Transcription
Indeed, that's not a bad idea. Great. Now, another item for my list. Can you find me now? Yeah. Thanks a lot, Terry. Another thing on my list.
Hello everyone. Welcome to another episode of SEO Fight Club. I'm just popping out the YouTube chat here and I'm saying hello and reminding people to ask questions. Uh, be sure to ask questions, uh, early. You can also suggest topics. We don't limit it to anything. Um, but sometimes we run out of time and we can't get to everything. So, if it's at all urgent for you, be sure to ask it early rather than later.
All right. So, with that said, a lot of people have been waiting for an update from me. I'm getting very close to a prototype of the new technology. I don't want to give out too many teasers. Um, but uh, yeah, working, uh, uh, very long days, back-to-back, weekends included. So I'm on the mad dash, uh, to get a beta program started. Uh, so that'll be coming around in the next, uh, I'd say probably 2-3 weeks. Uh, uh, hopefully sooner if possible, but it depends how the coding goes.
Uh, Ted, is is this a new version of Kora or is this a brand new program? This is a brand new program. So, uh, it's I would call it the next incarnation of Kora. It's not an update to Kora. It's a reinventing of it given the way the world has changed with, uh, AI and what Google has changed in terms of, uh, requiring rendering to even see the search results, uh, taking what we know about volatility and putting it into the core. And everybody's, uh, most, uh, favorite feature request automation is built at the core. You give it a list of keywords, you set it off on the whole list. Uh, so yeah, total rearchitecture, whole new inventions.
Um, will this work in will this work in conjunction with Kora or does it replace Kora or are you still not sure yet? It's probably going to replace Kora because most of the industry doesn't understand what's actually changed and what it means. There's a lot of people I've I've even tried to tell and they're just in denial about it. So, I figure once we get the tool set out and people can start seeing the difference and finding better opportunities for their time, uh, that'll make all the difference. uh, but you know, just explaining academically, uh, why their process is broken, uh, doesn't help them really do anything better. Uh, and so that's kind of the problem. Um, you know, recognizing the problem is something engineers like, uh, but businesses want solutions and so recognizing the problem doesn't get them far.
Cool. Will will this, uh, fight club be the place it's announced or fight club forum or do you know yet? Yeah. Yeah. I'll probably dedicate a whole episode, uh, to the technology. Um, uh, cuz yeah, people will want to see what it's about and, uh, yeah. So, uh, it'll bring a lot of new opportunities and that's the cool thing.
That's awesome. Do you, uh, without committing yourself, do you have a general ETA? Are we talking days, weeks, months? Um, I'd say the the beta would probably, uh, close beta probably starting in a matter of weeks to one month. Um, and, uh, then you know, as soon as possible, uh, as long as you know the advantage is real and the program is stable. Uh, so that those are my two criteria. So, uh, if you know anything about Kora, I mean the whole idea was advantage. It wasn't pretty, uh, but it was effective. And so, you know, I've always felt that the product in our industry is the advantage. Uh, everything else is a nice to have. And so, same thing here, except it's, you know, it's got some workflow consideration for automation and it's got new things that people have never seen or done before. And so that's that's going to be fun to show.
All right. I'm looking forward to it. That's awesome. Yeah. Yeah. So, uh, it's all coming together nicely. Um, let's see. I'd, uh, like to mention that I figured out how to get that, uh, uh, eyebrow filter off. So, I got pranked, uh, uh, last week. So, uh, no funny eyebrow filter this episode. Uh, I was wondering if like you shaved them or something. I noticed that they were a little little less bushy than they were last week.
Less bushy. Uh, so yeah, if you missed that, that was last week's episode. If you need a good laugh, I didn't take it down or anything.
Um, all right. Uh, anything to announce? Uh, Terry or Charles? I was going to, um, extend kudos to you, um, as you know, we spoke I guess it was yesterday, wasn't it? Yeah. Yeah. About your new gig a support ticket but, um, I want to advise people don't sign up for support ticket. Let them work on this new stuff. So well, I mean if you're stuck, get support. Um, uh, but yeah. Yeah. The world's changing. You know it. I know it. I'm trying to do something about it. Innovation isn't easy. Uh, but it's coming.
All right. Um, I have a beta beta announcement, Ted. Um, or alpha, I don't even know what you'd call it. So, uh, for those out there that are interested in expired domains, I I have come up with a process. I'm thinking of developing a service to help people find them. If anybody wants to has domains they're looking for, types of domains they're looking for, hit me up. I will pretty much do it for free for you because I want to test out my process. Um, you agree to a case study that you can share. Exactly. Exactly. Um, and again, I say pretty much free because if 20 people hit me up, then I'm going to have to charge somebody something. I only have so much time, but but um, I want to make I want to make sure I I want to make sure I have a process that's solid before I ever start asking anybody for any money. I also want to make sure I'm develop, you know, providing something of use to folks. Um, so yeah, so anybody has any any need just just hit me up on Teams, Discord, you'll see my QR code. You can, uh, scan it during the during the deck today. And and, uh, again, this is like not even beta. It's more like alpha testing. I'm just trying to figure out if this works as well as I think it works.
Awesome. Yeah, let us know how it goes. Yeah, for sure. All right, with that, uh, we have a, uh, presentation from Charles and after that we'll do our Q&A. Remember to get your questions in there early. And, uh, take it away, Charles.
All right. Um, so this was a test I ran and, um, I was originally going to call this the I Blame TED test because this was inspired by you. Um, but, uh, I ended up finding a mystery and I wanted to share it with everybody cuz I think it's kind of cool. So, previously on Fight Club, I had shown a test and, um, during one of the log files, I noticed there was a 404 and and I realized, oh, wait a minute, I put a bad link in my canonical. Google crawled the canonical, found a bad link, and returned a 404. I'm like, "Okay, that's cool." And Ted mentioned, "Hey, we might have found an exploit for for getting things indexed." And so then I say, "Tested to the limit." I we we started talking. I think this was actually even after hours, we talked about a little bit. And this is where parts that I blamed Ted because I said, "Well, I can test canonical. What about more than one canonical? What about HF links? What about OG tags? What about real alternate links? What about schema?" And then I went through all the different places I could come up with links. And I'm sure I missed a whole bunch, but those are just the the batch I came up with in in 30 minutes. And I know I know you're crediting me, uh, but all the glory goes to the tester. So, you know, that's the way it goes. The pain goes to the tester, too. So, so yeah, I said if I'm going to test, I'm not just going to test one thing. Let me test a whole bunch of stuff and just see. I'm a big fan with testing. I know I know Ted in the past you've done a lot of like, uh, you know, will it index and and and what ranks better. I've always been very curious about what will Google do or what can Google do. Uh, that's always fascinating me even more than the zones mainly because the zones have been covered pretty well. Um, but anyway, so I'm always curious what will Google do. So here was the question like all right if I have I know Google loves crawling URLs on pages. It'll grab URLs you don't want it to grab. So, I said, "Well, let's just put out in a whole bunch of different places." And where will Google crawl? Um, the only thing I found is that you'll see under schema there's a URL field. You can't put five in there. You can only put one. I double tested it with the schema tool and with Google rich snippets and it only let me put one. So, then I just took it down to one and everything was fine. Um, so I submitted I so I made one page. Um, and I submitted it every place. GSC, XML, sitemap, HTML, homepage link. grabbed a bunch of indexers and and threw it out to a whole bunch of indexers. So, shout out to some of these indexers here. Uh, I I wanted there to be no chance that this page was not crawled. And then it's kind of the waiting game, just waiting for results. Oh, shoot. I showed too quickly. Oh well. I was going to make you guess, Ted and Terry, which ones you think Google crawled and didn't crawl. I don't know how fast I Let's go back to the the safe list. All right. So, I think Google crawled canonicals href lang. Think they crawled the real alternate. Um, I am skeptical that they crawled the schema links, but I'd be perfectly fine being wrong on that. And geez, I I know that Paul Har, uh, said they're using the open graph at SMX West like five, six years ago, but I don't know if they still are. Uh, you would get a visit if somebody shared it on a Google place that supported open graph. So, all right. Yeah, I'm going to say canonicals, href, langs, re alternate, and open graph. How about you, Terry? Do you have a I'd say the schema and the canonicals? The schema I know was just one URL, but um and the canonicals. Actually, let me ask you guys this. Are there What are the fields you don't think Google would crawl? like Google will never crawl for example real alternate links or will never crawl schema same as is there anything you think Google will not crawl well I I think that, uh, uh, schema is the oddball one because we don't know a lot about how Google's ecosystem handles schema because there are parts of schema that are for image processing and parts of schema are for relevancy Y and parts of schema are for, uh, you know, topical significance and disambiguation and so you know, is it that different systems grab the part of the schema they need and they're done is there one big schema processor and is that in the crawling loop, um, so I I have engineering doubts about schema, Uh, but again, I'm happy to be wrong on that.
Okay. So, uh, we waited and after so this was a test. This was a funny one where so every Sunday night I check my log files and in my log files I found this test page. I'm like what the heck's this test page? I'm like oh that's right. That's the test I started almost a month ago. Cuz this is a conversation we had at least 6 weeks ago. I totally forgot about this. So, I found this about a week ago, maybe a week and a half ago, after having forgotten about it for several weeks. This morning, I just recrolled the log files, and that's where the mystery comes in. But when I checked it about a week ago, I found and I looked at my, uh, Screaming Frog and I said, "Okay, what is Google bot crawled?" Or actually, I looked at what are all bots crawled. And by all bots, what it does looks at Google bot, Bingbot, uh, Duck.Go, Yahoo. There may be another two, one or two in the mix, but no AIs or anything like that. Googlebot was the only thing that did any crawling at all. I had zero love from being yahoo duck. Go. And the only thing that Googlebot crawled were three of the five href langs. It did also, there's one item that's not in this list. I put a control in there. I put an href lang. I put an href link in the page to say, okay, is Google going to crawl the href link? It did crawl the href link. So, three of the href I'm sorry, yeah, three of the href langs and then one just link in the content was crawled. Now, this blew me away because number one, the whole test started because Google crawled a bad canonical and number two, Google, to your point, Ted, Google has said they'll crawl OG tags. I've seen them crawl OG tags, especially for titles and descriptions. And last week we had a, uh, case of, uh, LLMs.ext being visited through a real alternate from Bing. Yes. And I and I was always surprised and honestly to this day Google has still not crawled that real alternate. Um, Googlebot has not crawled that real alternate. Um, but I know that that's how you get RSS feeds found. Or I should I should clarify that's how Google tells you RSS feeds should be found. So if you have an RSS feed for a category, you're supposed to put the RSS feed in the category page, uh, in the header. And then schema same as I've run multiple tests and Google seems to crawl same as. I know Clint has mentioned significant link and significant links multiple times. Um, so this kind of perplexed me. I'm like, this doesn't seem right. Um, and I have some takeaways in in at the very end here like why I think maybe some of these things happened, but this bothered me a little bit. I don't know about you guys, but this bothered me a little bit. So this morning when I was Did these show up in search? Well, they're 404 pages. So I didn't create an actual page at the other side. I just wanted to see will Google crawl. That's a good question. So yeah, there I never even wanted to try to get anything indexed or found in Google. I was just saying so I'm just looking purely at So that's a good question, Terry. So what I was doing is I put these I put one there's really only one page in this mix right now and was linking to a whole bunch of just fake pages and if Google crawls the link, I'll see a 404 hit those pages. Yeah. So I was just looking for crawling. But that's a that's a great question. So, I did I was doing some digging this morning as I was prepping up this deck because I was like, you know what? Let me let me throw something together real quick so we can have something to talk about. And while I was doing some digging, I found something very interesting. Everything was crawled by something. And by saying something, I mean a single something. So a single something crawled all these links, my href langs, my main entity thing, my mentions things, my ogs, my re alts, my same as links. There was actually too much for me to this was the biggest screen I could put in there without condensing it down so small. So there was something that crawled almost everything and there was lots of actually bot activity on all kinds of things, but there was one thing and that one thing was Google other. So, the first thing I've learned, so I've learned a So, this is a great example, Ted, and I know we've talked about this many times both on and off the show, how when you test, and I know Lee and I have talked about this, when you test, you often learn things you didn't expect to learn. My goal in this test is to find out what will Google crawl and what Google won't crawl. That's really all I cared about. What I've learned is that there's another bot floating around, this Google other. And at first, I wasn't sure if it was really Google. I'm like, this can't be Google. If you look at the, um, uh, some of the version information, I'm like, it seems right. So, I checked the IP and according to Google, this is a Google IP. So, maybe it is, maybe it isn't. But according to Google, the Google AI overview says this is a Google IP. I was like, huh? So, then I did a little little deeper digging and Google does give a vague definition of Google other. Um, and it's basically just says, "Oh, yeah, we sometimes crawl other things and other services are crawling." And it's really down to Google other is a generic crawler that may be used by various product teams for fetching publicly accessible content from sites. Yeah. Yeah. I'm, uh, reading about it here and it's, uh, pertaining to things not search and so some of the possibilities are like when people use functions to pull web pages into Google Sheets, Google Translate, uh, some advertising things, uh, internal systems. So maybe there's some sort of fact-finding thing for scoring, you know, trust and relevancy that might revisit and look at some things. Um, and they even mention, uh, the possibility of some AI tools. So, the the first my first big takeaway from this is if you're using something like Screaming Frog Log Analyzer, um, Scream Frog's Log Analyzer does not put this Google other in with the Google bots. So, by default, their Google bot, uh, um, uh, profile does not look for Google other. And then it made me think, how many other Google bots out there does are not covered by that profile? I bet a lot. So if you're looking for adsbot or imagebot or any of these other bots, the log file that by log file analyzer by default may not be picking these up. So that's just something to be aware of that you have to you can't just take the default settings. I I've I've actually grown a little more, um, warm. I've actually grown quite warm of log filing. It's a lot more flexible I thought it was because you can customize things. So where does this leave us? So, kind of like learnings, takeaways, and questions because after this morning, I was like, "Okay, I got all kinds of questions." So, number one, the dangers in testing, uh, coming to conclusions too quickly. I was ready to hassle Ted for suggesting this test and be like, "Ah, Google doesn't crawl hardly anything." And I would have been 100% wrong. Um, also running tests once. I have one page. This was one page test. This was not even a five-page test on the same domain. So, this test needs to be run at least four or five more times just to see if the same things happen. Um, something I've run into a little bit lately, I think is a problem is worn out test domains. I've been running tests in the same couple domains for a couple for years and years now, and I'm beginning to worry that I've worn out that domain or Google just doesn't seem to crawl the domain like it used to. So, uh, that's some of the dangers in testing that I've learned from this test. I want other people to take away from this test that there's there's problems with testings and you can you can learn things. Um, does Google have a special crawler for schema and OG tags? Are they not using Googlebot? Um, is this crawler being used for discovery? Maybe maybe this is like a discovery bot or they've they've offloaded discovery onto this other bot or it's used as a, um, uh, a supplementary crawler to to help discover things. Um, so this unfortunately at the end of this test I now have even more questions than when I started. Uh, which is the cool thing about testing. Like, okay, number one, I'd have to run this test at least three or four more times on different domains to see does the same thing happen. And I'm setting that up right now to see does the same thing happen over and over and over again. Does Google others show up? Does Google bot only crawl certain things? Um, and then we still have the mystery of if these same things happen or even if they don't, we still have the question of like what was this crawler? Why did this crawler do all of this? And you again, is Google using this for some kind of discovery? Is Google using this for for something else? Is it another process that Google's running that has nothing to do with discovery or rankings or indexing? Um, I don't know. I don't know. I'm left with a lot of questions now. Yeah, Google's, uh, search central documentation say it's a, uh, generic crawler that may be used by various product teams for fetching publicly accessible
Content from sites. For example, it may be used for one-off crawls for internal research and development. It has no effect on Google search or other products. Um, so yeah, if you're getting crawled a lot by that, it sounds like Google is taking, you know, special interest in your content internally, which is kind of I have to look I have to look at all my test sites and I haven't had a chance to do this. I want to look at all my test sites to see is Google other hitting everything. I don't know. You know what I bet it is? I bet it is internal applications and I bet your content uh has fallen into some sort of training set.
So, it's either in like a web spam training set, uh, is an example of spam and they're revisiting it. Uh, or it's in a collection of test pages that they're using against a training set. Um, and in either case, they'd be coming back to your pages regularly, uh, to see if their changes to those types of algorithms would actually, uh, uh make meaningful differences. So, and that kind of goes to this, you know, have I worn out my test domains? Because I've seen I don't want to say bad behavior, but different behavior. I've I've kind of come up with a new theory on how I can test and not run into this worn-out test domain problems, but I have to run some more tests to see if I could test, if that makes sense. Yeah. So, I have to run tests to see if I can test. And you have to, Charles, you have to test this worn-out domain theory also. Is it just that Google is not crawling like it used to anywhere, not just your domains that you've been using for testing? So the crawling has changed.
Yes, the worn-out test domains came to me because I was doing a I've mentioned this a couple of times. I'm doing I was doing a little black hat test and what I was surprised was by how many of my pages Google was crawling and indexing. I mean indexing like you could find them and then I just looked a couple of days ago it still has out of the thousand pages I thrown out there's still well over 300 pages in the index sitting in the index and and this test is almost a month old now. So, that made me start thinking, "Wow, that doesn't happen on all my test domains. I'm wondering if I've worn out my test." That's where I started thinking I've worn out my test domains. To your point, Ted, I they could be using me as some kind of uh using as using my test domain as a as a uh spam set. I have to go back and now look at all of my sites and see is this Googlebot other all over the place? And actually, now I need to segregate this Googlebot other in my log file analyzer. So I can analyze just Googlebot other and look at the Googlebot other on each of my domains over time and see has it increased, decreased, is it equal across all of them? What's going on? Yeah, I I bet it's not personal. So I would wager that no human being over at Google is aware of your site at all. I think uh you met some requirement to be included on mass into some sort of uh study or test.
Yeah. So there's some great takeaways with this test. I think it's a great So number one I always say like this the dangers in testing. You come out to the conclusions too easy. Running tests once are no good. I my platform my platform may just be bad. So I I may be running tests on a bad bad platform. Um, but I really should have added a a bullet in here that this, you know, the benefits of testing are I've learned all kinds of things that I did not expect to learn in this test. My goal was to see will Google crawl it. That was really all it was. Will Google crawl this URL? That's it. That's all I wanted to know. And what I ended up learning is a bunch of other stuff that's now I have to put into my queue of things to be testing. Very cool. But uh yeah, if anybody wants to uh if anybody needs any help with SEO, you can always reach out to me. If anybody wants to do some uh alpha testing on my on my service, reach out to me. You can, you know, contact me that bit.ly or the QR, just Slack, Discord, a million places. Um I'm out. I'm always around. I'm easy to find. And uh only the first few uh takers on his offer. So uh get that in quickly. Otherwise, you might have to wait till the launch. I I will either cut everybody off, which is more likely, or I'll just start saying, "You're going to have to give me something because this is my time I'm spending."
All right. So, uh I've been uh uh working on a lot of development in uh in the AI space as it pertains to internet marketing. And what I've come to to learn is that especially with like chat GPT, there is this portion of time that the ecosystem takes for thinking. And what I've found as an engineer is that the more it thinks, the greater the likelihood I'm not going to like the output. And so what I've what I've kind of found is that that that thinking time means that your prompt left uh important things out and it has to consider the possibilities of the things you left out and decide for you. And that's how you get, you know, inconsistent results that don't match what you want. And it's kind of like the bigger the prompt, the more thinking time there is. And uh uh you know, as a developer, you know, as a software engineer, I want things to be fast. So I want to minimize time usage and I also want things to be reliable. I want to consistently get the correct answer to things I'm asking for. Um, and so tuning the prompt for minimal thinking time is uh one of my new strategies. And it's completely changed my mind. Like all those massive prompts, I don't want them. Get them away. Uh, I find that I'm now breaking things down into micro prompts that are super fast returns, very low thinking time, and always deliver 100%. Um, and so yeah, it kind of has me rethinking a lot of my interactions with AI, and I thought I'd share that. Uh, Terry, you work with AI a lot. What have you seen?
I um I agree the prompts need to be more exact. They need to eliminate and say I don't want this, not just what you do want. But I tend to use the deep research, the deep whatever the deeper version is on the various uh AIS when I use them. Um and I often co-pilot is my go-to for forming a prompt. I'll tell it I want an exact prompt that'll do this, this, and this. I don't want anything but the prompt that I can use in an AI, you know, agent. And I I get what I like and with mid midsize prompts, not micro definitely, but not those huge 31 page prompts either. Yeah. I'll I'll give uh things like examples of the output I want. I'll actually send them uh you know an example of the response I want to get back and I'll tell them exactly what labels to use and uh it you know just the the less it has to think it's kind of like the faster and better it is for me. Um, and I find that uh a lot of the uh the cheaper uh models that you can use through the APIs uh are a lot faster and they do a better job or just as good of a job the more micro and specific you get cuz what they don't have is the advanced reasoning type stuff. Um but yeah, yeah, if you need quick, uh you know, got to tune that thinking time.
All right, only a couple of questions today. Um, let's see. Create and collab is uh uh giving us an LOL about the eyebrow filter I got pranked with last week. Um, yeah, that's pretty silly. Uh, somebody's asking who is the TEDK K of social media. There are, uh, two, uh, like on Facebook. Uh, one of which is me and the other one is a Chinese account claiming to be me. Uh, so the uh the one that has 5,000 friends is the me one. Um, is the fake Ted outdoing the real Ted? Uh, they they even use my profile picture, but like they don't have more than 5,000 people, do they? No, it's been inactive for a few years. And I don't use a lot of social media. I'm one of those old-fashioned grumpy people that isn't thrilled about social media. And the the reason I'm not so thrilled about social media is it's very polarizing. You know, people are uh putting up a front that isn't reality or they're uh taking a side on an issue, but it's an all-or-nothing thing and they're saying unfriend me if you disagree and and and I'm, you know, I'm not a bridge burner and I don't want to engage in bridge burning.
Um, so I just I found the rage bait is just out of control. It's like and it's it's the rage bait. It's it's all rage bait and it's it's it's all made for clicks and somebody posts something or they repost something and they don't realize that it's all about getting Ted to agree with me say I agree with you 100% and then getting Terry to hate me. Um cuz you'll both click on my thing. It's Yeah. No, it's an algorithm and you play into it when you when you click on that stuff. You you know I I tried it for a while and it just made me feel horrible and I I stopped. I said, you know what? I I shouldn't be doing this. I I should just let this be what it is and if somebody has something urgent for me, I'll respond to that. And other than that, I just try to pay no mind. And you know, I've been a lot happier and healthier since I did that. And so I made this rule. Uh, the rule is uh social media is for business only and reality is for friends and family. And so I like that rule. It works well for me. Um, you know, uh, if if you think I'm wrong, uh, tell me why in the comments. I'm open to debate on it.
Um, let's see. How do you guys use Kora for YouTube SEO? Um, while it is technically possible to do that, it's a real pain in the butt to do it. And so the the way you have to do it is you have to use custom search results and only some versions of Kora have that. Uh, which means you have to take the search results uh manually out of YouTube, put them in a spreadsheet, and then paste them into the right settings in Kora. And then it'll run the correlation of the factors on that result set. And technically, you could do that with any search engine. You could do it with Amazon, you could do it with Etsy, you could do it with eBay. Um, but uh it does require a lot of special configuration to get it to work well. So it's very advanced Kora usage. I know of maybe one maybe two people who have even tried to do it. Um, it's mainly there for SEO testers who would have uh no other way to get the data they need to see if there's an important correlation. And that's pretty much the only case that I would ever use for for using it. So very very difficult uh uh thing to do. Um, but my reinvention of Kora that I'm working on, I'm planning on it being a lot more multi-channel and so things like that. Uh, you know, there's going to be some multi-channelannel at the beginning. Uh, there may be uh more as time goes on. And so that's one of the other aspects of how the tool set is changing.
I'm going to Ted, if you don't mind, I'm going to jump in real quick for Juan's answer as well. So, one, I do a lot of uh YouTube SEO. A couple of things I would point out. Number one, um I've done a lot of testing with YouTube. It's amazingly simple. Um there's not a lot of on-page to do. It's all about engagement. Now, the big thing to remember with YouTube is if you rank for a keyword, that's great, but that's only about 30% of the traffic on YouTube is coming from keyword searches. I think it's like 60 or 70%, I don't know the exact number, but it's around those numbers come from the feeds, people's personalized feeds. So, it's all about engagement. It's all about getting people to click and watch your video and then Google will then show more of your stuff to those people and will also see the persona. So, Ted has a certain persona on YouTube. If he watches certain videos, you want to be associated and and associated with those other videos so that you can be shown to that persona. So, it's it's in many ways it reminds me a lot of how discover works where there's a very small amount of optimization. There is a little bit of optimization you got to do to get yourself in the game, but it's really an engagement uh factor. So, uh Juan, feel free to ping me if you need help with YouTube SEO. I'm happy to, you know, give you some tips, but um don't think of YouTube as the I'm going to rank number one and then all this traffic's going to come in like with Google because it's very very personalized. So, it's actually more like Google is YouTube's more YouTube has been like Google is now for years. Um, where it's very very personalized and you can crack in there, but you're going to have to get more than just a ranking. You're going to have to attract clicks and views.
And uh one piece of advice is do not run your YouTube channel like we run SEO Fight Club because not only do I not make uh you know compelling thumbnails for the videos, I've even stopped giving them titles. So uh just know that for us this is a uh labor of love. Uh it pays no bills. We're just here to answer questions and and to give back to the community. And somehow in spite of ourselves, we managed to get about 7,500 subscribers. Um so, uh any amount of effort should do better than what we're doing in time. And time is the other thing. If you quit, you never get there. So, uh the main thing that all YouTubers who succeeded did was they didn't quit. Um, and a lot of them have stories about how, you know, they got next to nothing for a long long time. Uh, and then one day it popped and then it was, you know, off to the races from there. So, you just got to kind of fake it till you make it and not give up. Um, the other thing is on the advertising revenue. I've talked with uh many people. I can tell you a channel like SEO Fight Club would do, you know, 100 150 bucks a month in ad revenue. It's kind of coffee money. Um, definitely not quit your job money. I talked to people who have, you know, 60,000 to a h 100,000 uh subscribers and they're making, you know, $700 to $1,000 a month in ad revenue. So, unless you have something else to offer, it's going to be very hard to make meaningful money uh with YouTube.
Yeah. I I've worked with a few influencers. To give them some advice and they've shown me their numbers and there's, you know, they'll get like a 100,000, 300,000, a million views on something. You're thinking, "Wow, that's great." And then when you look at the numbers, they made $5,000 on it. Like it's not a lot of money. It's like like, "Yeah, I'll take 5,000, but they put a lot of work into getting to that point." So, um, there's a lot of people out there on YouTube saying, "Oh, how to have a, you know, a YouTube money machine kind of that's clickbait." It's you're pumping out a lot of content to do that. Uh, and you still might not hit it. Yeah, but if you're if you're looking for clients and you found a, you know, clever way to to corner a bunch of interest in in those types of clients, then yeah, you know, if you sell your service, uh that'll make you more money than the ads will. Um and you actually might get more people considering your service if you turn the ads off. Uh, so that's the other thing is, you know, weigh it all out. And don't and don't don't forget the with a YouTube channel, a YouTube video can rank in Google search and get you more views than while your Google channel or your YouTube channel hasn't built up yet. So, it depends on whether your focus is to turn that YouTube channel into a profit machine for yourself or using it to funnel people to your money page or affiliate offer, whatever it is, because you might want to just focus on ranking it in Google, a particular video, not a channel. And definitely network. So, if you're making an SEO uh channel on YouTube, let me know. I'll help you promote it with my channel. Uh I'm, you know, pretty good about promoting things. I have a successful track record of getting people to about their first thousand subscribers. Uh then it kind of peters out. You kind of need a bigger influencer to get further. Um, but yeah, that uh reminds me, remember to go check out uh SEO this week. That's Clint Butler's channel. Uh the SEO video show. So that's Dre Divera's YouTube channel. Uh don't forget GBP Roast. That's Jordan Pierce's local SEO channel. Uh, and if you want to be promoted too, uh, just, uh, let me know your SEO channel, not any channel, and I'll promote that as well.
Um, all right, let's see. Uh, Wix doesn't allow customers to get into their HT access file. Could you share uh why they should actually allow that and what benefits could be? Well, the the problem is uh they have done a lot of clever engineering hacks to get a uh a million websites hosted off of a collection of Blade servers. And I would bet that the HT access, if they gave you just raw access to it, would probably let you do really crazy things with other people's websites and they're trying to prevent that. Um cuz there's a awful lot of edge cases that you can cover in an HT access. And so I bet uh there's something about their engineering setup where it's too risky or they can't cover all the bases or they tried and it blew up uh for their customers in a bad way so they don't do that anymore. Uh they should have some way to do some basic things. So, they should make a web interface for doing very specific predetermined allowable things. Um, and if they're not doing that, then, you know, eventually migrate off of Wix. You know, get your own VPS, learn how to admin. Um, so yeah, that's, you know, kind of your options there. I put a note in the chat, um, for Abigail that if if you're trying to block stuff, you can try using the robots.txt, assuming you can access that. It's not as effective as htac, bots can ignore it if they want, whereas they can't ignore htac, but it's a, you know, second option maybe. Yeah. One one thing you can do if you uh uh absolutely have to use Wix but you don't want Googlebot to uh engage with it uh use uh JavaScript on the page and uh set an interval timer to draw the content after about 5 seconds. Uh usually at that point, Googlebot will no longer be looking for changes on the page. And uh actually that might not work. Uh you'd have to use a 5-second loop of iteration. You'd have to count to 10 billion. No, that probably won't work anymore either.
Um, yeah, that's a hard one to solve. Uh the reason these things won't work is that Googlebot has its own custom version of JavaScript. And so one of the things Googlebot does is it skips pointless iteration. It accelerates time so the JavaScript intervals happen instantaneously. Um, it's crazy cuz like when you uh benchmark how long it takes to run some JavaScript and you get that indexed in Google and you print out the benchmark time as part of the title text that shows up in Google. Google has these amazing benchmarks. It's like they have this alien technology that can iterate to 50 trillion and 0.0, you know, milliseconds while on your computer it took, you know, 15 minutes. Um, so yeah. Yeah, that's that's a tricky one. I I think your only option is probably to change hosts or to convince them to give you an interface for what you need.
Um, let's see. Uh, let me rephrase my question. Uh, who is a genius just like Ted, but instead of SEO does social media? All right. Yeah, that's a different question than I answered. Um, who for social media? Uh, you know, I I know of one person, but they're not in the communities and I'm not really in contact with them. Um, and the way they do
The things is, you know, at least at the time I knew him, it predated AI, so I bet all of those things are different. He ran teams of content writers, and you probably wouldn't do it that way now. Um, yeah.
So, I think what you really want to do is look at what's working in like TikTok and Instagram and Facebook, and uh, yeah, and try to learn from what's working out there. But yeah, I don't know any scientist in it that has modern knowledge. How about you, Charles? Have you found someone?
No, I have not. I don't like looking at it from a scientific point of view. No, I've heard lots of people say, "Oh, here's how you get people to engage," and you know, kind of like the psychological side, which it's not scientific. I've met dozens of SEOs that know a trick or two. Yeah. Uh, but none of them would say they are the expert in the field. Um, let's see. So sorry on that one.
Uh, Ted getting copied in China means he's internet famous. No. No. I probably pissed someone off, and/or somebody wanted to take a shot at me or something. Um, it seems more petty than anything else. And like I said, that account's been dormant for years. So, the only problem is people try to engage it thinking it's me. So, if that was you, just know I'm the other one.
Um, all right. So, that's all the questions. I'll give you about 30 seconds on any more questions. If not, we're going to end early, and I'm going to get back to coding. Uh, so, thank you again for watching the show. Thank you, Charles, for another great presentation. And thank you, Terry, for being our co-host for today since Lee was out. He had a parade. Um, haven't seen Clint in a while. He's probably busy. Uh, so, you know, if you need your fix of Clint, uh, go see SEO this week. All right. Thank you everyone.