Outsmart Google’s AI Shift: See Every Ranking Jump from Keyword to AI Overview with SE Ranking
As Google's AI overviews fundamentally change how search results display, traditional SEO rankings no longer guarantee traffic.
Matt Bailey draws lessons from past algorithm shifts—from the early 2000s multi-engine landscape to Panda, Penguin, and Hummingbird—to show how the core principles of user-first content, topical authority, and intent satisfaction remain timeless.
Using a real client case study, he demonstrates how AI overviews can eliminate clicks despite top-10 rankings, forcing a strategic pivot toward informational searches and owned channels.
The key to winning in the AI era is understanding customer intent through direct conversation, structuring content into scannable chunks for both AI and human readers, and building sustainable visibility beyond search rankings through email lists and direct audience engagement.
[Music] All right, let me pop my screen on here. So, yeah, what we're going to look at here is the the ranking jumps between SEO and AI. I'm also going to talk to you about uh a client I took on who was having some trouble and I was working with them in the middle of this AI overview shift and just crazy stuff is happening. But let me just you know Rejoice asked me to introduce myself. So yeah, I've been in this industry for like ever. Uh I have worked with in an incredible amount of companies. I've done training at a number of companies. Now, I do a lot of teaching and training. I don't take on a lot of clients, but I take on the ones that have a good story. Uh, and this was one of them. But here's how I'm going to
approach this talk. That what's old is new and what's new is old. And I say that because of what I mentioned earlier that it is just like being in the early 2000s. Right now, there is so much happening with AI overviews and trying to figure out, well, how's it going to show up in AI? How do I increase my visibility in AI? But I also want to keep my rankings. And so, there's all this information swirling around. I've got to tell you, it's just like it was in the early 2000s because back then, we had five search engines or more that we were trying to track rankings across. Google wasn't dominant at that point, but it was becoming dominant. And so you're trying to figure out how to rank in Alta Vista, in Yahoo, in Ask Jeeves, in Google, and you're trying to play all these off each other, but also learning, you know, how do I do this? What's working? The key here that I would say
back in the early 2000s, the key to learning and applying information was community. Back in the early 2000s, the vast majority of SEOs were on forums and sharing information that this is what I've been doing. This is what I'm seeing. They were sharing screenshots, debating. there was a full-on community that would debate and evaluate and you could get feedback from on what you were doing. And so a lot of shared information helped people get through that. Uh now I will say right now what we have is better tools. Back in the early 2000s, I mean it was Excel spreadsheets, it was basic analytics. you were just trying to work your way through and and by the way most of your rank checking services were illegal or not sanctioned by Google and so you had a very limited amount of information but now we have fantastic tools like SE
ranking. Now, one of the things I learned in those early days is sometimes a lot of information is shared and it becomes assumed. And back around 2008, there was a lot of assumed information about something called Google's sandbox. And what the sandbox was is new websites were supposedly put into sandbox and they weren't allowed to rank until they essentially kind of proved themselves whether it was through links or getting visits something like that. That was the concept of the sandbox. And a lot of SEOs were promoting this idea that new websites just aren't going to perform well because Google puts these sites into the sandbox. In my agency, one of the things I had every employee do was build a website. And then once they built that website, they had to make money with it. That was the second thing
they had to do. I had an employee and she was in the equestrian industry. She rode horses. She taught people how to ride horses. She competed. And so, of course, she built a website about horses, about equestrian events. And what happened in 2008, she built her website and about 6 months later, which is this ideal sandbox time, the Beijing Olympics, the summer Olympics came on. and she was looking to see when the equestrian events were going to be televised in our country and they were nowhere to be found. So she contacted the Olympic Committee. She she made call after call after call to find out when will the equestrian events be televised. She finally found someone. They gave her the answer and she took that information and put that on her website. Within days,
her traffic starts to increase. And we're about 3 months away from the Olympics at this point. And you could just see the amount of traffic increase over time. Her rankings increased. And we're looking at this thinking, you know, all this information about the sandbox. This is blowing it to pieces. By the time the Olympics came around now, her site has only been live for about 9 months at this point. She's a 20,000 visitors a day and she's monetizing with ads on her site. So, it was fantastic. But what this taught us, if you have information that people need, you will be found. And that's the bottom line when it comes to search and and AI. If you have the information, if you are the authority and you have the content people want and need and you've got it all, they will find you because it's in
the search engine's best interest to give people the content that they want. So regardless of the age of the website, this threw the sandbox theory out, it was about information. It was about are you giving people unique information and are you the source of that? Now, I want to talk to you about this client that we we took on. They had a story, you know, and and and you know the story. They've been through 10 companies, lots of promises, lots of activity, no no results. So, we took them on and we started optimizing for them. We got, you know, more than 200 terms. can see here went from almost nothing to getting into multiple rankings, getting into the first page. We're excited. This is happening early spring this year. This is probably about March, April, May. We're starting to see the rankings go up. We're starting to see traffic go up. And then all of a sudden, Google AI
overviews enters the room. And as you can see here, and this is right out of their search console, uh you can see the impressions in purple going up and the clicks in blue going down, which was exactly the opposite of what the industry does every year. Clicks should be going up. It's a it's a uh a trending content and it's a trending site. Summer is always their biggest season and that's when most people are searching for their products. But instead, we saw a tail off on our keywords, on our clicks, but the impressions were going up. Getting into our rankings here to find out what's going on. And for the primary keywords, going for those, we've got first page rankings. We've got rankings in in the top 10. We've got a few in in you know, but what happened is the AI overview. And so for the terms
that we were going after here, and these were going to be the focus terms for the first phase, and then we were going to grow from there. But all of a sudden, what we're seeing is the terms that were bringing in visits all of a sudden stopped. Why? Because of the AI overview. Look at this AI overview page. It's like three scrolls long. I don't care if you rank number one or number two. No one's going to look for you because everything is here. Google AI overview took all the information from all the ranking sites and loaded it into this. Essentially, what Google created was a landing page and we've got citations through here. So, now we've got to reassess. What are we doing now? Because all these great rankings are essentially useless. No one's clicking through. So we went through and and what we started to do was look at what terms are not triggering AI overviews. And what's interesting is when we started
getting into some short tail terms, what we saw, so for a couple of these large short tail terms, check it out. We have images. The images are coming from blogs. They're coming from Pinterest. They're coming from articles. Then we have more articles. Then we have Reddit. Oh yeah, there's Pinterest videos. And the videos are coming from Tik Tok. They're coming from YouTube. Uh they're coming from uh Facebook. And then actually we get into some e-commerce stuff here. And then even further below that, more articles. So what it showed us is yeah, there are some areas and and so now we're readjusting strategy in order and and what it's forcing us to do for this specific client is to go further up the funnel. We're trying to get more intoformational, but also finding thoseformational searches where
we're not competing against Google's AI because at this point, people at that top of the funnel, they're searchingformationally. They're they're at the very beginning of the cycle. It wouldn't make sense for Google to give an AI overview at this point because what people want to see are pictures, videos, lists. That's what people want to see there. And so finding those areas where we can readjust off of that and pivot into there's no competition from Google. It's going to take a while, but we're going to go after some of the big terms. Now, also what it showed us too is how to pivot with the AI. I showed you that big long page of Google's AI overview as well as you could see, let me jump back into there. There's citations all over it. all kinds of citations here. And so that's where we want to be found. And so using SE ranking, we were getting into some of the words. Where are we showing
up? I love this to find out the organic to AI overlap. That is looking at your organic results and how many of those organic results are also showing up in AI citations. So this can give you a good sense of who you're competing against and how many sites. And also what we were able to do in readjusting the content, readjusting how we were presenting content and and applying that. Uh you can see we increased our link visibility uh from 15 up to five. And so that's now where we can start looking at how some of this will work. Now here's the issue with some of this and and you can see here we've got a lot of similar type of words. The problem with some of this is that you can ask AI the same question maybe an hour later, maybe a day later and you'll get different information. I I see this all the time. I especially
see it with my college students. They think that they can just type it in and you know sometimes that they'll get the same information but depending upon and and Rejoice did a great job talking about how you prompt and if you're using just general prompts general questions the results are going to change. So, while I like this as a a kind of an overview kind of a look, you know, today it's alternative to wood deck today, you know, and and tomorrow it might be substitute. It depends on how the prompt is formed and that could affect your visibility. It could also affect just based on numerous other factors. But just giving a sense of, hey, I'm showing up. I could be very happy about that. but also uh looking at SE rankings. I'm I'm we're we're doing the beta on SE visible and this is the visibility in AI which I absolutely love. So now I can look with specific prompts in the
industry. What are people looking for? How where's the visibility and where's the average position and the competitors? It's exactly what Rejoice was talking about is go look at the competitors that they look at their visibility. what are they doing and so we can incorporate that as well. So this is going to be a fun fun product to help enhance what you're looking for in AI, what your visibility is, who you're competing against for very specific prompts that people are searching for. Now I want to now what do we do? How do we respond to this? and buy this. Here's the best advice I can give you. And and like I said, today reminds me of the early 2000s. So, what happened from there to here? If you look at the past, it'll help you predict the future. Here's what I mean. In my university courses, I make students do some research. research all the Google
algorithm changes and tell me who got hurt and who was rewarded and very simply explain these things and and so we can look at the evolution of search and algorithm technology and so here's what I want to present to you and stop me if this sounds familiar so back in 2011 Google panda maybe some of you remember that what did we learn from Google panda build original comprehensive of user first content. Does that sound familiar? In 2012, Penguin that taught us earn editorial and relevant back links, avoid the link schemes. Okay, so now that that focuses us on PR, it focuses us on developing relationships with other higher authority websites to get citations back to our site. 2013, Google Hummingbird. This is the one and this was probably one of the
first major algorithms to focus on content. And what did it focus on? Content that answers questions rather than matching keywords. This is hilarious because I see this this quote this this this finding build content that answers searcher questions rather than matching keywords. I'm seeing this headline articles that were written yesterday, that were written last month. It's like we've known this since this is what we learned from Hummingbird. So, and and this is what we learned and and so this should have been preparing us for what's going to happen rather than just matching keywords. Look at this is what came out of Hingberg. Evaluate pages for answers, not just facts. So going beyond that keyword match and giving a comprehensive topical overview of your content synonyms natural content that reads
conversationally conversational queries that's exactly what we're dealing with with chat GPT with claude with Google AI overview. This is where it started to really come in and what we learned from Hummingbird still relevant and it's still part of the advice you're getting today. RankBrain 2015, what did we learn? Google is going to reward topical depth and relevance. None of this is new. This has all been coming at us over the past 25 years. Google BERT, this was AI language analysis. What did it reward? It rewards writing for humans and using contextual relevance. So when you're building your pages, you're building out your topics. You're associating them. You're develop you're linking them together to build that topical authority in great well-written content that drives people to a result.
So and and this is what we've dealt with just in the past couple years with the helpful content. What are we learning from the helpful content update? Original, insightful, authoritative content, real expertise, unique value, and intent satisfaction. Now, that right there means a whole lot. Intent satisfaction. That means I got my answer and more. How do you give more than just the answer? How do you make someone give how do you give them the answer to what they're searching for and give them that satisfaction? Well, here's a couple of things and what I have done is I I'm now I'm not calling it SEO. I'm not calling it GEO. I'm not calling it is persuasive content development. This is what we're doing. Create content. And here's what you do. Number one, understand the intent of your searcher. And by that I mean if you never talk to your customer,
you will never understand their intent. Get out from your computer, make a phone call, visit with them, do one- on-one talking with your customers to learn how you solve their problem. Learn how they perceive you. Learn what would they do if they didn't have your product or your company. Understand their intent. What made them search and then how did they find you? These are things that can only be found through research for interaction with your customers. The better you understand your customer, the better you will understand intent. You'll never get it from just staring at a a screen and using search tools. You have to talk to your customers because now you'll understand the questions that they have. You'll understand those questions at a much deeper level. You will be able to have more topical dominance because you're getting your data from actual
humans. And then when they come to the page and this is how you get intent satisfaction. How will you engage their attention? You can engage their attention with videos, with downloads, with quizzes, with interactive elements like calculators, things that cause people to stay on the page longer, to click around, to investigate. This is where you're engaging their attention. And there's a great question about, you know, how do I move them through that buying cycle? You can get them on the top of the funnel. You can get them at that information search if they come to your site. But then also what you're doing is directing progression. Directing progression in that customer journey from just getting the basic information to, oh, there's something I can download. Hey, great. A video, a time-lapse video that shows how it
works. very quick bites that help them decide and directs progress call to action to a form to any lead generation device that you've got on there. But you just can't serve up the content without a context of what where to go and what you want them to do. Contentwise, AI is going to look at chunks of your page. SEO is going to look at the entire page. So, think about that. As you design your page, you build a page with topical expertise, but you're also breaking the content up into subheadings. This is basic SEO. It It's And as I said, I came out of journalism. This is how content is built. headings, subheadings, bullet points, breaking content up into easy to scan and read pathways for people to get through the page. And then also any authoritative
references. If you're citing studies, research, quotes, facts, cite it because that helps show that you're not just making this stuff up. You're also closing the loop by linking to people who you should link to. That has to show up as well. Now, just to give you a look at, you know, even I took my own medicine. All right, I redid my entire website, which hasn't been done in 10 years because now with AI and also with SEO, it's how do you balance these two? And so, just an example of how you you know, you chunk the information into very specific things. You've got your headings, but then also short paragraphs, easily scannable information. So, I've got, you know, your your benefit statements here, but then also getting into who are perfect students. And everything is its own encapsulized information, but it's part of a greater whole. And so, when you're
developing these pages, easily scanned, easily structured, understandable, how can people find the answer to their question? You'll notice I'm doing a lot more question and answer as I develop these pages. Now, the cool thing is I saw rankings increase. So, for this course, for this OMCA, it's a a digital marketing certification. And my rankings have gone up significantly from redesigning and redeveloping the site. Not only that, you know, finally, you know, I love seeing that, you know, 40. Uh, so if anyone is looking for information on Reddit about the certification, uh, went from 41 to 40. I love that. This is, you know, this is what makes an SEO heart warm is seeing those increased rankings. But now also getting into the AI overviews, the difference between before the launch and after the launch, seeing that visibility uh, show up. So now you can develop your
content for both SEO and AI overviews because a lot of these things are based on what we've learned in the past and it's how it's applied today. And so I love that you know with the SE ranking we can track these things. We can look at the overlap and it gives you a sense of where to focus in order to increase your visibility in those AI overviews. Um and then also just to give you a sense also the SE vvisible here as well. Um hope that comes public soon. Um but protect your brand and that's what I love about the SE vvisible product here is it enables you to look and see if people are searching based on your brand name. What's your visibility? Where's your position? What's the sentiment? And so fantastic way to protect your brand and see is it positive in the AI overviews how are you who are you competing against what's your position
so it gives you very specific areas that if you want to increase if you want to develop something from here great thing I saw is you know probably you know my lowest one here is uniqueness as well as uh success stories gives me great ideas for additional content content to boost that position, but also it's additional topic support that you may not have thought of, may not have developed from there. So, here is the summary. If you're going to optimize for both SEO and AI results, research your topics and remember what I said, talk to your customers. They're the best source of information. Focus on intent and providing comprehensive answers. Design your pages in subtopical chunks. Remember, SEO is going to look at the entire page. AI is going to look at a chunk. So, make sure that you have independent chunks of information in
context with the overall theme and and and topic that you're writing on. Cite all of your information, but keep the visitor on the page as long as possible. downloads, videos, CDAs, anything like that. And also that progression path. Where do you want them to go after thatformational view? Do you have things that will move them to that next stage and continue inviting them to take that next step? And then also what this is, all of this has learned, and you've probably heard this, go beyond the rankings. Don't just focus on AI overviews. Don't just focus on rankings. Now's the time to develop a lot of other channels that maybe you've not been developing before. Uh I I and many others have been hammering the email lists and that has been right now probably one of the better sources of engaging existing customer or repeat
business for ongoing. And so it's building and developing better email structures, better email communications, and getting that inquiring because here's the thing, it's direct. It's my audience. It's it's my email. So everything I do in email is a direct communication. So many other things are indirect. So anything you can do to build your own audience and develop it from there. I am amazed at how many companies right now in order to get a discount, you have to give both your email and your phone number in order to get the discount. And so now I get SMS texts all the time from brands. Um my favorite one here is I bought a shower head and I'm getting texts every day about buying another shower head. I don't need another shower head, but they are directly communicating with their entire audience. and what I can't think of a better way to do that because now it's going to drill in my head that
brand and if I need another one I'll go through them. So thank you so much for attending this session and looking forward to your questions. I would love it if you have any questions you can connect with me on LinkedIn. Uh also have a podcast called the endless coffee cup which you can find on the website at sitelogic.com. Well, Matt, that was amazing. That was fantastic. A few of the comments were sort of really um loving your voice and I have to sort of agree. It was, you know, it's so good you don't do commercials. I would have been buying anything anything you sold trying. So, we do have a couple of questions. We do. So, the first question is from James Goodwin. So, James has asked, "How do you split brand versus non-brand and risk mapped keywords by AIO exposure to decide where to double down, hold or exit? Also, what review cadence actually works in practice?
>> We don't know because it could change next week. Um, that's I I at this point my best advice is I I am committing to things about 70%." because it's so fluid. This is what I mean. It's like this is reminds me of back in the day. I'll listen to 20 people's advice. I'll decide what to do and I'll probably only go about 7 8% and then I watch that. That's the best advice I can give you. Um now brand versus non-brand. Um, you're and that's the thing I So, I'm looking at brand where I know my brand will be searched and so that I focus on. If my brand shows up for non-brand, that's awesome. But again, I can't depend on AI to be consistent in how they answer that because it could change
next week when they start using Bing instead of Google or start favoring Reddit less. So that's what I mean. You commit about 70% and watch. There's nothing you can do right now that's going to be 100%. And there's no one who knows what's going to happen next month. >> No, that's No, that's very true. I think how AI changes a lot faster than traditional um SEO. So it's very hard to pinpoint you know that >> yeah it was only two months ago that uh chat GPT was using Bing and you know now they're using so you know that's what I mean you can't commit heavy um you know be light on your feet >> that's like a boxer just be light on your feet advice um Jacob asks what's your evidence bar before you say this traffic swing is due to AIO visibility versus seasonality, brand noise or quite sight
changes. >> Great question, Jacob. That was the uh seasonality. Um when we looked at the past four years of traffic to the site, uh it always May was huge and then it peaked up into July, August and then came down around October every year. And so when we see we're going up in May and then all of a sudden the week that AI overviews is launched even the day it starts just going flat and and flattened down very obviously something wrong and and then you can go just go look at the page there's no ranking site showing up within three or four screens. So, you know, where you used to have that visibility without a scroll on desktop to now you you you're a dozen scrolls down. I mean, that that's that's pretty clear evidence. So,
>> that's that's very clear. Um, and we have one last question from Reuben. Reuben asks, "When AIO's take a chunk offormational clicks, which asset types have best defended CTR for you? video tools, calculators, images, news, and how do you prioritize what to build first? >> Oh, good stuff. Good stuff. It It's very interest. And and I'll say that because it's it's going to be, you know, it's going to be found in other places as everyone learns from each other and starts to compete with each other. Um, so what I am seeing is content that answers questions authoritatively. So it could be text. Um, I'm not seeing video video type content be picked up as much. Um, it still seems to be revolved around text. Calculators and tools are good for
keeping people there. Um, and and so that's going to play into the SEO side. >> Yeah. and and so I'm still relying where I have seen the most results is specifically a question and answer chunk and that answer is either bullet points or very clear statements >> of of of and that that is what gets that's what gets taken more than anything else. >> No, absolutely. No, absolutely. And where can people connect with you? >> Uh you can connect with me on site logic sitic.com. Uh and that's the best place. I'm also on LinkedIn. Uh and so yeah, love to uh connect with anyone if you got additional questions. >> Amazing. Thank you so much for that. Honestly, it was very very impactful. So appreciate your time with that.
Key takeaways
Chapters
Q&A
Commit about 70% to your strategy and watch the results carefully, since AI changes rapidly and unpredictably. Focus on brand searches where you know your brand will rank, but stay flexible—no strategy is 100% certain right now, and priorities can shift within weeks. — Matt Bailey
Compare historical traffic patterns across multiple years to establish seasonal baselines, then look for anomalies that align with AI overview launch. Check whether your pages have disappeared from visible ranking positions within the first few screens of search results. — Matt Bailey
Question-and-answer content with clear bullet points or concise statements generates the most results. While calculators and tools are good for engagement and SEO, video content is not currently being picked up as much in AI overviews. — Matt Bailey
Quotes
“The key to learning and applying information was community. Back in the early 2000s, the vast majority of SEOs were on forums and sharing information, sharing screenshots, debating. There was a full-on community.” — Matt Bailey
“If you have information that people need, you will be found. That's the bottom line when it comes to search and AI. It's in the search engine's best interest to give people the content that they want.” — Matt Bailey
“If you never talk to your customer, you will never understand their intent. Get out from your computer, make a phone call, visit with them, do one-on-one talking with your customers to learn how you solve their problem.” — Matt Bailey
“There's nothing you can do right now that's going to be 100%. There's no one who knows what's going to happen next month.” — Matt Bailey