Campfire Chat with Joe Hall

Noah: There we go. Everybody welcome to our sixth edition of Campfire chats. Welcome everybody. I'm joined today by one of my favorite people working in digital. Joe Hall, a few words to describe him. Master auditor of websites, technical SEO expert that's tons of brands, including buffer. He's worked with some of the best agencies around, including fractal.

He's an accessibility leader. He's a speaker. I was lucky enough to share the stage with him at MozCon in 2022. He's an educator. He has tons of videos on his website, hallanalysis.

com for you to check out. He's a consultant, a strategist, a writer, a builder. One of my favorite things about you, Joe, is that you're a food. No, no, no.

He's like a master junk food aficionado.

Joe Hall: Yes. Yeah, thank you for clarifying that. Yeah.

Noah: He's this, yeah, he's a South Carolina Gamecock fan, a gentleman, a scholar. He even got a big data and social analytics certification from MIT, which is super rad. I'm really stoked to welcome you to our edition of Campfire Chat. How you been?

Joe Hall: Thank you.

Joe Hall: Thank you.

Joe Hall: Thank you so much, Noah. I'm great. Thank you so much. I'm so happy to be here. It's a huge honor to be a part of this and to engage your community, which I am a huge fan of. I'm in such awe of that community and the people behind it. It's amazing to see. And I'm so happy to be here. It's just a pleasure.

to talk with you today.

Noah: you too. So I want to jump right in. I think there's all kinds of stuff that we can talk about. I know that you came ready to launch into something that you care a ton about. So do you want to start us off with that? And then I'll sort of lead us in an organic direction.

Joe Hall: I can, yeah we can do that, I gave you little preview of it already, so if you want to stop me at any point and be like, Joe you're sounding a little off, that's fine, we can go to a different topic, whatever. But basically the big thing that I was thinking about today is, other people have spoken about this before,

Noah: Yeah.

Joe Hall: But I've just really been thinking about this lately. I think that as an industry and a community, as SEOs or people that care about organic traffic, we are quickly approaching a situation where we don't have very many reliable sources of data anymore.

And it's important that we understand that, not to say that, you know, Medina we're looking at is not correct, but it's just maybe not as reliable as it once was. it's not, when I say reliable, I don't mean it's not correct. I mean that it is simply just not known if what we see is what's really happening.

And I know that sounds wacky, but I think there is some issues that have happened in the last several years that are serious causes for concern that we need to think about as a community and really kind of wrap our head around what is happening so that it doesn't get more out of control going forward in the future. Yeah, I feel like I need to

Noah: Tell me more, walk us through this.

Joe Hall: provide some notes here. So basically, I feel like a lot of this for me started happening when Google Analytics 4 rolled out. There was a lot of criticism with it when it rolled out. Some of it was genuine criticism. A lot of it, though, was just people not wanting to change and not wanting to go through the process of

learning a new system. totally understand that. But to give IntelliX more its props where it's due, it does, in my opinion, it does give more control over the IntelliX experience than Universal did.

It takes more work to set it up. It takes more work to understand it.

It is a better piece of software than universal analytics. With that being said though, my main concern with analytics for is that there is an obvious inherent bias within the way it is designed and the way that it provides data towards

paid search channels and that could be paid search, could be, I mean, paid traffic channels. That can be paid search, could be display ads, could be paid social, but there is an inherent bias towards that paid model. And that is understandable.

I mean, that is kind of the point of Google offering it for free.

is because their main business model is around advertising, right? you know, it would make sense if they would put that inherent bias in there. It does, however, feel like SEO and organic traffic in general has been somewhat, just thought, afterthought with GA4.

An example would be that when it first launched, there was no landing page report.

Noah: Mm-hmm.

Joe Hall: and that there was no landing page dimension. They eventually added the dimension and so that you could then go in and create your own report. And then eventually they added the landing page report as part of the default setup under the engagement collection. So that was a good step in the right direction, but overall, that was a very small thing.

that they kind of gave the SEO community, you know? And I think that even with universal analytics, I think the kind of the trade-off that they, what kind of the bone they threw to the SEO community was that you can integrate Search Console and that somehow that fills a lot of the gaps. But even people like, you know, other analytics people have said that there are

large parts of Google Analytics 4 that are not available to people who don't have ads accounts. And some of those tools can be helpful for SEO. You wouldn't know it if you didn't have an ads account, you know?

So that's another thing to consider. Another issue that I've looked at is that the session duration of a default setting is way too low. It's only 30 minutes.

That's ridiculous. It's one reason why direct traffic in most people's acquisition reporting is through the roof. It just doesn't make sense why you'd have that much direct traffic, you know?

And it's all because the default setting for session duration is only like 30 minutes or something, which means that the sets are...

Noah: How many tabs do you usually leave open?

Joe Hall: Exactly, like I have way too many. I have like one two, three, four, five, six. I have seven open right now. And I literally turned on my computer maybe 20 minutes ago to get prepared for this. you know, and that's just an average user is going to have a lot of tabs open. It's going to reset the duration, reset the default channel grouping on each session reset. So I think that that is

That's definitely getting in the way of organic reporting. And then another thing, and this is my last point about Antimod Explorer, is, and we've known this forever, this is not a new thing, but it's something that not many people talk about. Last Quick Attribution is horrible for SEO because generally speaking, the consumer journey is

With SEO is long. I mean it usually goes through some sort of discovery phase a research phase a discovery phase a Comparison and product you have a the marketing funnel works. So it's yeah, if you think about that that attribution model last quick attribution It's really only geared towards Like a paid experience, you know It doesn't work with

Noah: Sure, it's messy.

Joe Hall: Organic SEO very low right and you can tell that a lot of Google Internet Explorer is kind of all focused around that model of attribution. So that's my whole stick about Google Internet Explorer and it's fine for Google to create a biased piece of software that doesn't really that's not an example of any ill intent or any you know conspiracy or anything like that. That's just

Noah: Mm-hmm.

Joe Hall: basic business logic, Since then though, there have been a series of things that have happened that at very best we can assume that this is just a oversight or lack of interest on Google's part. One of the things that happened was there's this issue with merchant center tagging.

Noah: Mm-hmm.

Joe Hall: So for those of you that are not aware, Merchant Center allows you to upload a feed of products, if you're an e-commerce website, into this Merchant Center feature that allows for those products to be displayed in Merchant Center and search results, and you can easily turn them into advertisements as well.

there is an option within Merchant Center to do automatic tagging, which essentially adds a perimeter at the end of the product URL inside Merchant Center that helps Google Analytics report that any traffic from those URLs are coming from Merchant Center. Recently, or maybe last year it started,

users started seeing product pages ranking in regular organic results include that same perimeter. And that was really weird. Like that shouldn't be happening.

should only be, that perimeter should only be showing up within merchant center results.

They went and looked into it and basically came back and said that they were aware of the problem, but we're not going to change it or did not seem like there was much interest in fixing the problem. And it might seem like a small, miniscule problem, but for a lot of customers, are using Merchant Center. They're now looking at analytics and

worst-case scenario they don't know this problem exists and so now they're seeing that their immersion center traffic is going through the roof when it might not really be it might just be coming from organic results or if they do know the problem exists then they may just not understand how much of that traffic is genuine or not and I think that that's a problem as well is that all these issues that I've talked about already

Joe Hall: and that, you know, I have a few more talk about. The problem is not simply that like the traffic may be wrong. The problem is that we don't know if the traffic is wrong because there is no clear guidance on if these issues are going to be fixed or if there's any sort of end result to this. Another one was

There is no reporting for AI overviews in Search Console or anywhere else. So a lot of people are just kind of wondering, am I, you know, am I, do I have an impact there or not? They're not sure.

They've now taken a much more aggressive approach at blocking third party tools from accessing search results with rate tracking.

Scrapers, they've always done that. They've always done that. That's always something they've done, but it's become a bigger issue now that you have AI crawlers probably scraping their results as well.

And all of this, each one of these issues that I mentioned can be easily explained by someone.

sound very innocuous and not really anything to be concerned about. But when you take all of what I just said and then you consider all of the stuff that we learned in the Department of Justice case last year about how they are increasing queries to drive ad impressions up and how they have

shifted results in ways to provide more click backs to the result landing page and all of these things that we were told they weren't doing. We were told they didn't use click data in rankings. All of this stuff we were told they were not doing.

We were always told that there was a very hard line between search quality and ads department.

Joe Hall: When we see all of this other stuff happening behind the scenes, and then all of this other at the very best are kind of over reluctance to address real issues that SEOs have, you start to wonder if maybe these are not bugs, but they're actual features. That they have, okay, okay, okay, okay. I'm gonna take a sip of coffee.

Noah: All right, all right, I gotta stop you. I gotta stop you. Hold on, so can I reframe? Yeah, you take a couple sips of coffee. So here's what I'm hearing. We don't have the data that we need. And what you're saying reminds me of two of my favorite people. One is Renee Bigelow, and what she would talk about is something called attribution crack. And that...

All we want to do is to be able to properly attribute all of our revenue to specific channels so that we can argue for budget and all that stuff. That's attribution crack. And then another one of my favorite people is Robin Allenson.

And he is a great saying when, when he gets in these discussions with clients where they need to know exactly where the money's coming from his comeback to that is, we optimizing for revenue or are we optimizing for certainty?

Joe Hall: Yes. Yes. Yeah.

Noah: So I guess the question that I have for you is what's really important here? Like, shouldn't we be thinking, and I've been reading through a bunch of your writings and it's like, we always start with business goals and the business goals typically involve money. So like, shouldn't we be thinking about that more than our ability to prove value or how do we kind of navigate the waters there?

Joe Hall: Well, you're right, Noah, but all of this goes back to business value and business goals because because my my hypothesis And by the way before I go any further I'm well aware that I sound a little naughty today. Okay, so Just just put that out there. I'm just you know, I'm just

Noah: Yeah. Yeah.

Noah: You don't, you don't. I'm just like, I'm sitting there thinking as you're talking and.

Joe Hall: Well, I always say that because I'm getting ready to sound a more down here. I think that when we get to this business, this approving, optimizing for revenue, the bigger picture here is that even with this merchant center tagging issue, there's always a clear path

back to buying more advertising. And if we are looking at ways to optimize revenue, then we need to focus our efforts on the channels that have the highest RLI. And time and time again, for as long as I've been doing SEO, everyone, including people that don't do SEO, says that organic traffic has the highest

are alive. And so there is a very clear, in my opinion, there's a very clear

business objective for Google to start to cast more doubt on that fact. If Google can create more what they call FUD, know, fear, uncertainty and doubt around the SEO process, then it just puts more people into paid search because they buy more ads if they don't think

Organic is a reliable source of traffic. And if we consider after 24 revenue in that kind of lens, then if more and more businesses make that decision to go to paid search or any other paid marketing channel over organic, then it's likely we'll start to lose

Joe Hall: uh, more money, um, because the ROI just won't be there like it used to be there. Um, so that is kind of like my answer to that is that, that fundamentally this isn't just about like optimizing for certain media or accuracy or whatever. I mean, to a certain extent, you can't really ever do that. Uh, even with the best, you know, most honest platforms out there, uh, you can't do that.

But from a purely business standpoint, you have to ask yourself, are you being, like, is your view of reality influenced by the people you pay? And I think that's what's really important to think about with this, you know?

Noah: Hmm. Because Google has a vested interest in us buying ads. Yeah.

Joe Hall: Yes, they have a huge investment interest in the ad market. And like the Merchant Center thing, back to that, there is a much shorter path from that problem to turning those Merchant Center listings into ads than there is with any other organic issue. So the fact that they're not going to fix the problem

It just makes me believe that there's a real reason for that. And again, like I know it, you know.

Noah: Mm.

Joe Hall: Conspiracy theories are annoying, but I have always been a true blue, organic SEO. I've never touched paid search. And I think because of that, maybe I'm biased. Maybe I am not seeing the situation like it really is. But I also have always been pitted against paid search my entire career.

It's always been about, if organic can't deliver, then we'll put more into ads, you know?

Noah: I like where your head's at about this. Specifically what I'm hearing and what I've been feeling for a long time is that doing what we do, relying on precision and accuracy is never going to win. know, like we're constantly playing poker and the data is obfuscated in such a way that we'll never have real truth about anything. And because of that,

Joe Hall: No.

Noah: I feel like we need to make decisions based on the information that we do have. And directional information is often enough to make bets on what types of content or strategies we need to test and execute on. And...

Joe Hall: Yeah, absolutely. Now, yeah, I mean, what you're saying is actually true. to be clear, I don't think that the data we have now is unusable or taken beyond, you know, taken beyond utility. But I think it's something that more people need to talk about at least that there is this, there is a there is definitely a bias and a

A conflict of interest on Google's part for all this, you know.

Noah: Mmm.

Noah: Let's shift gears. I got a whole stack of questions. I got a stack of questions because you're just a fascinating guy. Okay. one of the things that strikes me when I look at your website are the different services that you offer. One of them is SEO audits. It seems like that's been a cornerstone of what you've been doing forever. Tell me about like, what is the value of them now?

Joe Hall: Okay, yeah, let's get off that. Yeah, please.

Joe Hall: Yes, yes.

Noah: Has it changed over time? How has your analysis changed over time?

Joe Hall: Well, it's interesting because...

Joe Hall: the search engines have caught up, I think, to a lot of the development community since I first got started, at least. When I first got started, you know, they did not process or crawl JavaScript. It was, you know, always just flat HTML. And

Noah: Mm-hmm. Mm-hmm.

Joe Hall: canonical tags do not exist and there was a lot of issues with canonicalization, you know. So I think the search engines have evolved a lot to be better at not making these same mistakes, you know, while crawling and indexing. So I think they've changed a lot in that regard. However, you know, I still come across

sites. did this a couple days ago. came across a site still that was using client-side rendering and none of the pages were being rendered by Google. And I mean, oftentimes, you know, Google's pretty good at that nowadays, but it's still an issue.

And so I still have to check for all that kind of stuff, you know.

Noah: oof

Joe Hall: So I think the audience have changed a little bit in that they do tend to be more strategic than tactical as they used to be more like, okay, you you need to fix this issue with your knuckle URLs because you don't have knuckle tags or, you know, the search engine can't crawl this page because it's in JavaScript or whatever. It used to be, those used to be bigger problems than they are now.

Noah: Mm-hmm.

Joe Hall: And so now my audits still focus on things like that. That's like the bottom, you know, that's like the floor where we start. But then the audit usually tends to be more strategic and utilizing technical SEO in a strategic way. And so, for example, I'll say, look, you know, these pages you can no index and instead

Noah: Mm-hmm.

Joe Hall: let's push more of your internal Lake Equity to these other pages. And I'll show the breakdown flow of Lake Equity throughout the site and show how to better structure their information architecture and stuff like that to make it make a better, you know, SEO performance. But yeah, I would say it's less tactical and more strategic now. The tactical stuff really...

Noah: Yeah.

Joe Hall: still happens in every audit. But I think that the search engines have done a much better job figuring out how to crawl and index just about any type of site now, whether or not there's problems or not with it.

Noah: Tell me what you think most other SEOs miss when they're looking at websites, because I'm sure you inherit other people's work all

Joe Hall: I do, yeah. So I think a lot of SEOs, well it really depends on your background really. I like to think of SEO as, we're looking into those three areas, technical, content, and links. And I think there's a lot of SEOs that do content. I think there's less SEOs that do technical.

With content people, I work really well with content people. Some of my favorite clients to work with are folks that do incredible content work. And they are very good at developing content and brand identities and stuff like that that bring in all of the right elements for SEO.

Like they bring in users, bring in links oftentimes, but they may miss a lot of the technical foundations that can really punch up that content and get that content into where it used to be. So a lot of people miss like doing the right optimizations for things like featured snippets and structured markup and rich snippets displays, kind of stuff.

Noah: Good.

Joe Hall: And then a lot of people still to this day, even technical SEOs don't think a lot about the influence of information architecture and internal linking and understanding how that has a major influence on crawling and indexing and your whole SEO performance. So I see that missed out a lot as well.

Noah: got one for you. The best SEO tool you use is the one between your ears. Talk about that.

Joe Hall: Yes, yes, I say that all the time. Yeah, the woman. OK, thank you for I was about to say, my God, we're just alike. So yeah, so the best SEO tool you can use is the one between your ears. And what that means is basically anyone can run a site through screen from right and screw from is amazing. It's probably.

Noah: I know, I know I, you know, I did a little research.

Joe Hall: My second favorite SEO tool is Scrummivrog. The one between my ears is my most favorite. But anyone can put a site into Scrummivrog, set the configuration the right way, and get almost all the data they need, right, for a technical SEO. But understanding how to utilize that data the right way and to make informed decisions.

takes experience and it takes understanding what works best for the search engines. A great example this would be with Screaming Frog, you can get a crawl-directed map of internal links on the site, which looks super cool.

But if you're not familiar with the type of crawl pattern that typically aligns with specific algorithm updates, then you wouldn't know that that pattern needs to be fixed. So for example, I've noticed that a lot of core updates, one of the similarities I see a lot is having

sections of content that are siloed off from the rest of the site architecture. And in that crawl graph report in Scurrying Frog, it shows up as sort of like dandelions. I call it the dandelion pattern because it's like these, you know, shoots of winks that go out and then like there's a huge cluster of pages over here.

They're all winked together, but they're not really well winked together elsewhere. That type of pattern.

I think is indicative of a core update. And sometimes that pattern in the graph is not fully accurate of what's happening in reality because sometimes those pages do link together in ways that the crawler doesn't show in the chart. So knowing that and knowing what to look for I think is incredibly valuable. And knowing how to interpret

Joe Hall: other metrics inside those reports in a way that provides value, think is critical. And that's what sets an SEO apart from just someone that exports CSV sheets and sends it in an email. Having an understanding of what the data means and how you can best utilize that data to make decisions, I think is critical.

Noah: Let's go deeper here. So do you have a decision making framework for how you like to think? Do you think in a structured way or unstructured?

Joe Hall: Okay.

Joe Hall: That's a good question. So I've thought about making a framework, but I don't have one. I probably think sort of unstructured.

But usually my process is this. If I go through the steps to do the audit by collecting data, crawling the site, and then as I'm making a list of all the tactical things, like duplicate title tags and stuff like that, as I'm making that list of things that you need to do to be fixed, I'm kind of looking through the data itself.

And I'm making notes on anything that looks out of ordinary and anything that would indicate that there's a bigger problem at hand. And then kind of going back, and then I take a break after I do all that and come back to it and try to look at the whole thing from a bigger picture kind of lens.

To see what are the bigger problems. I'll give you an example a client

had Quiet last year, I guess, and they had about three or four million URLs. And they were legit. was legit, you know, Quiet. The content on those URLs were legit.

Noah: Mm-hmm.

Joe Hall: you know, decent content on those pages. The problem was that the search engines had that thing, you you see in search console called discovered but not crawled. And it was massive. was like almost all of their pages were like that. I think they only had about 250,000 indexed and they wanted to fix that problem. That was what they came to me for.

They wanted to get all their pages indexed. I explained to them, well, at 2 million pages, you're not likely to get all of them indexed, but we could get more than 250,000, you know. So, through...

I went through the standard process and doing audit. In doing that, I realized that they had...

on the pages that were not getting indexed, most of them had a massive number of requests per page. So it was like, on average, was like 97 requests per page. And so I explained to them that, you know, this is, when you talk about crawl budget and the amount of crawl budget the search engine has to go through,

Noah: Thanks.

Noah: Mm.

Joe Hall: to index a page like this, a site like this, you're not looking at like 2 million URLs. You're looking at more like 200 million because you've got an average about 93 requests per page. and they had like, I shouldn't, why? I don't really matter. I'm not telling you their name, but they had like these ridiculous, like little sprites all over the top. was like,

In my 26 little PNG files, they had about 30 something JSON files that loaded through Next.js framework. It was just over the top. Way too many requests needed to made.

Noah: So is this about the same time that you posted on social? Does anyone have a cool and easy tool for checking requests per page? Because I saw that on Twitter.

Joe Hall: Yeah, I that's right, Yeah. God, are you stalking me? This is a little bit weird. Yeah? Well, you did a good job.

Noah: No, I just wanted to come prepared.

I did a bad job promoting the episode, so I wanted to kill it in the room.

Joe Hall: Well, that's cool. That's fine. But yeah, but yeah, you're actually right. That is exactly, that is exactly around the same time that happened. By the way, you can get that number with Screaming Frog. just, the number, the request per page is in Screaming Frog and then you can just give the average in the spreadsheet. But anyway, so yeah, I mean, that's part of my process is going through this, going through the normal steps.

and keeping any issues or problems that the client's having in the back of my mind while I'm doing that and then looking through the data while I'm going through the steps and finding evidence that could contribute to those problems. And so the solution for that client was to consolidate their sprites into one PNG

You can do that and then you can like use CSS to basically move the PNG as a background, but you just load one PNG for like 30 images. Do that and then preload that in the HTML so it's preloaded. That cuts down on their cumulative layout shift like crazy because all those little sprites were loaded individually. And then I said,

you need to open up some these JSON files about half of the JSON files that are reloaded through Next were blocked by robots.txt so I had them open that up and then had them I had them uh what else did I have them do I had them oh try to minimize the number of requests per page so if there's any way you can consolidate some of the JSON into one request so you're not just pulling like

you know, 30 JSON files, you know, per page, that would be ideal, you know. And so their engineer team went back and they started implementing that. And after about a month, we started seeing indexation rise significantly.

And I don't know where they're at now, but it really didn't make a huge change in difference. But all of that was a result of going through the data with a fine-tooth comb while I'm...

Joe Hall: Going through normal steps and everything making notes of what's going on Keeping their specific problem in the back of my mind And understanding that the entire audit is meant to fix that one problem, you know And keeping those issues in the back of my mind as I kind of make notes for what's going on, you know So I guess it's not really structured But like I said, it is like that tool

between my ears is the most important tool because like I wouldn't have known to to identify certain things if it weren't for that experience and that knowledge, you know.

Noah: But also it's like designing a process in a way that allows you to systematically uncover things. it, right? Like that's, that's pretty rad.

Joe Hall: Yes.

Joe Hall: So that's my, so that's actually been something I've thought about for the last five years. Can I design a process that can replace me? Right? So like, can I design a process where I can hire someone to basically replace me if they just follow this process? And I've always kind of come to the conclusion that I can probably have them do like 60 or 70 % of the work.

I still have to go back myself and go through my own process to fill in the gaps, you know, because like I said, I can't pull that tool between my ears out and give it to someone, you know? It's stuck there. So I have not done that yet.

I have not replaced myself yet.

Noah: We got all kinds of questions in the chat. So I want to, I want to throw one in. So how many requests is too many?

Joe Hall: yeah, let's hear the questions.

Joe Hall: It just depends, man. I mean, it depends on the site. As few as possible is great. The example I provided just now is, like I said, he had 2 million pages. It really comes down to, you know, we've heard Google talk about this before, that most websites don't need to be concerned about crawl budget. I think that's true. But I also always tell clients like,

Noah: as few as possible. mean, it's...

Joe Hall: Yes, it's true. You don't need to be concerned about crowd budget as a metric or as a concern But that doesn't mean that just because you don't have diabetes You should eat like 12 cookies during lunchtime, right? So if you you still need to be kind of mindful of Things like average request for page. I think that most sites

can get away with, I don't know, like 20 or 30 per page. But really, as few as possible is better, because that's going to really help users and crawlers access the page.

Noah: I'll never forget working on this one website that was an e-commerce website that was selling road bikes and they had a tiled grid that was probably a thousand pixels by 400 pixels with, don't know, eight pictures in there. Each one's 200 pixels wide or whatever. However, each one of those images was 5,000 pixels wide and scaled and the page was like six.

Joe Hall: Yeah.

Joe Hall: my god. So they were like two megabytes each or something? No.

Noah: Yeah, it was like, it was like a 65 megabyte, page weight. And the site had ranked number one for like specialized, which is a brand road bikes. And then when they got a new photographer who didn't know what he was doing and didn't downscale the images to be 200 pixels or whatever the max width they'd ever be used, all of a sudden their rankings went from, you know, like ranking number one to ranking 28.

Joe Hall: Yeah.

Noah: and all their revenue tanked like overnight. Super brutal.

Joe Hall: Yeah.

Joe Hall: I think that, you know, and that's a good point, you know, when it comes down to like page performance and page speed and core web vitals, I, it's, you know, I don't, always talk like, you don't need to obsess over that. You don't need to have a perfect score, but you absolutely should not go to the extreme in the other direction. Right? So like, you don't need to be perfect, but you also don't need to have a page that is 65 megabytes either.

Noah: Mmm.

Joe Hall: I mean that's just too much, you know? So really just keep it in check. It doesn't need to be perfect, but doesn't need be kept in check, you know?

Noah: Tell me what does a bulletproof infrastructure, you can't see I got an air code here. What does a bulletproof infrastructure look like?

Joe Hall: That's a really good question. think a bulletproof infrastructure is going be different for every site, but it is definitely going to be one that has the path of least friction for the crawler to every page of a site. I'm a big fan of HTML sitemaps, and I think if they're done correctly, they can be really effective.

not only for SEO, but also for user navigation. I think that, you know, for SEO, having an integration architecture that allows for the path of least resistance for the crawler is critical, but also it has to help the user as well. And I used to get really upset when people would talk about like user experience for SEO.

But I am a huge advocate for it now. Not only because of how search engines now render pages completely during indexation, but also because you have no idea who visited your site is a potential link magnet or potential linker, right? And the easier the site is to use

and the more accessible and the more easy to access content, the more likely that content is going to get linked to. And that's critically important for that other part of SEO, is part of your, I guess, theoretical conversion model should be how much links are we getting out of

the traffic to the site, you know, because that's the other part of SEO is the links and create a website that's easier to use where people can find what they're looking for as fast as possible. It's just gonna increase the likelihood that you get those links. So that's also a critically important thing. But I think those are the two things is, you know,

Noah: Mm.

Joe Hall: It's for users and then also for the crawler, the path of least resistance for everything.

Noah: Let's talk about a couple more things. Keep this answer super short. How many lunchtime cookies is too much?

Joe Hall: Alright, no comment.

Noah: Okay, I agree.

Joe Hall: Yeah, that video is here.

Noah: Yeah, sorry, Mark, Mark Alves, but we refuse to comment. Okay. So the, the next thing I want to talk about is tied to what you were just talking about. What the hell did you build that gathered 2 million links?

Joe Hall: Okay. Now...

Noah: You built a widget on a site that gathered something like 2.4 million links. What the heck was that?

Joe Hall: Oh, God, you've really done your research. You should go on that show Hot Ones. So yeah, that must have been a really old tweet I put out there.

Noah: I don't remember. It stuck out though. couldn't drop that. had to learn more about that.

Joe Hall: So, yeah, so I can't tell you specifically what it was, but it was a WordPress widget, or a WordPress, yeah, it was a WordPress widget. And back in the day, and actually, there's a little bit of a history behind this. So back in the day, the plugin team at WordPress.org

were not very SEO savvy. And you could insert links into plugins. yeah, so you can do that. And you basically just create a plugin, upload it to the repository.

And then in the footer of the site, there's a link that goes somewhere.

Noah: I did it too.

Joe Hall: And the way that I did it was not... It was super authentic in my mind because the link essentially just went back to the site that created the plugin and back to the plugin rating page so they could download the plugin. It wasn't like a real commercialized link or anything like that. It was a very authentic link.

But that link basically generated about 2.4 million links. I can't remember the number of domains that it came from because, of course, it was a site-wide link on every site.

So it was a massive amount of links, but I don't remember the actual number of websites. Now, I'll give you a little bit of a backstory behind that.

Noah: Mmm.

Joe Hall: So I did that and I was tickled pink with it because it worked so well and my client was super happy with it. They went through the roof in their rankings after that kind of taking off. The plugin was very popular though, I won't mention that. And also they put a little bit of public relations behind it so it really caught wind. However, after I did that...

I started noticing other things popping up and like I started noticing like and plugins to add the Google Plus button to sites and then like all of a sudden after you added it to your site there was like a link in the bottom to like an online casino or something. There was that kind of stuff going on and back then I was really

Noah: F

Yeah.

Joe Hall: uh, infatuated with open source. I still am. I still am just in love with open source. Uh, but back then I really felt like a sense of duty to protect the WordPress, uh, project. Um, and I knew a lot of core developers back then. Um, I still know some of them. Um, and so I started to help report those issues.

to the plugin team because they were not versed in SEO and they did not understand why those links, I mean can understand that it was like a, they looked at it as like a free ad basically. They didn't look at it as like a manipulated search results. So I worked with some of the people on that team back then and now they have much stricter guidelines for plugins.

Noah: Mmm.

Joe Hall: and links found in plugins. And so now I don't think you'd get away with that at all. But in large part, I feel like it was because of my early education with that team. Like, hey, you guys need to be able to look out for this. And I will admit, I did not in that vein, I did not mention my own plugin. But I definitely wanted them to identify anything coming forward in the future.

Noah: Okay, I got one for you. Keep your answer short on this one.

Joe Hall: Yeah, we're coming up to an hour. Yeah.

Noah: What's your favorite underappreciated national food chain?

Joe Hall: my god. I don't know if they're national, but I really love Bojangles. And I also really love Arby's. Arby's gets a lot of hate, but I like them. I really love Bojangles too though. I don't know if Bojangles is national though, is it?

Noah: Okay.

Noah: I don't know. Hey, so I got a question. I went to Tech SEO Connect last October and I was introduced to boiled peanuts. And I know you love boiled peanuts. Help me understand what's cool about them. Cause I had a hard time getting into it.

Joe Hall: Okay.

Joe Hall: Okay, cool.

Joe Hall: Yeah.

Joe Hall: Where is TechSU Connect? Where is that at?

Noah: Raleigh, North Carolina last year.

Joe Hall: Okay, okay, so you're the good the good part of the country, okay, so So yeah, I live in South Carolina and yeah, if you were in Raleigh, you probably had access to bull peanuts. I love bull peanuts There's a handful different types you can get But it's just you know, it is what it's it is what it said. It's boiled peanuts They are kind of salty. They have a real good What do they call?

Noah: Yeah.

Joe Hall: the in the money that later or

Noah: a wet salty slime all over your fingers that you really want to deal with.

Joe Hall: Yeah, it's very good. Yes, it's very good. I love it.

Noah: That was the hard part for me, was this consistently wet salty slime that I just couldn't get off my fingers.

Joe Hall: Well I don't know why, you know, it sounds like, do you know who made the nuts?

Noah: I don't know, but we went out with Matthew Kay, who's a total foodie, and he was like, this is the best restaurant in Wally. And we walked across the city to get there.

Joe Hall: Okay.

Joe Hall: So the nuts, okay, so the nuts were at a restaurant. Okay, so I will say that if they were slimy, it means that they had been sitting out for a while. Because, I mean, it's peanuts, they're not gonna go bad. But still, if you get them fresh right out of the boiler, they're not slimy at all. But still, I will pull out a bag.

Noah: Yeah, totally. They boiled them there.

Noah: Mmm.

Noah: Yeah.

Okay.

Joe Hall: of whole peanuts out of the refrigerator and they will be slime on them and I will still eat them.

Noah: Yeah.

Okay. one of the things that's always super attracted me to you is just how you're a builder. You and I both love creating things and we get a lot of value from that.

So, I want to know how you're using LLMs these days, what your coding kind of setup looks like, or using cursor, windsurf, VS code, like what kind of LLMs are you using if any?

Joe Hall: Yes.

Joe Hall: So I'm primarily using ChatGPT. I have an account there. I really like it. I probably would really love Cursor, but I still use Atom. For anyone that follows me on social, you know by this point that I am kind of grumpy all the time. And my personality is, yeah.

Noah: Okay.

Noah: Mm-hmm.

Noah: No.

No.

Joe Hall: My personality though tends to shy away from new things. And I know that cursor is pretty great. I know a lot of these new AI editing tools are awesome. But I just haven't gotten into them yet. I started using ChatGPT for content stuff and other things and playing around with the API. And then I tried using it for code.

Noah: Mm-hmm.

Joe Hall: And the coding has gotten a lot, lot better. It has gotten so much better. And so I have used it for some coding work. I think of it in the same way that I used to go search on Google. For example, I might search on Google how to

Noah: For sure.

Joe Hall: Just an example, might say like how to loop through a multi-level array in PHP, right? I used to type something like that into Google to see like what Stack Overflow said, you know, that kind of thing. Instead, I'll type that into Chatch UPT now and it will give me as good of a answer as Stack Overflow did. Now that is not saying

Noah: Minus the arguments. Minus the arguments.

Joe Hall: Huh?

Minus the arguments and then that is not to say that stack overflow is the best answer either But it is it is I think when it to coding I think that chat CPT does provide a good enough solution to build out at least a proof of concept and a working application That is ready to be a viable

product. And then later, if you want to change it, refine it, optimize it, then yeah, you could do that. to get it started to develop your project, think that AI, LLMs are definitely capable and are a good place to start. I don't like the idea of the no code people using LLMs because it seems so

Noah: Yeah, I agree, actually.

Joe Hall: dangerous. I've seen so many, you have no idea what it's producing. I've seen so many instances where the code will leave open, will leave it open for SQL injections, their security problems. I mean, it's just stuff that, that if you have coding experience, you know that, that needs to be fixed, you know? But if you are a no code person,

Noah: You have no idea what it's producing,

Joe Hall: It just seems so dangerous to not have that level of experience and use AI in that way, you know.

Noah: Rad, okay, we got to wrap up. I want to do a couple things. I want to read one of your favorite, one of my favorite tweets that I read from you. And then I'd love it if you could share how people can follow you. And then we'll kind of take it from there and I'll do a final wrap up. But one of my favorite tweets of yours was, the older I get, the more confident I am that no one really knows anything about anything.

We were all a bunch of weirdos trying our hardest to act normal, whatever that is. Love it.

Joe Hall: Yeah.

Yeah, I mean that... Do you want me to comment on that or...?

Noah: Okay, sure.

Joe Hall: So that actually goes back to the beginning of this talk. think that from an SEO perspective and a data perspective, I find myself questioning things so much now. And the things that I thought I knew related to data, but also to work and life and everything, I find myself questioning as I get older.

And I think that that's probably natural for most people to go through this period of confidence when you're young. And then the older you get to start to question that confidence going forward. And it doesn't mean that like I'm unsure of myself or anything like that.

It just means that the ways that I'm doing things and the ways that we thought about the world, maybe in our youth.

are not so concrete, you know? And we need to have a more open perspective and open mind to the way the world works, you know? Yeah.

Noah: Mmm.

Noah: I agree.

Joe, this has been an amazing hour. Love it. Can you let everybody know how to follow you on socials?

Joe Hall: Yeah, sure. You can find me, uh, at Joe Paul on Twitter. Um, and then I'm on Blue Sky as well. And my Blue Sky handle is... It's so complicated. I made it too complicated. But I think it's BSKY.Holoanalysis.com. Uh, I should have just used my name. I don't know where I was thinking. Uh, but anyways, I'm on, I'm on Blue Sky. I'm on, uh, Winton.

LinkedIn is where I have more serious conversations about SEO. My other socials I'm not really that serious on, but LinkedIn I have more serious conversations. But yeah, those are the best places.

and then also clawanalysis.com is my website. I don't update the blog as much as I'd like to, but there are...

Noah: Mm-hmm. Yeah.

Joe Hall: There's a blog and there's some tools and some education content in there. So yeah, check me out on those places.

Noah: Awesome, you rock. So I get to just say a couple quick words of thanks to everybody. I want to thank all of our viewers. I want to thank Amy because she's our amazing producer rocking behind the scenes. I also want to thank all of our community partners who've made this all possible. One of which is Aerobs that we're going to be joined by in two weeks on our next community chat. And the last thing.

Joe Hall: Thank you.

Noah: is I want to thank all of the members because we just crossed 4,000 members today, which is just super rad. was like, you know, it's a vanity metric among vanity metrics, but like it was super fun. Yeah, it's neat. So everybody, thanks so much for watching and we'll see you in two weeks in the next one and we will see you soon.

Joe Hall: Wow, that's awesome.

That's so cool. Yes. That's so cool.