A (0:00)

If you're a small business owner, you already know what it takes to keep everything moving. You're juggling customers, invoices and about a hundred decisions every day. Thankfully, taxes don't have to be one more thing on that list. With Intuit TurboTax, you can get your business taxes done for you with a full service expert. TurboTax matches you with your dedicated tax expert who knows your industry, understands your business write offs, and gives you the personalized advice your business deserves. Upload your documents right in the app, hand everything off and still feel like you're in the loop the whole way through. You can even get real time updates on your expert's progress right in the app, which makes it so much easier to stay on track. And you can get unlimited expert help at no extra cost, even on nights and weekends during tax season. Visit turbotax.com Ticket get matched with an expert today Only available with TurboTax full service experts.

B (1:06)

Hello and welcome to SEO 101 on WMR FM. My name is Scott Vanac. I am the Senior SEO at Step Forth Web Marketing. Ross is away today, so you've got me all by myself for a solo show today. This is my second take at this episode because for whatever reason I did an entire show and it did not record. Oh how wonderful is that? So maybe this one will be even better. I don't know. Same topics but my second go through. Not the first time that's happened. Love it when technology fails me. But that's okay. We've got a little bit of non SEO news to get started and this is with OpenAI moving on paid ads within ChatGPT. We don't talk a lot about paid ads, but this is, this is relevant. I think. OpenAI is preparing to launch ads inside of ChatGPT as early as February. They are asking some advertisers for some big commitments, including a small pool of advertisers where they're asking for about a million dollars in spending commitments each over the period of the next several weeks as a trial period. Ads are expected again to start appearing in February, but ChatGPT does not yet have the technology for advertisers to easily buy ads, so everything is kind of done manually. Not sure exactly how that works. Someone over at OpenAI is just manually doing the work I guess. Pricing details have not been revealed, so we're not really sure what people are going to be paying at least to start, but it will be based on a pay per view impression based model and not click based. Glenn Gabe posted To X pay per view. Oh my God. What does that tell you about clicks from ChatGPT? I would expect not a lot of clicks per impressions. Like it's usually done by per thousand impressions or something like that. If the click through rates are low, as long as the cost, you know, accounts for that and they're not paying a crazy amount per click at the end of the day, maybe it'll work out, maybe it won't. It's really hard to say. But yeah, like are they not expecting a lot of clicks so they want to build by impressions? Kind of sounds like it, but I guess time will tell. Hopefully this becomes a good means of advertising for you people out there listening and for our clients and anyone else out there, it's always good to have a little bit of competition and maybe this will help get some more traffic. Okay. There's another study that shows that organic search Traffic is down 2.5% year over year. So a company called Graphite, using data from similar web, looked at data from more than 40,000 of the largest US based websites. The study compared organic traffic from February to December of 2024 against January to November of 2025. They saw that on average year over year Traffic was down 2.5% which is far below the often cited 25 to 60% range that we see in a lot of studies and articles. I do have a bit of concern with their date range though, because it was February to December compared to January to November. December is one of those months. You know, we've got the seasonal aspects of Christmas and all that taken into effect and you, you aren't comparing to the current December. So I, I think the data that they're looking at is, is going to be a little bit off here. I don't know, I guess it depends on the industries that they're looking at. But anyone who's seasonal knows December is either a big vague month or a quiet month. So we'll see. A Google statement back in August does support this 2 1/2% reduction saying Google had said in August that click volume was relatively stable year over year, so could be true. Of course results will vary significantly here. Sites that are highly dependent on informational intent queries are the hardest hit with AI overviews. AI overviews is displayed for. I can't remember what the figure was, but it was something like 80% of all informational intent searches. It's displayed much lower for commercial and transactional and so on. So if you are, if your website is very focused on informational intent based content, you're probably seeing a lot more than 2 1/2% in reduced click traffic from Google. If your site is more commercial or transactional, you probably haven't been hit quite as badly. So I don't know if this site looked primarily at sites that are not informational because then that would make a bit more sense to me. But interesting to see a study that shows kind of what Google's been saying and how click through rates haven't been affected as much as a lot of people say they have. Up next, a study. Well, there was a report by the Guardian that discussed health related topics and AI overviews. So they reported recently. The Guardian reported recently that Google AI overviews sometimes provide incorrect or dangerous health advice. Some examples include flawed guidance on pancreatic cancer diets, misleading explanations of liver blood tests results, and one I saw last year and we talked about it on a previous episode where I can't recall the exact search that was made in Google. But it effectively AI Overview's response recommended that putting crushed porcelain in breast milk and feeding it to your baby is a beneficial source of extra calcium. Now hopefully anyone listening knows that obviously that's ridiculous. It's not only a good source of extra calcium, it's a good way to kill your baby. Probably. Like I don't really know what the effects of eating crushed porcelain would be, but I can promise you it would be not good. Thankfully I did try to reproduce that search result today, just out of curiosity. I couldn't get AI overviews to talk about it in any way. It never appeared for anything. So either Google has fixed it because that one story did get a bit of press. It was all over search engine land and SEO community. A lot of people knew about it, we talked about it. So I don't know if Google manually fixed that one or if the algorithm just naturally has improved. I mean it has improved with certainty, but has it improved to the point where it's not showing that bad of a result? I don't know. But it is important to also note that with AI overviews you really, really? I say this all the time. Maybe I should stop saying it. But you've got to be careful. Don't take it word for word. If AI overviews recommend something, make sure that it's correct. Especially when it comes to health oriented stuff, it does get stuff wrong. I see AI overviews being displaying incorrect information almost daily. So you know it's great when it's correct. But do you know if it's correct or not? I don't know, so be careful. Google did dispute the findings from the Guardian. They did say that most AI overviews are accurate and linked to reputable sources. Again, my experience is maybe not use of the word most. Many. I would say many are good, but definitely not most. The Guardian also examined where AI overviews were getting their information from. The study reviewed more than 50,000 health related searches or citations that you would have seen in a overviews or, sorry, 50,000 health related searches in Germany and found that 2/3 of the citations used came from sources without strong medical or evidence based safeguards. They also found that YouTube was the single most cited source. No surprise there, at more than 4%. YouTube has cited more than reputable sources such as hospitals, clinics, health associations, really the authoritative places you'd want to see cited in any kind of health oriented AI overview. Additionally, they also found that only 36% of the cited sources appeared in the top 10 of regular organic results. I find that kind of surprising actually because Google tends to like I tend to see the top, you know, three to five organic appearance citations quite often, but not with this study, not with health oriented stuff. So there we have that and let's move on to our next topic. Does LLMs txt Matter? We've talked a lot about this in bits and pieces over the past few episodes. We recently reported that Google had placed LLM LLM. I just can't say it. I don't know why LLMs txt on a few of their websites. John Mueller said that it was not intentional, it was related to CMS updates that were made. John also repeated that the file does not matter. Well, Anna Fernandez and her team from Pre Visible did a small study. This is really low scale, but it's something where they analyzed LLMs txt from 10 websites. Of those websites they saw that AI traffic increased from two of those sites, but that it was not because of the file. They tracked the 10 websites for 90 days, added the text file, tracked for another 90 days. They measured AI crawl frequency from ChatGPT, Claude, Perplexity and Gemini. With the results they saw that two sites saw an AI traffic increase, one at 12 and a half percent, another at about 25% but state that the LLMs txt file was not the cause. Seven of the sites saw no measurable changes and one site declined by 19.7%. For the two sites that saw growth. Those could both be attributed to other work done by those two sites, including new content that was developed, some technical SEO issues that were fixed, and other things for the site that dropped they do not believe that was related to usage of LLMs txt. They didn't really go into why they believe that, but they don't believe it is. So basically at this point, no major LLM providers are officially committed to parsing these files. This kind of goes with everything that we've been saying, with what Google's been saying. They're really kind of a nothing right now. But there may be some outliers that are using them. You know, it's probably similar to when other things are released like XML sitemaps. Not everybody used them right away. You know, just stuff takes a while to be adapted. So my suggestion is implement them if you've got a moment. It's not critical. I see zero chance that having an LLMs txt file will hurt you in any way. I also don't think it's going to help in the short term, long term, big picture down the road, maybe they start using them and you've already got it in place and you're, you're good to go. You don't have to think about it, you don't have to adapt, you don't have to react to the changes. It's just already there. So low priority item implemented. If you feel the need, I, I would suggest it, but don't stress it. And now we are going to take a quick break. When we come back, we'll talk a little bit about keyword search usage within Chat GPT, the importance of core web vitals within ChatGPT or within AI and Google search console impressions and a couple other things. So stay tuned. All right, welcome back to SEO101 on WMR FM today hosted by myself, all alone, Scott Van Eck, Senior SEO at Step 4th Web Marketing. I've got a problem. In my notes it says CEO. I'm not the CEO. I need to change that. I almost read it three times now. I think Senior SEO at Step Worth Web Marketing, please remember we do have a show notes newsletter you can sign up for@seo101radio.com don't miss a single link. Refresh your memory of a past show at any time. In addition, we do invite listeners to connect with Ross on LinkedIn. You can connect with him@LinkedIn.com in website marketer. Do a search for his name. I'm sure he'll come up there as well. You could probably go to our website stepworth.com and you will find links to his social accounts there. Also, before the break, we were discussing a bit about paid ads on ChatGPT click through rate declines, AI overviews. And now we're going to go into 75% of ChatGPT users rely on keywords for local services. So there's a lot, I guess there's a big push where people are saying SEOs are saying we need to move away from keyword based SEO. You know, we're not really focusing on a single keyword for a single page. It's more about topics and that sort of thing. And while that's not totally incorrect, the keyword per se is not dying. This, this study helps illustrate that I believe it. I think most people probably still believe people search keyword based a lot and this data kind of backs that up. When it comes to using tools like Chat GPT, the assumption is often that keyword style searches are really on the way out. Most people use longer conversational prompts. I tend to be different. I don't like to speak to my phone necessarily. I'm still typing things out like a, I don't know, like an old school type person. I don't know. So keeping it short makes sense. I tend to keep things shorter most of the time. And to test this, Frank Olivo from Sega Pixel spearheaded a test where they observed participants using ChatGPT to find local service providers, primarily in the healthcare and aesthetics fields. Here's what they found. They found that 69% of users still searched using keyword based queries, 25% searched using natural language, and 6% used kind of a hybrid of the two. So it really does show that keyword based searches are alive within your favorite LLM and should not be ignored. Yet I always, you know, say, you know, don't forget about these keyword based searches and including keywords in your content where it makes sense to do so, all that kind of stuff. But conversational prompts are still growing. They're big, they're important, you know, FAQ based type content where you, you're posting the question as somebody would ask it and then answering it. All that kind of stuff is really good for the future. It's still very useful now, but it may not be as big as keyword based still at this point at least. So keyword is still or keyword phrases that are simple are still delivering a lot of traffic. So they shouldn't be ignored yet. All right, here's what 107,000 pages reveal about core web vitals and AI search. So Google uses core web vitals as a partial signal into whether a page provides a strong user experience. And user experience is a big part of Google's algorithm. Since AI also looks for a positive user experience, it does stand to reason that they too should look at performance metrics like Core Web Vitals. A study of 107,352 AI visible web pages by Dan Taylor, head of SEO at Salt Agency in the UK, found the relationship between core web vitals and AI performance to be real but limited. They found no strong positive correlation. However, they did find a measurable negative correlation in extreme situations. So if you have a website and the performance was really, really bad, really low Core Web Vital scores that could have a negative impact on your performance in AI. However, most sites now have decent performance and if you have sort of an average performance compared to someone who has exceptional performance, there's not really any difference there in how your ranking is within AI results. So basically if your rankings within Core Web Vitals, your speed test scores, that sort of thing are terrible, improve them, improve if they're average, you don't really have to worry about it if AI search is the goal of this performance increase the article over at Search Engine Land the link is in our show notes, so definitely check it out if you want to learn more. It does go much further in depth and dives into various core Web Vital data points like Cumulative Layout Shift and Largest Contentful Paint and has a bunch of fancy charts and graphs that show everything and and if you're interested in knowing a bit more about this, how they made their determinations based on everything, the article goes way further in depth than we are going to cover on SEO 101. All right, and finally we have a Mueller file and it's about URLs in search results and the impressions they create within Google Search Console. Mark Williams Cook, the director of SEO at or the director at the SEO agency candor, shared on LinkedIn a discussion around understanding how AI overviews affects visibility metrics When a page appears multiple times. John Mueller confirmed that the same URL in both AI overviews and in regular search and this is at the same time, same search, same screen, Google Search Console Console will count this as a single impression, not two. And so to extend this a little bit further, the same holds true if you scroll down away from the result and scroll back to the link if you encounter the link in other multiple elements on the page. Perhaps people also ask for featured snippet or a local result or something like that. Your result will still only count as one impression in Search Console, something I think a lot of us knew or at least suspected. But there you have there you have it directly from Google once and that's it. Well, that was a bit of a shorter show than normal because I'm on my own today and I tend to talk too fast. Ross slows us down a bit, which is good. He yeah, he's maybe a little bit better at time management when doing the show, but I'm trying to get better at this. If you have any feedback, what you think, what you think I should change, please go to our Facebook group and let me know. I'm open to suggestions. I do listen to reviews and what people have to say. So anyways, on behalf of myself, Scovanek, Senior SEO at Stepforth Web Marketing, thank you for joining me today again. Remember, we do have that Show Notes newsletter. You can sign up for it seo101radio.com don't miss a single link. Refresh your memory of a past show at any time and if you have any questions you want to share with me, thoughts, comments, please feel free to post them on our Facebook group. Easily found by searching SEO101podcast on Facebook. If you enjoy the show, we appreciate any feedback on Apple Podcast Stitcher or your favorite podcast stream. We try to be everywhere and we do listen to what you have to say. Have a great week. Remember to tune into future episodes, which are every week on WMR fm. Thank you for listening everybody.