to those
59:17things. You could do activities, other
59:19things like you could have a a one of
59:20those hub type pages that has nice
59:22information on it and good information
59:24and it makes sense to do campaigns
59:26towards those, but then that's going to
59:27spread throughout that entire category
59:29or all of those services or all of those
59:31locations or or however you're or
59:33whatever that page is supposed to cover
59:34for you. So that's usually the easiest
59:36way to do it is through that. In your
59:38tests, this is an interesting question.
59:40In your tests, have you found anything
59:42that's good for SEO but not good for
59:45humans?
59:49>> [laughter]
59:49>> Good for SEO but not for humans. Uh well
59:52probably a page of Lauren Mipsum is not
59:54great for humans.
59:55>> Yeah but the but the principles of
59:57having like the relevant terms in that
59:59page still that's
1:00:01>> something good for SEO not good for
1:00:02humans.
1:00:04>> Uh well I mean hidden text still works.
1:00:08Um uh having fake drop downs still
1:00:11works. You know none of that's going to
1:00:13be great. And you could have content
1:00:15like um you know it appears within say
1:00:17the search results you know because
1:00:19Google's pulling it in from the page and
1:00:21then it doesn't exist on the page at
1:00:22all. That's a pretty miserable
1:00:24experience for humans where basically
1:00:26the the the lure that got them in like
1:00:28oh yeah this is the page I need to see
1:00:30cuz I think that's the information I'm
1:00:31looking for and they get to the page and
1:00:33they can't even find it because it's
1:00:34completely hidden on the page and was
1:00:35only there for ranking purposes.
1:00:39>> Do you have any favorite niches to do
1:00:40SEO in?
1:00:43just I want to get paid.
1:00:46[laughter]
1:00:48>> Anybody that's got cash
1:00:49>> or like maybe maybe niches that you that
1:00:52you more commonly do SEO in.
1:00:55>> Uh I've been very fortunate. I've been
1:00:59doing SEO full-time since 2012 and I've
1:01:02spanned just about every niche. I have
1:01:04not personally done a lot in gambling.
1:01:05I've done a little. Um that's probably
1:01:07the only niche maybe I don't have a ton
1:01:09of experience, although I do have a fair
1:01:10amount. Um, but after that I've done I
1:01:14feel like just about every niche you can
1:01:17imagine and they all have their own
1:01:18little challenges. Most of the time it's
1:01:21the biggest issue is bureaucracy, right?
1:01:24Um it's nice when you're dealing with a
1:01:26marketing team rather than uh a founder.
1:01:31Uh so like uh startups for example often
1:01:33the founder you know they want to touch
1:01:36and it makes a lot of sense or some of
1:01:37the small smaller mom and pops they
1:01:39they're much more involved and that
1:01:41makes a lot of sense. Um it's just you
1:01:42have to take those things into
1:01:43consideration when you're you're putting
1:01:45your plans together how you're
1:01:46presenting things um the time you're
1:01:47going to put in and stuff like that. So
1:01:49all those just things to navigate. None
1:01:50of them are necessarily bad. I I don't
1:01:52have any problem working with with with
1:01:53anybody, but you just have to be mindful
1:01:55of like where those companies are coming
1:01:57from, their perspective, their needs, uh
1:01:59how much they need to get out of this,
1:02:01you know, is that kind of and just you
1:02:03kind of have to just be mindful of of
1:02:05the different corporate politics or
1:02:06small business politics or whatever
1:02:08politics. And once you can get through
1:02:10that, you I think you can handle just
1:02:11about any niche is fine.
1:02:12>> Have you have you found that SEO changes
1:02:16when you're going from a non YM niche to
1:02:19a YML niche?
1:02:22Uh it seems that uh when you're playing
1:02:26in some of the dirtier niches, the more
1:02:29like payday loans kind of thing, it does
1:02:31seem that um the competition is much
1:02:33more willing to try to take your site
1:02:35down.
1:02:36>> Whereas like, you know, if you if you
1:02:39have,
1:02:40>> you know, bakeries are not going after
1:02:42each other online. you know, they might
1:02:45have beef somewhere else, but that's
1:02:46more on the streets than uh than um
1:02:49trying to do something actually
1:02:50malicious, but there are some niches
1:02:52that you do see where it feels like
1:02:53people are trying to be malicious and um
1:02:55that's unfortunate and u but that does
1:02:58exist in those types of
1:02:59>> Well, I had Yeah. So, the first time
1:03:00that James Douly came on this podcast,
1:03:02he's like, I Yeah, what?
1:03:05>> James is great.
1:03:06>> Yeah, James is awesome. and he's like um
1:03:08he's like yeah like I don't like playing
1:03:10in some of these affiliate niches that I
1:03:12used to play in because like these fake
1:03:14DMCA requests they they've they just it
1:03:16just makes it not fun.
1:03:18>> Like he's like some of the some of the
1:03:19dirty things that people will do that
1:03:22that like man I just I don't want to do
1:03:23that. It's just he's like it makes it
1:03:25not fun.
1:03:27I was working for a mattress company and
1:03:29they accidentally let their GMBB slip
1:03:32and um the next thing you know one of
1:03:34their competitors claimed it and said
1:03:36that there wasn't a mattress company
1:03:37anymore. It was a piano tuner.
1:03:39>> Wow. Wow. [laughter]
1:03:42>> They completely changed the categories
1:03:43to piano.
1:03:44>> Wow.
1:03:47>> Who would take the time? [laughter]
1:03:51>> But like those niches do. So
1:03:53>> yeah. Yeah. Yeah. Let me ask you, um,
1:03:55when the when Google's May API when
1:03:58Google's May 2024 API leak happened, was
1:04:02there anything that surprised you in
1:04:03that or were you just like, yep, all par
1:04:06for the course here from what I'm
1:04:07seeing?
1:04:09>> More than anything, it was just like,
1:04:11yeah, we've been [clears throat] we've
1:04:12been screaming this. You know, my
1:04:13favorite leak though was the Yandex one.
1:04:15That's the one I liked a lot more. And
1:04:17the reason I like that is because, you
1:04:20know, the guys that put Yandex together
1:04:21were Google engineers and they were
1:04:24trying to approximate Google and as
1:04:26close as they can for the least amount
1:04:27of money. And we found out that they're
1:04:30using BM25. And BM25 is just TF plus
1:04:34word count. And I've been screaming that
1:04:36for years. And I was like, vindication.
1:04:39Like that was my favorite part because
1:04:41that means that formula is going to get
1:04:43you very close to what Google is trying
1:04:45to do. you know, close enough, right, in
1:04:47terms of like the quality of results.
1:04:49And I'm like, that's I've been screaming
1:04:50that it's basically TF plus word count
1:04:53for years. And that to me was actually
1:04:56much more validating than anything I saw
1:04:58on the Google. The Google was just like,
1:04:59all right,
1:04:59>> go deep into that for for noobs who
1:05:01haven't even heard of that.
1:05:03>> TF is a way to sort um results, you
1:05:08know, um uh it's it's the term frequency
1:05:11uh divided by the inverse document
1:05:12frequency. So how many times a term
1:05:13shows up versus how many documents are
1:05:15available is kind of the idea and it's
1:05:17not anything that uh anybody's ever used
1:05:19directly I think within a search engine
1:05:21but it approximate it approximates
1:05:23things very closely to what search
1:05:24engines are actually doing because
1:05:25humans only think in a certain way. We
1:05:27can only categorize stuff in so many
1:05:29ways that make sense so that the vast
1:05:31majority of people can search properly
1:05:33and get their answers. So basically
1:05:34these types of ways to do search are
1:05:36going to be very very similar. So then
1:05:39BM25 is just another system, but it
1:05:41basically takes TF and then it adds the
1:05:43concept of word count. So like how long
1:05:46is this copy shorter or longer? And
1:05:48that's going to factor in it's going to
1:05:49weigh into what is the most relevant
1:05:51result here. And it's going to say like,
1:05:53hey, you know, all of these are about
1:05:551500 words and all of them use uh this
1:05:57term about this many times, more so than
1:05:59the rest of the of the corpus that's
1:06:01available and they're all over the
1:06:02place, so we like these the best. And um
1:06:04I think it's a really good way to
1:06:05approximate how most search engines are
1:06:07doing it. But then when the Yandex one
1:06:10came out and they were showing that this
1:06:11is that's exactly what they're using, I
1:06:13was just like, "Yes, I knew it. I knew
1:06:16it."
1:06:16>> Awesome.
1:06:16>> I knew it.
1:06:17>> Do you think Google favors any CMS's
1:06:20WordPress, Shopify, Web Flow, other
1:06:22platforms?
1:06:23>> No. And that's what's actually great. Um
1:06:26they I think they're completely platform
1:06:28agnostic. But the one thing I will tell
1:06:30you is that different platforms have
1:06:32different amount of hidden words and
1:06:34words that Google is counting and a lot
1:06:36of people aren't aware of that. So they
1:06:38might switch from one platform to
1:06:39another, keep everything identical, keep
1:06:41their URLs exactly the same, but this
1:06:43platform over here was actually doubling
1:06:45your word count and you didn't know like
1:06:47maybe in the CSS it was doing like a
1:06:48mobile and a desktop version and then
1:06:50you cut over to this next one and it
1:06:52doesn't do that. You just lost half of
1:06:54your content on your site
1:06:56and then those sites tank.
1:06:57>> Wait, wait, wait. And wait, why why say
1:07:00that again? What what caused you to lose
1:07:02half the content
1:07:03>> in like one example that I I saw? So
1:07:05like one template on one platform was
1:07:08presenting a desktop view and a mobile
1:07:11view and that essentially doubled the
1:07:14word count on every page
1:07:16because Google was reading all of that.
1:07:19So then when they went to a different
1:07:20platform that wasn't handling things
1:07:22that way, they lost half of their
1:07:24content. So like every page that was
1:07:262,000,
1:07:27>> right? every page that was 2,000 words
1:07:29just went to 1,000 words. Half of their
1:07:32content just got cut off the site. And
1:07:34>> even though even though it was basically
1:07:36duplicate content, but but yeah,
1:07:39>> that's not how it's being seen.
1:07:40>> Yeah. Right. Mhm.
1:07:42>> Yeah. And so those are things you need
1:07:44to be mindful of on platforms. Some
1:07:46platforms have a little amount of of
1:07:48hidden words that are going to be
1:07:49counted. Some have a lot. And it's not
1:07:51good or bad. It's just that's how those
1:07:53platforms work. And it's something you
1:07:55need to be mindful of when you're
1:07:56creating content. How is my like uh
1:07:59let's say you decide you need 1500
1:08:00words. Well, how many words are in your
1:08:02template now? You probably don't need
1:08:041500. You probably need to write a
1:08:05thousand because there's going to be 500
1:08:07on the word on the page just to start,
1:08:10you know? So, you need to think about
1:08:11where those counts are coming from. And
1:08:12then al you also need to think about
1:08:13like what terms the are there just as a
1:08:17base and then which ones are they adding
1:08:19once the content gets put in place as
1:08:21well. So you your word counts might be
1:08:24very different from what you think they
1:08:25are because like in your WordPress, you
1:08:28know, little thing or whatever you're
1:08:29using or or your Google doc, it says
1:08:31there are this many words on the page,
1:08:32but when it actually goes live to what
1:08:34Google is seeing, what other search
1:08:35engines are seeing, there's actually a
1:08:36significantly different number.
1:08:37>> Yeah. And you could so you could just
1:08:38view page source and give this to Claude
1:08:41and and just ask.
1:08:43>> Claude actually does a pretty decent job
1:08:44with it because um a free little tool
1:08:46that I like is um SEOQ. Uh they do a
1:08:49really decent job of showing you the
1:08:51words that Google will be counting. It's
1:08:52a free
1:08:53>> I used to use SEOQ. Yeah, SEOQ's great.
1:08:55>> I do it as a sanity check, you know,
1:08:58like because like pop counts in a
1:08:59certain way very close to how that
1:09:01counts and then cloud will count
1:09:02correctly as well. Uh but I'll check
1:09:04across those like when you see like cuz
1:09:06sometimes you'll see like man there's
1:09:07only a thousand words on this page but
1:09:08pop is saying there's 4,000. What's
1:09:10going on? I'll go to SEOQ and like okay
1:09:12SEOQ has it. Ask Claude. Claude sees it.
1:09:15So then you know like there's something
1:09:16going on with this template where it's
1:09:18just churning out, you know, the content
1:09:20again and again and again for SEO
1:09:22purposes.
1:09:23>> Which CMS's have the most hidden text
1:09:26and which have the least?
1:09:28>> Uh I don't I don't have a a a specific
1:09:31number. There were ones like um a lot
1:09:34when the the page builder craze came out
1:09:38uh when there was that huge flood a few
1:09:40years ago with all these page builders
1:09:42coming out every day. A lot of those uh
1:09:44had a lot of issues. They a lot have
1:09:45cleaned up. I don't have any in
1:09:47particular that that might be better or
1:09:49worse. I think they've maybe hopefully
1:09:52taken care of some of the problem or at
1:09:53least people may be more aware of the
1:09:55problem. One thing that does come up
1:09:57with a lot of those page builders though
1:09:58is that um if you're using like drop
1:10:00downs or sliders or some of that, they
1:10:02often strip out the HTML and they'll
1:10:05automatically by default put your text
1:10:07into an empty div. Google will
1:10:09eventually get there, but you're really
1:10:10shooting yourself in the foot because it
1:10:11takes Google a long time to get there
1:10:13and will often discount text that's not
1:10:16properly within HTML tags. So, you might
1:10:18think in your mind, oh, I put an H2 here
1:10:20and then I've got all this this P copy
1:10:23and paragraph top copy and um it turns
1:10:26out you don't. It's all just empty text
1:10:28on on on the page. And so, you're not
1:10:30getting near the the credit that you
1:10:32should for the work that you did. Google
1:10:33will eventually read it, but it will
1:10:35take a long time.
1:10:35>> For people who want to like test that
1:10:38themselves to make sure it's not an
1:10:39issue. How could they do that?
1:10:42>> Well, even within Search Console, you
1:10:43can get a snapshot of what Google can
1:10:45see within the HTML. That's an easy way
1:10:47to do it. Um, if you're not sure if
1:10:49Google is reading content, um, put a
1:10:52unique phrase, uh, 7, eight word phrase
1:10:55that's unique, that's only on there, put
1:10:56it in that element, that the page index,
1:10:58and then search for that phrase. Your
1:11:00page should come up. And if it isn't,
1:11:01that means there's a problem with Google
1:11:02reading it. So that's like if you have
1:11:04like a JavaScript element and you're not
1:11:06sure if Google's seeing the text in
1:11:07there, uh, or if you're not seeing think
1:11:09maybe it's not being seen properly, you
1:11:10can do those types of checks on your
1:11:12page, um, to see if it's if it's working
1:11:14or not.
1:11:15>> Yeah. So you would do site colon your
1:11:16site.com or fflorist and then uh, space
1:11:20and the the thing that you put the
1:11:22>> or just do the just do the search for
1:11:24that like that eightword phrase. Just do
1:11:25it in quotes and then make sure your
1:11:28page comes up as it should.
1:11:30>> Yeah. Yeah, Kyle, do you think SEOs are
1:11:33too obsessed with schema?
1:11:36[laughter]
1:11:38>> Probably. A schema. What's great is
1:11:41schema's always been the next best thing
1:11:42or like it's always going to take off
1:11:44next year. You know,
1:11:46>> it's going to take off next year. It
1:11:49>> it's always next year's next year's
1:11:51schema is here. I'm telling you, it's
1:11:52coming.
1:11:52>> 2027 is the year of schema, everybody.
1:11:54>> The year schema. It's the You heard it
1:11:57here first.
1:11:58uh [laughter] it's always about to go.
1:12:01Um
1:12:04as a ranking factor in and of itself,
1:12:06it's not. But there are a lot of things
1:12:08that can help with. You know, there are
1:12:10different elements on the page that can
1:12:11now show up in SERs, different SER
1:12:13elements. Anytime there's a new SER
1:12:15feature, they often rely on schema uh to
1:12:18show it. So, if you don't have that
1:12:19schema on your page, then um you're not
1:12:21going to appear. So you're sacrificing
1:12:24at a minimum real estate surf uh real
1:12:28estate within that SER. Um you're you're
1:12:31potentially sacrificing that. It's not
1:12:33that hard to do anymore with AI
1:12:35honestly. So
1:12:38uh do all the things as uh like Matt
1:12:40Diggity says, just do all the things.
1:12:42Don't worry about it. Just get it done
1:12:44and um and then don't worry about it.
1:12:47But uh is it gonna
1:12:49>> But the thing is SEOs will work. The
1:12:50thing is SEOs
1:12:52A lot of people don't like to say do it
1:12:55because SEOs do worry about it like like
1:12:57you know how SEOs are. We go we go deep.
1:12:59We we obsess like some of the best SEOs
1:13:01are hyper obsessive. So you you might
1:13:04say listen
1:13:06>> dude or dude that just spend just spend
1:13:09five minutes on it and that's it. But
1:13:11like it's not going to be five minutes.
1:13:12It's going to be three hours.
1:13:15>> Yeah. I feel like you could do three
1:13:17hours on something else.
1:13:18>> Yeah. Yeah. Now, I know I have a lot of
1:13:20SEO friends that do a lot of great work
1:13:21with it, and I I don't want to say
1:13:23anything,
1:13:25but from my perspective, it's okay. It
1:13:29It's worth it for the potential to show
1:13:31up in other things, for sure.
1:13:33>> Yeah. All right. So,
1:13:34>> and maybe and maybe ASA is going to use
1:13:36it more, you know, that's always
1:13:37possible, too.
1:13:39>> No, but I mean, it's possible.
1:13:42>> Exactly. Why? Like, why why would they
1:13:44use it more?
1:13:46Maybe just um uh cost saver.
1:13:49Uh if you only read the schema, not the
1:13:51the whole page, you'd save a lot of
1:13:53money.
1:13:54>> Yeah, but like but but what So the
1:13:56content on pages that have schema is is
1:13:59better than the content on pages that
1:14:00don't.
1:14:01>> That's that's what
1:14:02>> No, I'm just saying they they just might
1:14:04decide to be super cheap about it. Just
1:14:06read the schema and not the content.
1:14:09>> That'd be interesting.
1:14:11>> I would do it if I were them.
1:14:12>> Would you?
1:14:13>> I'd save money. Yeah, I'm cheap.
1:14:15>> Yeah, but in a in a like [laughter]
1:14:17hyper competitive niche where like
1:14:19people are trying to where these
1:14:21companies are trying to have the best
1:14:22output,
1:14:24>> then they should have taken the time to
1:14:26do the schema as Kyle Roof told them
1:14:27[laughter] on on your on your show.
1:14:29>> Yeah.
1:14:31>> Um so yeah, we were talking about uh we
1:14:34were talking about AI content before um
1:14:36and you you explained why AI content
1:14:40underperforms.
1:14:41So, what are um and we were talking
1:14:43about this a little bit, but I wanted to
1:14:44go deeper into it because honestly, it's
1:14:46just such an important topic because
1:14:48look, you could tell people not to use a
1:14:49lot of AI content. They're still going
1:14:51to try to do it. So, what what are the
1:14:54right ways to do it and the wrong ways?
1:14:58The wrong way is just say, "Hey, write
1:14:59me this article and SEO it please." You
1:15:02know, and then some people like, "Well,
1:15:03I prompt really well." Okay, well, what
1:15:05are you using as your benchmarks?
1:15:08Prompting without any kind of benchmarks
1:15:09is also going to be problematic. So the
1:15:11right way to do it is have benchmarks
1:15:12for what you need for each individual
1:15:14piece of content that's going after that
1:15:16keyword. And however you get those
1:15:18benchmarks is is going to be better than
1:15:19anybody that that's not using any
1:15:21benchmarks whatsoever. You could use the
1:15:24worst tool in the history of tools. You
1:15:26could make your own tool that is
1:15:27actually awful, but you're still going
1:15:29to be way ahead of people that aren't
1:15:31doing that at all. So you need to have
1:15:32some form of benchmarking of what we
1:15:34need to get into this article to uh to
1:15:38show you know search engines that this
1:15:40is worthwhile. Uh so they can understand
1:15:42it, they can index it and then they can
1:15:44rank it. That's the key. So uh it's not
1:15:47just like I gave a really good prompt
1:15:48but you also have to give it benchmarks.
1:15:51It needs to get this these parameters
1:15:53for this content. What's really great
1:15:55about AI too though is it writes really
1:15:56good within those within that framework.
1:15:59you get really great content either way.
1:16:01It's very readable, but you can get it
1:16:03so that it satisfies also what search
1:16:04needs. It's just taking the time to set
1:16:06that up. But it's so much easier just to
1:16:08kind of mash buttons and then get output
1:16:12in a few minutes rather than taking the
1:16:13time to do that.
1:16:14>> What do the what are some good
1:16:16benchmarks? Like what do good benchmarks
1:16:17look like to you?
1:16:19>> Those benchmarks are um how many words
1:16:21do I need on this page? You know, if
1:16:23you're if you're selling something
1:16:26rather niche, you know, you you might
1:16:27only need, you know, 750 or thousand
1:16:29words. If you're trying to rank for
1:16:31something financial, you know, that
1:16:33might be a 4,000word article. You know,
1:16:35that that those are different things.
1:16:37So, I want to know how many words do I
1:16:39need on this page. And then how many
1:16:41times to put that target keyword in very
1:16:43specific places, its variations in very
1:16:45specific places. And um then what I what
1:16:49I need contextual term wise within the
1:16:50paragraph text. Those are all numbers.
1:16:53They're very
1:16:54it's they're easy to count, you know,
1:16:56but it's not just it's not just the
1:16:58average, too. Some that's another thing
1:16:59that people miss. They're like, "Oh,
1:17:01you're just doing averages." No, no, no,
1:17:03no. There are times when you want to do
1:17:04a little bit more. There are times when
1:17:05you want to do a little bit less. There
1:17:06there are times when you want to do what
1:17:08they're doing. And there are different
1:17:09indicators on the page that show you
1:17:11when to do that. Now, it's often
1:17:13depending on what else is happening
1:17:14within that SER, but uh it's not just
1:17:17taking like a page one average and
1:17:19you're going to be good. That's not good
1:17:20enough either. Um
1:17:23uh whatsoever. You know, you need to
1:17:24think about what is the range that's
1:17:26available for me to use and what's an
1:17:29appropriate range for these types of
1:17:30terms in these places.
1:17:33you uh you argued that your existing
1:17:36content silo strategy was already well
1:17:39suited to AI query fanouts.
1:17:42>> Yeah, it seems to be working really
1:17:44well.
1:17:44>> What what's the what's your content silo
1:17:46strategy?
1:17:48>> Well, it's the idea of um if you have a
1:17:50target page, you know, whatever main
1:17:52concept you're going for, um you want to
1:17:54write out pages that will support it.
1:17:56So, um, one of the keys is you want to
1:17:59figure out how strong your site is
1:18:00because there's a there's a level that
1:18:02Google will just rank you, right? You
1:18:04write a piece of content, it just ranks.
1:18:06So, that's your level. So, you want to
1:18:08identify those types of levels because
1:18:09as you gain impressions and keywords and
1:18:11all that, those pages gain strength. And
1:18:13then, if you have interlinking done
1:18:14properly between those to a target page,
1:18:16you can now take advantage of the
1:18:18strength that you've built from your own
1:18:19site and it it takes the place of some
1:18:21of the backlinks you will need to rank.
1:18:24Um, but I like to say we did, you know,
1:18:27three supporting pages. A links to B. B
1:18:30links A and C. Uh, C links to B, and
1:18:34then they all link up to that target
1:18:35page. And then as they grow, they pass
1:18:38that juice to each other. They pass that
1:18:40juice to the target page. But about
1:18:42three uh internal links of the same
1:18:45value are equal to one external link
1:18:46over that value. So just as a random
1:18:49number, your site's strength of however
1:18:52you like it is 30. If you have three
1:18:54internal links at 30, that's equal to
1:18:56one backlink of 30. And so you can put
1:18:58out content on your site to take
1:19:00advantage of that and limit the amount
1:19:02of backlinks you need. It's safer, it's
1:19:04faster, it's more effective, and then
1:19:06you're actually getting traffic out of
1:19:07it, too, if you're doing it through
1:19:09supporting content that's interlin. Um,
1:19:11but then also when those questions get
1:19:13broken down, that's one of the things
1:19:15you can use within those silos is is put
1:19:18those questions as the breakdown within
1:19:20that supporting uh page, that silo
1:19:22concept. So then you take advantage of
1:19:24when the LLMs are are are going through
1:19:27the site, they see that you've got those
1:19:28answers there. You're providing that
1:19:30you're that's your authority, that's
1:19:32your relevancy.
1:19:33Getting back to the earlier thing, um
1:19:35but you can you can see it within um
1:19:38within that content. And that's it's
1:19:39it's worked well for Google. It works
1:19:41really well with the LMS. And with now
1:19:43this going on, it's um it's it's working
1:19:45better than it ever has, actually. It's
1:19:47it's it's great. It's great for us.
1:19:49>> So, the a a concept might like with a
1:19:53with a real niche. Maybe you have
1:19:56something like best mattress. I've I've
1:19:59had the founder of uh Nap Lab, which is
1:20:01a top mattress affiliate and he's been
1:20:03on the show and um they rank I think one
1:20:06or two for best mattress and uh they're
1:20:08they're doing the strategy actually. Uh,
1:20:10and so, um, yeah, so they have this like
1:20:13megalong best mattress article. Um, and
1:20:16then each, I think it's each H2 links to
1:20:20a subpage, which is a lot longer tail,
1:20:23best cooling mattress, best mattress for
1:20:27X, best mattress for Y. These link to
1:20:30each other. They also link back to the
1:20:32best mattress page. And the long tails,
1:20:34because they're longer tail, they're
1:20:35able to rank. They're targeting real
1:20:37keywords, so they get clicked. They get
1:20:39real traffic. They have positive
1:20:43behavioral signals because they're
1:20:45they're
1:20:46>> they're often more buyer intent, too,
1:20:47because they can get very specific
1:20:48towards a very specific buyer.
1:20:50>> Yeah. Yep. And and then and then they're
1:20:53passing their page rank back to their
1:20:55more com to the more competitive best
1:20:58mattress page.
1:20:59>> Yeah. So like while you're waiting for
1:21:01that big one to rank, you're often
1:21:02getting you're the right traffic is
1:21:04already trickling in and um it uh it
1:21:07allows it kind of like builds like a
1:21:09moat of safety around your site too. Uh
1:21:12you become less volatile. You become
1:21:14less susceptible to Google updates cuz
1:21:16also when you're doing this there's
1:21:18nothing for Google to take away. You're
1:21:19not doing anything outside of the
1:21:20guidelines. You're not doing anything
1:21:22that is can be affected by an update.
1:21:24You're simply putting content on your
1:21:25site and you're interlinking it
1:21:27properly. And uh so you'll see a lot
1:21:30less variance when like some sites take
1:21:31big hits. You're going to take a small
1:21:33bump in a lot of situations. So there's
1:21:35only just so many good things that come
1:21:37out of it as as a a technique or a
1:21:39strategy to use. And it's not that's not
1:21:41the end of SEO, but it's a great
1:21:43foundation. And I think that's kind of
1:21:44how I start with my SEO is I'm thinking
1:21:46about which are my target pages, okay,
1:21:48which are my supporting pages and how am
1:21:50I interlinking them? And then kind of go
1:21:52from there as that's we're going to
1:21:53cycle through this every month. We're
1:21:54going to do other things. There are
1:21:55other techniques that we're going to do
1:21:56this or that. But that's kind of the
1:21:58foundational part that I know will
1:22:00always work and will carry us through
1:22:02the good times and the bad.
1:22:04>> You think optimizing for um for AI for
1:22:07chat GPT is most mostly traditional SEO
1:22:10with a different measurement system
1:22:12>> for that's all it is. I I I don't think
1:22:14they can do much different differently
1:22:16than than what Google's doing. I think
1:22:18they're a less sophisticated version of
1:22:20of the Google that we have now because
1:22:22they aren't built for search. So they
1:22:24have to approximate I think what they
1:22:26think Google is doing or what they can
1:22:28do with the systems that they have. Um
1:22:31they do a lot with uh backlinks in the
1:22:33same way that Google did a lot with
1:22:34backlinks in like the early or the late
1:22:362000s and you know 2020 2010s. Very
1:22:40similar concept you know in terms of
1:22:42yeah spam really works in terms of links
1:22:44for the LM because they're they're just
1:22:46less sophisticated. Google had to find a
1:22:48way to commotion. Yeah, it really is,
1:22:50you know, and and that very similar
1:22:52things that worked back then, I think,
1:22:53are very effective now.
1:22:54>> And and Google has all the surfaces,
1:22:58they have phones, they have Chrome, they
1:23:00have everything to to tell like to tell
1:23:03us somebody if somebody's trying to
1:23:04spam. LM don't have
1:23:07>> No,
1:23:07>> no, not at all. No, there's there's no
1:23:10LLM browser necessarily where they can
1:23:12get all that data.
1:23:13>> They've tried to do browsers, but no,
1:23:14but people haven't adopted it,
1:23:16>> right? I mean, why would you leave
1:23:17Chrome?
1:23:18>> Yeah. Why would you leave Fire? There's
1:23:20Firefox there. I guess there's Edge.
1:23:23What's left? Brave.
1:23:25>> Yeah, Chrome is still the market leader
1:23:28by a long shot.
1:23:30>> By a stretch.
1:23:31>> Yeah, by a stretch.
1:23:31>> I'm on Chrome right now.
1:23:32>> Yeah, me too.
1:23:33>> I'm pretty I'm pretty sure I'm logged
1:23:35into my Gmail and it's probably
1:23:36recording this whole thing.
1:23:37>> Yeah. Yeah. Um and uh oh and so what do
1:23:42you think about about uh any um any AI
1:23:47company building its own index like lots
1:23:49of people are saying oh open AAI is
1:23:52making its own index. This is going to
1:23:53be a game changer. What's your take
1:23:56there?
1:23:56>> I think they I think they should give
1:23:57AHFS a call and see how that goes
1:24:01[laughter] because AHF was doing theirs
1:24:03um was it yes or Yep. something like
1:24:06that. Yep. I think um they were
1:24:09launching their own thing and uh that
1:24:11did not go very well. So, and I think
1:24:13they spent a lot of money on it, a lot.
1:24:17Uh and they're an SEO company and they
1:24:19know what they're doing. Yeah.
1:24:20>> And they ran into, you know, like those
1:24:22they're smart people and they know what
1:24:23they're doing and they ran into a lot of
1:24:24problems. Now you've got some people
1:24:26that have no idea what they're doing and
1:24:27they're going to try to do that. And and
1:24:29just as an example of how they really
1:24:31have no idea what they're doing, look at
1:24:33how they accidentally like all these AI
1:24:36companies always accidentally index
1:24:39shared conversations
1:24:41like and and this isn't like oh no that
1:24:43was like 3 years ago. This was like that
1:24:45was last year and now you're trying to
1:24:47build your own index. You're going from
1:24:49not understanding
1:24:51like how not to index private shared
1:24:54conversations to within a year building
1:24:56your own index.
1:24:58I wonder why why is that? Is it is it
1:25:02like the um build fast and break stuff
1:25:05attitude?
1:25:06>> Absolutely. And it's a competitive
1:25:07market. Also, they've raised so much
1:25:10money. They're like, "Well, we got to do
1:25:11something with with all of this." Like,
1:25:13>> we've got all this cash, you might as
1:25:15well.
1:25:15>> Yeah. Like, uh we can't just put in the
1:25:17S&P like, "Well, we got to do some
1:25:19something with it." Like, [laughter]
1:25:21>> and might as well.
1:25:23>> But I mean, also, I mean, why is it?
1:25:25It's because it's getting a lot harder
1:25:26for them to scrape Google's results.
1:25:29That's probably the like it's just like
1:25:31we can't
1:25:31>> You saw the article recently about uh
1:25:33Google poisoning SERs, right?
1:25:35>> Dude, I I made a podcast about that. I
1:25:37like a really widely watched podcast
1:25:39about that.
1:25:40>> Yeah, Barry Schwarz.
1:25:42>> He was on this Barry Barry has been on
1:25:44many times now and and he he talked
1:25:46about that. Um, and it's the Yeah, it's
1:25:49it's and but you see it like if you do
1:25:52SEO regularly, man, like dude, there
1:25:54there was a period where I was dropping
1:25:58in my in my third party rank tracker for
1:26:01all of my branded terms.
1:26:04>> Meanwhile, in
1:26:04>> for branded terms,
1:26:05>> the branded terms. Meanwhile, Google
1:26:07search console like everything's normal.
1:26:09Everything's great. In fact, rankings
1:26:11are going up, organic traffic is going
1:26:13up. Then you look at you look at your
1:26:15third party rank tracker and things are
1:26:18just steep declining down. It it it's
1:26:20horrendous. And and like if you just
1:26:22looked at the third party rank tracker,
1:26:24you'd be like, "What the you would you
1:26:26would be sweating. You would be cursing.
1:26:29You'd be swearing." And then you look at
1:26:30Google Search Console, the source of
1:26:32truth, everything's fine. You look at
1:26:33the SERs, everything's fine. So yeah,
1:26:36and I I will say um because I've I've
1:26:38had uh I've also had people from MA on
1:26:41the show and they're like, "Yeah, you
1:26:43know, like we're we're taking steps and
1:26:45we believe that we can combat this." And
1:26:47I and I will say that I'm seeing third
1:26:49party rank trackers improve. I am like
1:26:52now my branded terms are returning. Um
1:26:55other other terms that that I had lost
1:26:57in the rank trackers that I hadn't
1:26:59actually lost are coming back.
1:27:01>> Are you are you seeing that yourself?
1:27:05I haven't seen any too much direct
1:27:07evidence of it. The thing that we have
1:27:10had like within the tool some people
1:27:11have said like these results are weird
1:27:14like you know they they run they I want
1:27:16to rank for whatever keyword and then
1:27:18they they look at the results they pull
1:27:20back and like these results are weird
1:27:21results I don't
1:27:22>> right and then we look at it and what's
1:27:25funny though is like cuz we we were
1:27:28seeing this before like say that article
1:27:30dropped which was a really great article
1:27:32um but we were seeing it in this these
1:27:34kind of situations cuz then we would
1:27:35search ourselves to double check like no
1:27:38the results are correct and then we
1:27:39compare what they're getting and we're
1:27:41getting we realize oh different data
1:27:42centers are bouncing off these different
1:27:45types of SERs like they're giving just
1:27:47bad results. And so what we've done is
1:27:50[clears throat] um we've implemented a
1:27:52few things on on how to check to see
1:27:54like if this is a potential um bad
1:27:57result and then we'll bounce off a
1:27:59different data center is kind of how
1:28:00we're approaching it right now to see
1:28:02like okay these look wrong based off of
1:28:04this criteria that we have and then
1:28:06we'll do an additional check. How can
1:28:08people at
1:28:09>> um
1:28:09>> or is there a tool that you that you
1:28:11made for this?
1:28:13>> This it's just built into what we do. Uh
1:28:15it's not like a specific tool at all.
1:28:17It's just how we can try to get the best
1:28:18SER results for our users. is
1:28:21>> are you manually looking at the the SERs
1:28:23to be like yeah this doesn't look normal
1:28:25and then you and then you have to check
1:28:26or is this like
1:28:27>> we um we identified uh certain
1:28:30indicators and so if it there's like say
1:28:3215 indicators and if if six hit then we
1:28:35automatically check
1:28:36>> that kind of thing.
1:28:38>> Yeah. I mean, it was, dude, it it's it's
1:28:40real. Like, um, so so I've been I've
1:28:43been just saying to everybody like um
1:28:46just vibe code something with Google
1:28:48Search Console API. Take your historical
1:28:50rank track data, merge that with the 16
1:28:54months that you have from Google Search
1:28:56Console,
1:28:57>> and then from now on, save uh your GSSE
1:28:59data at at the very least for your
1:29:01tracked keywords.
1:29:03Yeah, if if you have that historical
1:29:05data, that's perfect from our
1:29:06perspective because we're always coming
1:29:07at this almost from the first time. We
1:29:08don't have any historical data, so we
1:29:10have to go off of other indicators to
1:29:12say like this SER looks suspicious.
1:29:13>> Yeah. Yeah. Yeah. Well, you could you
1:29:17could still you're you're talking about
1:29:18via third party tool versus GSC,
1:29:22>> correct? Yeah. So, since we don't access
1:29:23have access there's there's
1:29:24>> Oh, you don't have access to GSC with a
1:29:27>> not like not a users uh kind of thing.
1:29:29They just come and say, I want to rank
1:29:30for this particular like user is not a
1:29:33client. So this is different. A user of
1:29:35your
1:29:35>> Yeah. Yeah. A user. Yeah. Yeah. They
1:29:36come in and say, "I want to rank for
1:29:37this keyword." We don't have Yeah.
1:29:38>> historical data
1:29:40>> on that at all. Um so we have to kind of
1:29:44look at it from another perspective.
1:29:45>> Have you thought about allowing them to
1:29:46to connect their GSC's?
1:29:49>> We do it. We have some done for you
1:29:51products and uh in there they can get it
1:29:52because we get better data out of it. Um
1:29:54we're going to open that up to
1:29:55self-service users pretty soon. Um, it's
1:29:59a little tricky because it's not the
1:30:00easiest thing to connect and people
1:30:03aren't always the best at following
1:30:04directions and so we're trying to make
1:30:06it as easy as possible. That's actually
1:30:08been the biggest holdup.
1:30:09>> That makes sense. Um, you said, okay,
1:30:11this is really interesting. So, you've
1:30:13highlighted research showing that 95% of
1:30:16AI generated fanout queries of of query
1:30:19fanouts have zero reported monthly
1:30:22search volume. Um, should SEOs be using
1:30:26query fanouts to inform content or no?
1:30:29And
1:30:29>> well, it's not just that, but it's also
1:30:30like um people also ask in those types
1:30:32of terms and that's actually the best
1:30:34example because it can't have shown in
1:30:37the SERs unless somebody asked,
1:30:40you know, that's the whole the clue is
1:30:41in the name and it's not just person
1:30:44also asked
1:30:45>> people multiple that means it has to
1:30:48happen a few times for it to show up
1:30:50there. That means there is actual search
1:30:51volume and it is actual questions that
1:30:53people are asking. But you throw those
1:30:54into any kind of um you know search
1:30:57volume tool, they're always going to
1:30:58come back 10 or below and you know it
1:31:00has to be more than that. So the the
1:31:03biggest thing that I found in in a lot
1:31:04of like of the research we've done is
1:31:07keywords that are 200 in volume a month
1:31:09or less, it's a guess.
1:31:12The more search volume there is, the
1:31:14more accurate all the tools get for
1:31:16sure. So, if they say if they come back
1:31:18with like, you know, 4,500 and then you
1:31:20actually say ran an AdWords ad, you can
1:31:21actually see the actual number after you
1:31:23run it a little bit. Um, it might come
1:31:25out at like, you know, 44, you know, or
1:31:27like, you know, slightly off. But if
1:31:29like it comes out at like that 10 to
1:31:31zero kind of thing, who knows what it
1:31:33is. And so, that's a gold mine of
1:31:35opportunity because, you know, first of
1:31:36all, you know, people are asking it.
1:31:38These are your users, so you might as
1:31:39well answer the question for them.
1:31:40There's a decent chance more often than
1:31:42not it can blow up to much higher where
1:31:44you're getting hundreds of impressions
1:31:45off of a keyword where everybody's
1:31:46ignoring it because they think it's
1:31:47zero.
1:31:49>> So query fanat should people should
1:31:51people pay attention to them? If so like
1:31:53what should be what how should they be
1:31:55using them to do SEO or everybody hates
1:31:58this term but go like yeah
1:32:02>> SEOs love to label don't we?
1:32:03>> Um because if we get a new label then we
1:32:06can charge more money.
1:32:06>> That's well that's what it is. That's
1:32:08that's why that's why these people are
1:32:10saying that geo is so different from SEO
1:32:12and I don't know if you if you've seen
1:32:13so yeah you're going to be on stage with
1:32:14David soon. So, we have this um
1:32:17recurring series on this show where like
1:32:19so this is a pretty popular SEO show and
1:32:21I get pitched all the time by people
1:32:23with quoteunquote geo agencies and um
1:32:26there's even though this is a pretty
1:32:27popular show, plenty of people who don't
1:32:29know me, plenty of people who don't know
1:32:30this show and they're like, "Yeah, I'd
1:32:32love to come on your show and talk about
1:32:35how geo is a completely different beast
1:32:39than SEO. SEO is so different and
1:32:41honestly, I believe it's outdated. I'd
1:32:43love to come on the show and talk about
1:32:45that. And I say, great. Would you like
1:32:47to have a conversation with somebody who
1:32:49maybe a little bit of a debate with
1:32:50somebody who believes that SEO or maybe
1:32:53it could get a pretty heated debate here
1:32:54previous episodes. They don't watch the
1:32:56previous episodes even though I give it
1:32:57to them with somebody who believes that
1:32:59SEO and G are the same thing. They're
1:33:01like, I'm down. Sounds good. They just
1:33:02want to get on the show because they
1:33:03think they're going to be able to plug
1:33:05their their geo agency. And I give them
1:33:08I say, "These are the previous episodes.
1:33:10You can go watch them. It's going to be
1:33:12a debate and there they don't watch then
1:33:14they come on the show and so David David
1:33:17Quaid will be the person that they have
1:33:18to talk to. I'll be the moderator and I
1:33:20will basically just ask one question and
1:33:22then for the rest of the episode shut up
1:33:24and I I I will for the rest of the
1:33:26episode I just shut my mouth and I let
1:33:27them go at it. And so we've done all of
1:33:30these we've done all of these geo versus
1:33:33SEO debates where David annihilates
1:33:35these people like like the last one was
1:33:39the last one was insane like really
1:33:42really there were some the comments for
1:33:44that episode were crazy. Um it was like
1:33:48just yeah like um and that person who
1:33:50came on the show he even said he
1:33:53literally said I am trying to make geo
1:33:56seem like something different from SEO
1:33:58so that I can make money off of the way
1:34:01that
1:34:02>> ended up admitting it at the end so
1:34:03defeated.
1:34:04>> Yes. Yes. and and like and you know
1:34:07people might think that that oh yeah
1:34:09it's pretty harmless but it hurts our
1:34:11colleagues because because now like um
1:34:15marketing department CMOs uh employers
1:34:19they think that they need to hire
1:34:20somebody different and the people who
1:34:22are honest about no these are really
1:34:24just the same things you're doing a lot
1:34:26of the same tactics the honest people
1:34:29are the ones suffering
1:34:30>> and uh
1:34:33>> that's a good point I would have brush
1:34:34it off as harmless less uh normal
1:34:36industry stuff. But you make a good
1:34:38point there.
1:34:39>> If if if they buy in and say that, then
1:34:41that's taking budget away from people
1:34:42that are actually being honest and
1:34:43transpar.
1:34:44>> Well, and then you see and then you see
1:34:46other companies falling for these spam
1:34:48tactics that these so and that these
1:34:50quote unquote geo companies are are
1:34:52recommending. They're like, "Okay, Gio
1:34:54is a completely different beast. We need
1:34:56to spam hundreds of AI generated
1:35:00self-promotion listicles." And like and
1:35:02so they're like falling for these
1:35:04tactics that are going to hurt them. And
1:35:06yeah, you might make the argument. Well,
1:35:07then they, you know, then they get hurt
1:35:09and then they go to to people like us
1:35:11who are like, "Please help us. We got to
1:35:13we messed up. We listened to the wrong
1:35:14people." But like um [laughter]
1:35:18so just real quick getting back to we
1:35:21got off on a
1:35:22>> Yeah, sorry about that.
1:35:23>> A good tangent. A good tangent. But um
1:35:26I'm not worried about query phantom and
1:35:28I know that it was a much more popular
1:35:30term last year or maybe earlier this
1:35:32year than it is now in terms of what
1:35:33people are saying and and maybe some
1:35:35correlational studies that have come
1:35:36out. I'm not worried about that at all.
1:35:38I'm worried about answering the
1:35:39questions that my users have. And
1:35:41there's a huge uh overlap in the vin
1:35:44diagram of people also ask questions
1:35:47that people are asking and what query
1:35:48found out is. And so I'm not concerned
1:35:50at all because it's just something
1:35:51should be on your site anyway. If the
1:35:53term is really highly connected to your
1:35:56your target term, then that's probably
1:35:57going on the target page, right? You
1:35:59know, that's going to be an H2. That's
1:36:00going to be something on there because
1:36:01it makes sense. If it's slightly
1:36:03adjacent, that's perfect for supporting
1:36:05content. And that's what I want to get
1:36:07on my site because I want it to be a
1:36:08place that Google understands and LLM
1:36:10search understands and users understand,
1:36:12they can come and find an answer. So if
1:36:14it's query fan out, if it's people also
1:36:16ask, if it's supporting content, in my
1:36:18mind, that's all the same thing. And um
1:36:20it's great to do. And uh I if it if it
1:36:24benefits or not one thing, it's
1:36:25definitely going to benefit something
1:36:26else. And then that's all I'm really
1:36:28concerned about.
1:36:28>> Are you finding people also ask
1:36:30questions?
1:36:32uh we can scrape them pretty easy. But
1:36:34um the really great way to to this day
1:36:36people really don't know that if you hit
1:36:38that bottom arrow you get three more. If
1:36:40you hit the bottom arrow you get three
1:36:42more. And if you hit the bottom arrow
1:36:43you get three more. You can get three
1:36:45months worth of content you know in
1:36:47about two minutes. And um then pick the
1:36:50ones you like and go with it and you've
1:36:52got you've got a pretty decent content
1:36:53schedule. If you don't you know if
1:36:54you're bowling on a budget and you know
1:36:57you need to find ideas that's how I
1:36:59would do it for sure.
1:36:59>> Yeah. Mark Williams Cook. He's been on
1:37:01the show before and he has a tool also
1:37:03ask.com which is I don't know if you've
1:37:05tried it. Um
1:37:06>> and you you know Mark though
1:37:08>> and uh Yeah. Yeah. And so he is also
1:37:10ask.com which is really cool and it's uh
1:37:12it makes it really easy to get uh paa
1:37:15questions.
1:37:17>> Yeah, it's perfect. I mean yeah there
1:37:19are a lot of tools that do it. You can
1:37:20do it yourself that it's a great way to
1:37:22go. But um it's it's just a great piece
1:37:24to have is in terms of like what are the
1:37:26foundational things you're doing? Why
1:37:28aren't you answering questions that
1:37:29people might have?
1:37:30>> Yeah.
1:37:31>> Yeah. Yeah. Yeah.
1:37:32>> Like what
1:37:33>> what's the point of your website then?
1:37:34You know, if you're not being helpful,
1:37:36if you're not being useful, if you're
1:37:37not a place where people can come and
1:37:39find out if they should or should not
1:37:40buy this or they should or should not do
1:37:42this or whatever it might be, those
1:37:43questions should be on the site
1:37:44somewhere. You said in a in a May
1:37:47interview that AI search adoption
1:37:49remained relatively small and that AI
1:37:51visibility could be a vanity metric for
1:37:54businesses. Do you still you still
1:37:56believe that?
1:37:58>> Yeah. Um I've got the the the dad and
1:38:02Janice metric. It's an index that I
1:38:04created.
1:38:04>> Okay.
1:38:05>> Uh my my dad and his wife Janice, I ask
1:38:07them every other month, are you doing uh
1:38:09any AI search, guys? And they're like,
1:38:11"What's that?" Like,
1:38:13>> yep. Nope. Yep. Yep. That's so true.
1:38:16That's so true.
1:38:17>> They're re they're recent retirees.
1:38:19They're online every day. They take two
1:38:22or three trips a year. They buy gifts
1:38:24for themselves and their grandkids and
1:38:25everything and they're and they make
1:38:26online reservations. They're they do
1:38:28driving directions. They're constantly
1:38:30online. They have never heard of
1:38:31ChatGpt.
1:38:34Actually, that was the best. It was
1:38:35about a year ago. I took a screenshot of
1:38:38a text that I of our text thread. I'm
1:38:40like, "Hey, are you guys using chat GPT
1:38:42or any other AI search?" And Janice was
1:38:44like, "We've never even heard of that.
1:38:46Should we be doing that?" [laughter] you
1:38:47know, and the consumer, a lot of the
1:38:50consumers that have a lot of money just
1:38:52aren't there. You know, they just aren't
1:38:53there. They won't be there. It's a
1:38:55different type of consumer. Kids are
1:38:57maybe using a little bit more, but not
1:38:58that much. I have a son that's 12 and um
1:39:01uses it a bit, but not as much as you
1:39:04would think, you know, and kids have no
1:39:06money. So,
1:39:07>> so you you know, basically 50 plus,
1:39:10maybe 45 plus and under 18, those
1:39:13segments are gone. So now you're you're
1:39:15focusing on this and then I don't know
1:39:17just ask ask you your your cousin Susie
1:39:20you know you you know uncle Bob you know
1:39:23different people in their different
1:39:24places in their lives and ask them how
1:39:25much they're using chat GBT or Claude or
1:39:27anything like that and I I'm going to
1:39:28guess the answer is pretty low.
1:39:30>> There was uh some controversy in the SEO
1:39:33world over the last week and a half
1:39:35maybe you saw it. Um and Marie Marie
1:39:38Haynes wrote this article is geode a
1:39:41distraction I think that was what it was
1:39:43called. And then it was very
1:39:45interesting. She showed all this data
1:39:46suggesting that basically like you're
1:39:48saying, it's it's vanity metrics. Um,
1:39:51and Ran Fishkin liked this article so
1:39:53much that he made a he made a video
1:39:55about it. Um, and he and he shared that
1:39:57video across platforms.
1:39:58>> Um, and that video blew up.
1:40:00>> And so Mike King, who's friends with
1:40:02them both, um, Mike King said, you know,
1:40:05like, um, I love both of you. Here's my
1:40:07take. And Gio was not a distraction. Um
1:40:11and uh yeah and so uh it was just I made
1:40:14an episode about it a few days ago. So
1:40:16it's very interesting to see uh the
1:40:18conversation between between both
1:40:20people.
1:40:22>> Yeah. I mean they're all smart people
1:40:23too.
1:40:23>> Yeah.
1:40:23>> And they're all they're all successful
1:40:25people and you know they're they're all
1:40:27fun to listen to. Um
1:40:30>> I guess the controversy is what makes it
1:40:33what makes it all interesting, right?
1:40:34>> That's true. That's that's that's that's
1:40:36what that's what it is. Yeah. Yeah, but
1:40:38I guess we're all likeminded. It
1:40:40wouldn't work.
1:40:40>> My my point with um and my point with my
1:40:42video was like, you know, like
1:40:45I don't know who's right, but like we
1:40:47know the tactics that work at the very
1:40:49least. And uh yeah, I
1:40:52>> I just think, you know, next week they
1:40:53come out with this brand new channel
1:40:55called Meet Morp. It's going to be it's
1:40:58going to be the same. It's going to be
1:40:59relatively the same,
1:41:01>> you know, and and people are going to
1:41:03freak out. Yeah. What I really enjoy
1:41:05actually is, you know, a few years ago
1:41:06people were really complaining about
1:41:08Google having, you know, 104% market
1:41:10share and they're like, we just need a
1:41:12couple more channels to come in here to
1:41:14level the game.
1:41:15>> So true.
1:41:16>> And then a couple channels come in and
1:41:18people are
1:41:20like, go back to Google,
1:41:22>> kill, they hate it. Yeah.
1:41:24>> Yeah. Exactly.
1:41:25>> So things change a little bit, channels
1:41:27change a little bit. It all is
1:41:30relatively the same is really, I think,
1:41:31the bottom line, but it's easy to freak
1:41:33out. the sky is falling. I don't know
1:41:35what I'm doing, you know. Or like, you
1:41:37know, they had a technique that worked,
1:41:38but maybe it's not working now, but they
1:41:40really kind of missed the why part of
1:41:41it. Um, that's kind of maybe a bigger
1:41:43issue within the industry. But, uh,
1:41:46ultimately, if you've got a framework
1:41:47that's repeatable, you can grind on it.
1:41:50Grinding is the biggest thing. Worrying
1:41:51about SEO and doing SEO are two very
1:41:54different things, and people often
1:41:56confuse them, you know, and they they
1:41:58they they stress about their SEO a lot
1:42:00more than they do their SEO. What do you
1:42:02mean grinding is the biggest thing?
1:42:05>> Doing a mediocre plan to completion will
1:42:09beat the best plan of all time that it's
1:42:11only done, you know, 10%.
1:42:12>> Yo,
1:42:14that's that's one of the reasons a lot
1:42:15of people listen to me. I show up every
1:42:17day. I grind.
1:42:19>> I get the work done.
1:42:20>> I get it done.
1:42:21>> Even when it's like 3:00 a.m. and I want
1:42:22to sleep and my first export failed and
1:42:25it's a 2hour podcast that I have to
1:42:27export and it's going to take another
1:42:28two hours to do it. I'mma do it. It's a
1:42:31lot of fun, right?
1:42:31>> I show up. I grind. I show up every
1:42:33Well, it actually is like I'm really
1:42:34enjoying like
1:42:35>> actually it is. I do enjoy.
1:42:36>> Yo, I'm like I can't believe I get to do
1:42:38do these calls. Like I'm having such a
1:42:41good time talking to you. David Quaid
1:42:42and I we had this like great 90minute.
1:42:44We were vibing yesterday. Like this is
1:42:46amazing. Like what a great life.
1:42:48>> No, it's it's fantastic. Okay, I have a
1:42:49few more questions. Um not not many not
1:42:52many left. Uh you you were you were
1:42:54saying something um interesting. Okay.
1:42:56So, yeah. What's what is uh what's
1:42:58better for SEO or for or for geo? Um is
1:43:03it focused pages um or enormous
1:43:07comprehensive guides? So, like should
1:43:10super long pages
1:43:12uh such as articles like long articles
1:43:14be split into smaller pages?
1:43:17>> Yeah, the skyscraper thing it worked for
1:43:20a hot second. Um but the bigger issue is
1:43:23uh is I think to combat that Google uh
1:43:26really focused on what they call
1:43:28centerpiece annotation and that's the
1:43:30concept that they figure out what the
1:43:32main con the main idea of the page is
1:43:35and then from there they evaluate does
1:43:37everything stay within what we think
1:43:39should be for that and then as soon as
1:43:41you have content that gets away from
1:43:42that it devalues that content and then
1:43:45by extension devalues the page now since
1:43:48that first came out I think they've
1:43:50taken the concept of centerpiece
1:43:51sanitation and spread it across the
1:43:52site. And the idea is that if if pages
1:43:55on the site get too far away from what
1:43:57the main concept of the site is, those
1:43:59pages get devalued and then by extension
1:44:02the site gets devalued. So the problem
1:44:05is it's not necessarily having a page
1:44:07that's 10,000 words. The problem is not
1:44:09all those 10,000 words might be exactly
1:44:11onto the main topic and then you start
1:44:13hurting yourself. And so for that
1:44:15reason, you know, it's it's a much
1:44:17better plan to figure out, okay, we're
1:44:19going to do a couple thousand words and
1:44:20then we're going to break these things
1:44:21up and into other pages so that each
1:44:23page is staying within that concept of
1:44:25centerpiece annotation.
1:44:28>> Did you um you saw what happened to
1:44:30HubSpot's SEO?
1:44:33>> Not recently. No.
1:44:35>> Oh, how it um how it just completely
1:44:37collapsed. Oh, it was like
1:44:39>> Oh, I did see that. Yeah.
1:44:42>> Yep. But um man, weren't they like
1:44:44selling guest post though or something?
1:44:45Well, they do. They had just all that
1:44:47extra content.
1:44:48>> It was they had like this like
1:44:49motivational quote subfolder and like
1:44:52basically like a lot of the content that
1:44:54collapsed was completely unrelated to
1:44:56their core business.
1:44:58>> Yeah.
1:44:58>> And it was it was top of funnel. Great
1:45:00example. Yeah. Yep. Yeah. And so a lot
1:45:02of people they they said um the reason
1:45:04that their SEO collapsed for it was for
1:45:07terms that were unrelated. Um and they
1:45:10just didn't have the the topical
1:45:12authority. They were they're they were
1:45:13spread like they were spread too far out
1:45:16where it's like what does this even
1:45:17mean? What does this website even mean?
1:45:19>> They had spread themselves too thin.
1:45:21>> Yeah. I think that goes along very well
1:45:23with center centerpiece annotation.
1:45:26you've got a a website that's all about,
1:45:28you know, being a a CRM, you know, and a
1:45:32content publisher and uh motivational
1:45:34quotes is probably not, especially if
1:45:36it's a huge category where they're just
1:45:38cranking out. Was that just like them
1:45:40automating something to get content out
1:45:42there to to kind of
1:45:43>> this was in um so I think this was like
1:45:452024 2023 and they're like, "Yo, we're
1:45:48HubSpot. We can do anything that we
1:45:49want." Like uh yeah, we we need more.
1:45:51And so a actually during um at at uh
1:45:55what's the the big HubSpot conference? I
1:45:57forget. Oh in inbound HubSpot inbound
1:45:59something like that.
1:46:00>> That's I think that's right.
1:46:01>> Yeah. And so so they're like um after
1:46:02this happened they're like yeah we're
1:46:04going to focus more on the bottom of the
1:46:05funnel. We were too focused on the top
1:46:06of the funnel. And and so
1:46:08>> is that is that top of funnel though?
1:46:10>> Motivational quotes.
1:46:12>> Yeah.
1:46:12>> I mean is that even is that even is even
1:46:14that part?
1:46:15>> Oh yeah. Like it's even part of the
1:46:16funnel. Like dude it's as top as you can
1:46:18get. Like it's so unrelated. Yeah. It's
1:46:20so far top. It's on your website.
1:46:22>> It's just pure um
1:46:26name awareness. Not like even for what
1:46:28they do, just like, oh yeah, I I I don't
1:46:31know. This name seems familiar. Maybe I
1:46:33trust it a little bit more.
1:46:34>> And and yeah, and and so they went hard
1:46:37with like all of this stuff because
1:46:38they're like, we're HubSpot. We can rank
1:46:40for anything, so let's rank for
1:46:41anything.
1:46:41>> Yeah.
1:46:42>> And uh Yeah. And it and that really hurt
1:46:44them. And then there was this major
1:46:46scandal. I don't know if you remember
1:46:48where uh one of the founders of HubSpot
1:46:52went on a big podcast and said SEO
1:46:55doesn't work. It's all about AI search
1:46:57and we're the perfect example of this.
1:46:59We lost all of our clicks because AI
1:47:02stole them. And then SEOs came in and
1:47:05were like, "No, actually we can see that
1:47:08you lost all of your rankings and that's
1:47:10why you lost all of your clicks."
1:47:12>> Right? It'd be one thing if the rankings
1:47:15were holding tight.
1:47:15>> Yeah.
1:47:16>> Yeah. Which they were not. The rankings
1:47:18had completely collapsed.
1:47:19>> Yeah.
1:47:20>> And but the the this founder went on a
1:47:22major business podcast and said SEO is
1:47:25dead.
1:47:26>> And it was and and actually they they
1:47:28had to they had to change the name of
1:47:30the podcast because the name of the
1:47:32podcast episode said SEO is dead. Like
1:47:36HubSpot founder proves it. And uh yeah,
1:47:39[laughter]
1:47:39>> that's great. I don't remember that part
1:47:41of it. Yeah, I made I made I that it
1:47:44went viral on the SEO subreddit and I
1:47:45made a few podcasts about it because I'm
1:47:47just like, "No, it's not.
1:47:48>> No, no, it's not."
1:47:49>> That's hilarious.
1:47:50>> It's propaganda.
1:47:50>> You got to figure that that's a
1:47:52shareholders meeting situation, right?
1:47:54>> Yeah. That's what Oh, no, no, no. It
1:47:55wasn't our our bad strategy. It was AI
1:47:58like uh Yeah. Yeah.
1:47:59>> What are you going to do?
1:48:00>> Yeah.
1:48:01>> We We did the right thing.
1:48:02>> What do you What do you want us to do?
1:48:03Um All right. Two more two more
1:48:05questions. If somebody gave you a brand
1:48:07new domain and $1,000 and you had 90
1:48:11days to generate organic traffic, what
1:48:13would you do? What would Kyle Roof do?
1:48:16>> Kyle Roof do I don't know. 90 days, a
1:48:20thousand bucks.
1:48:21>> You're going to you're going to get an
1:48:22expired domain.
1:48:24>> Okay.
1:48:25>> Right.
1:48:25>> We're going to get a cheap one that
1:48:26that's decent. Right.
1:48:27>> I mean, I I was I would assume that
1:48:29that's what you're going to do.
1:48:31You know what you can do is um actually
1:48:32what's really nice with AI now is you
1:48:34can grab what it was from the from the
1:48:37way back machine or something similar.
1:48:39Uh redo all the pages, but you can
1:48:41optimize them and do them better. Put in
1:48:43your own images and probably get that
1:48:45traffic back probably within that 90
1:48:47days. And that's going to be your
1:48:48fastest uh route to dollar one out of
1:48:51that $1,000. And you can probably do it
1:48:52with that $1,000. You can do with less.
1:48:54You can probably do with a couple
1:48:55hundred bucks to do that.
1:48:56>> Would you change the brand name? or
1:48:58their copyright things to think about.
1:49:00>> Yeah. Yeah. All that stuff I would
1:49:02change and you got to change out their
1:49:03socials and you can't use any of the
1:49:04images that were on there before. But
1:49:06you can recreate all the URLs.
1:49:07>> Can you keep the text the same?
1:49:10>> No, I wouldn't. I I don't think there's
1:49:13any legal issue with I don't this is not
1:49:15legal advice. I don't think there's any
1:49:17issue with the text necessarily unless
1:49:19it's some proprietary sort of thing, but
1:49:22text seems to be text. Um, but you can
1:49:25change that easily cuz you you're going
1:49:27to want to reoptimize it anyway because
1:49:28there's a decent chance they weren't
1:49:29really doing as best a job as they
1:49:31could. So, take that opportunity to use
1:49:33the concept and just reoptimize those.
1:49:36>> No, you said but you also said change
1:49:37the brand name, but if you're buying a
1:49:39domain like the usually the brand name
1:49:41is pretty tied to the domain name.
1:49:43>> Yeah, but you can, you know, if it was,
1:49:45you know, Bill's bait shop, you know,
1:49:47now we're doing bait shop express, you
1:49:49know, kind of a thing. And even though
1:49:51the domain might be billsbaitshop.com
1:49:54um but then from there, so that's gonna
1:49:56be that's only a few hundred dollars. So
1:49:57let's say that's going to cost me 350. I
1:50:00got 650 in the bank, right? Um I'm going
1:50:04to do a pretty big content blitz. Then
1:50:06uh I'm going to determine how how strong
1:50:08I think the site is. And you do that by
1:50:10kind of the number of clicks that it's
1:50:12getting um per day on average. And then
1:50:15that's going to inform you that you can
1:50:17go after a keyword that has that monthly
1:50:19search volume.
1:50:21And then I'm going to blitz content that
1:50:24has that search volume to try to get
1:50:26some very very quick wins out of that.
1:50:30Um, so that's going to cost me uh
1:50:32probably 300 for probably about the same
1:50:34amount. So 350 350. So now I'm sitting
1:50:38about 300 bones left. What am I going to
1:50:41do with that?
1:50:44Uh, do I need do I care about the
1:50:47longevity of this thing or do you doing
1:50:50like this is like a Black Friday special
1:50:51one and done? Here we go.
1:50:55>> Uh, what would you Okay, let's let's do
1:50:57it for both.
1:50:59>> If I didn't care, I'm probably going to
1:51:01start doing a CTR manipulation campaign
1:51:05because you can get on page one pretty
1:51:06quickly.
1:51:07>> Would that would this be connected to
1:51:08your GSC or would you would you isolate
1:51:11it? Yeah, all your sites should always
1:51:13be isolated anyway to be as safe as
1:51:16possible, especially because, you know,
1:51:17if you've got client sites, for example,
1:51:19and your clients are doing something
1:51:20dirty, you don't want that coming back
1:51:21on your GSC. Um,
1:51:25but yeah, I would keep it separate just
1:51:26as a matter of course. Um, but yeah, so
1:51:30if if like he didn't care, he could burn
1:51:32it and you know, we just need to rank
1:51:35and this is that, then the CTR
1:51:36manipulation is going to happen.
1:51:38300 bucks that
1:51:40>> I I I don't have anything special to
1:51:42tell you, but um I'd say they're
1:51:43probably all about the same.
1:51:46Um 300 bucks is probably not going to
1:51:47get you too far, though. So, time it. Uh
1:51:51CTR manipulation will work for about a
1:51:53week.
1:51:54So, I would time it for about a few days
1:51:57before you need it and then within that
1:51:59window of time. If I cared a little bit
1:52:02more about the
1:52:02>> Yeah, let's say you care. This is you're
1:52:04you're you're actually trying to start a
1:52:05legitimate business.
1:52:07Yeah,
1:52:09I've got my content going. I've got the
1:52:11pages rebuilt.
1:52:14Uh, that's going to go that's going to
1:52:16last you pretty that's going to do
1:52:17pretty well.
1:52:18>> You got 300 bucks. Go to go get some
1:52:20cake.
1:52:20>> Got 300 bucks.
1:52:21>> Go to go get some cake. Get some coffee.
1:52:24Go to the amusement park.
1:52:27>> See, the thing is like if you would try
1:52:28to get like a backlink campaign going,
1:52:29that's a backlink. It's worthwhile,
1:52:31right? Um,
1:52:34you know what I would do actually is I
1:52:38would put that $300 into Google Ads and
1:52:41I would run ads on the keywords to my
1:52:43pages. It's going to help those pages
1:52:45index. It's going to get referral
1:52:46traffic in and I'm going to get real
1:52:47search volume numbers because you can
1:52:50get that out of AdWords. So that might
1:52:52inform and that's also maybe going to
1:52:54get me to dollar one and get recoup some
1:52:56of my money. So I'm going to get
1:52:58referral traffic, indexing, good data. I
1:53:01can maybe split test some things too
1:53:02with an ad copy to maybe work on my H1s
1:53:04to increase my CTR. So, I'll put a $300
1:53:07campaign into that and try to get some
1:53:09data out of there.
1:53:09>> For people who ask, cuz I literally I
1:53:11was just asked this yesterday. Is there
1:53:13a ranking correlation? Like, do you do
1:53:15you need to buy ads to have your organic
1:53:17traffic do better?
1:53:19>> No, but pages will rank faster or index
1:53:21faster
1:53:23for sure. Um I I've not I've not seen
1:53:26like a I'm running an ad, my page will
1:53:28improve, but um I have seen that pages
1:53:30that run ads do better. [laughter]
1:53:32So, but um you definitely the data you
1:53:35get out of it,
1:53:35>> but it's just it's just the indexing
1:53:37benefit and the data that you get out of
1:53:39it.
1:53:40>> The indexing for sure, the data for
1:53:42sure, a ranking benefit, eh, but um
1:53:46pages that start to get referral
1:53:47traffic, whether they're from ads or
1:53:48they're from, you know, social or
1:53:50whatever, tend to start to rank better.
1:53:52So whether you can say there there's a
1:53:54factor there or not or there's other
1:53:55things going on, but it is definitely
1:53:57part of of the mix. And so that's what
1:53:59I'm going to do. I'll take my $300 and
1:54:01I'll I'll do it that way.
1:54:02>> Well, you also I mean you show you've
1:54:04you we were talking about this that that
1:54:06no follow links matter if the page ranks
1:54:09and it's it's getting clicks. It's
1:54:10getting traffic. Yeah, it's getting it's
1:54:11getting traffic. And so Joy Hawkins, she
1:54:14she's like, I want to test this. And she
1:54:16did multiple tests and found all the
1:54:17same things. And so she put she put a a
1:54:22no follow link on her forum on her on
1:54:26her local SEO forum and it went to a
1:54:29page on her site. Um not on the forum
1:54:31but like actually on uh on Sterling Sky
1:54:35and
1:54:37uh that page also had a YouTube video
1:54:40embedded and this no follow link from
1:54:43this forum site that was ranking and
1:54:46getting clicks. This no follow link was
1:54:48so powerful that not only did the page
1:54:51that was linked to rank, her embedded
1:54:54YouTube videos started ranking as well.
1:54:57>> That great.
1:54:57>> Yep.
1:54:58>> That's so great. [laughter]
1:55:00>> That was a lot of fun. That was a fun
1:55:02test.
1:55:02>> Yeah. And she tested it multiple times.
1:55:04I I can't believe I didn't ask you this
1:55:05um yet and I'm so glad I remembered.
1:55:07What are your favorite ways to build
1:55:08links?
1:55:11Uh,
1:55:14you know, my all-time favorite is um SEO
1:55:18for good is is my my favorite approach.
1:55:21And there are a lot of different ways to
1:55:22do it, but one thing you can do is do
1:55:24like um a canned food drive for your
1:55:28local food pantry, right? Um, and you
1:55:31contact local businesses and you contact
1:55:33the Chamber of Commerce and you contact
1:55:36different charities and you have a page
1:55:37that says, "Hey, Tuesday, you know,
1:55:39November, whatever. Um, this is what
1:55:43we're doing." And then you send that out
1:55:45and what happens is people start to
1:55:46share it. They put it on their
1:55:48calendars. You're getting all these back
1:55:49links coming in. And at the end of the
1:55:52day, yeah, you got backlinks from a lot
1:55:53of people. You probably wouldn't have
1:55:54gotten from a lot of.orgs that are
1:55:56powerful. You know, maybe some.gov,
1:55:58some.ed. edus are all coming in now. You
1:56:01got them fairly and then uh you did some
1:56:04good for the community on top of it and
1:56:07it cost you time is what that one's
1:56:09going to cost you more so than going out
1:56:11and you're going to get a lot more links
1:56:12than you ever would have gotten on any
1:56:14kind of uh paid link campaign.
1:56:17I love that. I made an episode about a
1:56:19local roofing company that donated a
1:56:22roof to like hurricane v victims and
1:56:25they went super viral. They were in all
1:56:28the local publications
1:56:30um in lots of other publications. They
1:56:32were shared outside of the local area um
1:56:37in on social everywhere and like they
1:56:40got better links than any link building
1:56:42budget could buy
1:56:44>> for sure. And think about too with like
1:56:46short form content now as well cuz you
1:56:48want to maximize what you're doing
1:56:50across as many channels as you can. And
1:56:52so the buildup to the day, the cans
1:56:55coming in, you know, the kids with their
1:56:57families bringing stuff to help their
1:56:59neighbors, taking it to the wherever,
1:57:01the drop off, the hugs, the high fives,
1:57:03that turns into really great content
1:57:05that people would love to see. People
1:57:06need joy in their lives.
1:57:07>> They do.
1:57:08>> I think a lot more than they need
1:57:09bitterness. And there's a lot of
1:57:11bitterness out there.
1:57:12>> Too much.
1:57:12>> Do a little good. Do a little good.
1:57:14Bring a little joy. There's nothing
1:57:15wrong with it benefiting your business
1:57:17at the same time. you know, you have the
1:57:19resources to make it happen, you know,
1:57:21because you've got the business, but
1:57:22this allows you to give back. And at the
1:57:24same time, it's really great. It's great
1:57:25for your company.
1:57:26>> And how did you how did you say you're
1:57:28sharing these goodwills? Because a good
1:57:30like you you still need people to hear
1:57:31about it. So, how did you say you're
1:57:32you're sharing them?
1:57:34>> You could just reach out to like like
1:57:36the the different organizations, the
1:57:37different charity event uh organizations
1:57:39that exist within your area or
1:57:40nationally or or where whatever the the
1:57:42operation is that that you're doing. um
1:57:46uh the Chamber of Commerce, that kind of
1:57:47stuff that you can reach out if you try
1:57:49to if you can get on one of those
1:57:50calendars. A lot of people aggregate off
1:57:52of those calendars and so they go into
1:57:54other organizations calendars as well.
1:57:56Um and just a little outreach to say,
1:57:58"Hey, this is what we're doing. If you
1:57:59could post this on your Facebook, maybe
1:58:01if you could do a blog post about it, if
1:58:02you can share it here or there, that
1:58:04kind of stuff." And people are usually
1:58:05very uh willing to do that because it
1:58:07doesn't take a lot of time for them. And
1:58:08you know, again, you're doing good for
1:58:10your community.
1:58:10>> Yeah. Yeah. Yeah. All right. Last
1:58:12question. as this is it is 11 p.m. for
1:58:15you. So, thank you for for doing this so
1:58:17late. Yeah. Yeah. And
1:58:19>> that that went really fast. That was a
1:58:20lot of fun though. That was
1:58:21>> Well, it's not this my last I have one
1:58:23more question. I have one more one more
1:58:25question as we
1:58:25>> Is this where it all goes downhill? This
1:58:27isn't fun anymore.
1:58:28>> No, this is No, this is super easy
1:58:29question. It's a It's a fun interesting
1:58:31one. They're They're all fun. I try to
1:58:33make them all fun and interesting. Um,
1:58:34but yeah. So, okay, as we end on as we
1:58:38as we round on this twohour epic pod,
1:58:42what is your boldest prediction about
1:58:44SEO over the next 3 years that most
1:58:46people in the industry would disagree
1:58:48with?
1:58:50>> Um, LLMs we as we have them right now
1:58:53are going to die.
1:58:56They will become irrelevant because
1:58:57there will be a new system
1:59:00that will do things better. And that's
1:59:02coming within the next three years. And
1:59:04everybody that put all their money and
1:59:06time and effort into these LLMs might be
1:59:09screwed because again, they didn't learn
1:59:11the framework. They didn't learn the
1:59:12why. They just focus in on a technique
1:59:14that works for these. But another
1:59:16thing's coming soon and it's going to
1:59:17wipe these guys out. Think about going
1:59:20from like records to eight track to
1:59:23cassettes to CDs, you know, to digital
1:59:27to whatever is coming next to cloud,
1:59:29that kind of stuff. It's this is we're
1:59:31going to see the same kind of thing with
1:59:32this. and um you're going to have to
1:59:34reby your record collection, you know,
1:59:36in and into a new medium is what I think
1:59:40that's my wild and bold prediction over
1:59:42the next 3 years.
1:59:43>> Why why why do you believe this?
1:59:45>> Just how technology works. You know, you
1:59:48go back in time, you can just see
1:59:49exactly how that works. Every time a new
1:59:51technology is coming out, you get rapid
1:59:54uh development and invention. I remember
1:59:56the point when people said uh when I was
1:59:59way back in the dark ages uh CDs could
2:00:02only hold so much information and I was
2:00:04watching this news report that was
2:00:05saying the issue is they can't spin them
2:00:06fast enough and so they're like oh we
2:00:08need to find a way to spin them faster.
2:00:11No they didn't. They just found a
2:00:12different medium that could hold more
2:00:13data.
2:00:15You know then we got DVDs you know then
2:00:17we got something else and something
2:00:18else. And so I think if you follow the
2:00:20the history of technology and how the
2:00:22things go, what we have now is these
2:00:25LLMs are going to get wiped out by a
2:00:27completely different technology, it's
2:00:28going to do something very similar, but
2:00:30it's going to expand upon what's there.
2:00:31It's going to do it better, more
2:00:32efficiently, take less energy, uh,
2:00:35organize things better, be more
2:00:37compatible to what we're trying to do
2:00:38rather than a large language model is
2:00:40trying to predict words that are coming
2:00:42and then we're framing stuff around
2:00:44that. I think we're just going to get a
2:00:45new medium. I think that's going to come
2:00:46within the next 3 years. How how can
2:00:48people prepare for that?
2:00:51>> Understand that like the the search
2:00:52itself isn't going to change, you know,
2:00:54like how information retrieval doesn't
2:00:56change. And so understanding how that
2:00:57works cuz again it's the same principles
2:00:59as we're talking. The principles are
2:01:00going to be the same. It's just there's
2:01:02going to be a new uh uh AI, if you will,
2:01:06that that'll be running the show.
2:01:09>> I didn't expect such a great answer from
2:01:11you on that last one. I thought you were
2:01:12going to be like, "My prediction is that
2:01:14SEO is not going to change over the next
2:01:163 years." And you gave something way I'm
2:01:19literally after this I'm going to clip
2:01:20that into into a short form on on social
2:01:23because that was that was fire by that
2:01:25was unexpected like and my face like
2:01:27when you watch it back you you like you
2:01:29say that and I go
2:01:32that was [laughter]
2:01:33>> Yeah I didn't I did not expect that.
2:01:35That was a super super interesting
2:01:38answer. Um
2:01:40>> well cheers.
2:01:41>> Yeah. So so a new paradigm shift
2:01:44>> I think so. And I think you just look at
2:01:46how technology advances from one thing
2:01:47to another staying within the same vein
2:01:49you know the same it's the same type of
2:01:52technology but it is just completely
2:01:53different
2:01:55>> and so you just have to understand do
2:01:57you think there's going to be query
2:01:58fanouts or no idea like it's just you
2:02:00said you said it yourself search doesn't
2:02:02search retrie retrieval doesn't change
2:02:05>> correct how how humans think doesn't
2:02:07change you know we we like things in a
2:02:09certain organized structured type of way
2:02:11and so that always will have to be But
2:02:14the thing that's driving it, you know,
2:02:17that's driving say AI, I think will just
2:02:18be completely different type of AI.
2:02:22>> Kyle Roof, thank you for this epic for
2:02:25this gift of a podcast. That's what this
2:02:28was. This was a gift to the SEO
2:02:30community. This was a gift to me. This
2:02:32was a gift to the listeners. Gift gift
2:02:33to the community. Thank you for this
2:02:35incredible show, Kyle.
2:02:38>> I appreciate being here. I can't even
2:02:40talk anymore. I'm so I'm so taken. I
2:02:42can't [laughter]
2:02:44Uh, where should Thank you so much.
2:02:45>> Where should where should everybody find
2:02:46you and and and find your stuff and use
2:02:48your stuff?
2:02:50>> Yeah, pageoptimizer.pro. That's the
2:02:52place to go. That's where actually I
2:02:54hang out most of the day. So, I'm there
2:02:56and uh I I see most of the support
2:02:58tickets, too. Actually, if you have if
2:03:00you have an SEO question, you can toss
2:03:01it to support and say, "Hey, if if Kyle
2:03:03has time, if he can take a look," and I
2:03:04answer most of them, so that's okay.
2:03:07>> And um what are there any social
2:03:09platforms that you're really active on?
2:03:12Uh, LinkedIn is actually um that's where
2:03:15I spend most of my time. I've never
2:03:17gotten into X here. You're an X guy,
2:03:19though. I uh I never really figured it
2:03:23out. You know what I mean,
2:03:24>> dude? I I hadn't figured it out until
2:03:26recently. Only recently I started I
2:03:28started to So, actually, it's really
2:03:30interesting. Um, so X is taking short is
2:03:34taking vertical videos so seriously now
2:03:36that they're paying people $3,000 to
2:03:38post 15 videos in in 30 days. And these
2:03:42could be videos that they posted on
2:03:44other platforms. So if you're posting
2:03:45videos on Instagram or Tik Tok, you just
2:03:48repost them. And so then I had this
2:03:50theory, well, if X is paying people
2:03:52three grand to post vertical videos,
2:03:55maybe they're prioritizing them. And so
2:03:57now I'm seeing vertical video is doing
2:03:58really well on X. And um and something
2:04:01about X, I don't know if like if you've
2:04:03tried this, but if you post every day,
2:04:05it's just like a lot easier to pop off.
2:04:08And um yeah, so if you post every day
2:04:10and just respond to like one person a
2:04:12day on X, like it it works. And so now
2:04:15now I'm actually starting to grow on X,
2:04:16which for a long time was a black box
2:04:18for me.
2:04:19>> And uh
2:04:21>> that's where it is right now for me. But
2:04:22>> LinkedIn is great, too. LinkedIn is
2:04:23amazing.
2:04:24>> Yeah, I get a lot of traction there. And
2:04:25um and since we're marketing to
2:04:27marketers, that's true where they live,
2:04:28you know.
2:04:29>> Yeah. LinkedIn, they're all amazing.
2:04:30Instagram is great. Tik Tok sometimes
2:04:33great.
2:04:34>> I I I'm hopelessly addicted to Tik Tok.
2:04:36>> To Tik Tok. Yeah. [laughter]
2:04:38>> Yeah.
2:04:39>> I can't help myself. I start doom
2:04:42scrolling.
2:04:42>> I doom scroll sometimes. Um Yeah. All
2:04:44right, Kyle. This was a lot of fun. Uh
2:04:46this is episode 1,213
2:04:49of the podcast. 1,213
2:04:52days in a row doing this show. Let's go.
2:04:56>> Amazing.
2:04:56>> Thank you so much.
2:04:57>> If you watch this on YouTube, thank you
2:04:59so much for watching. If you listened on
2:05:01Spotify or Apple Podcasts, thank you so
2:05:04much for listening. And I will talk to
2:05:06you again. I want to talk to you again
2:05:10tomorrow.
2:05:11Bye now.

About this transcript

This page contains the full transcript of Kyle Roof Ranked Gibberish #1 on Google. Then Google Deindexed 20 of His Sites. by Edward Sturm, generated from the public captions YouTube serves with the video. The transcript has 25,922 words across 3,728 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.