A New Era in SEO with Cyrus Shepard

In This Episode

• [02:23] – Cyrus Shepard dissects the art of reverse engineering Google’s algorithms, hinting at methodologies that could redefine your ranking strategy.
• [07:09] – Cyrus shares insider experiences as a Google quality rater, revealing how these insights reshape his approach to SEO at Zippy.
• [12:24] – Learn why effort and experience on a webpage can make or break your site’s success in Google’s eyes and how to manifest these in your content.
• [16:28] – Cyrus reflects on his Moz years and how embracing a philosophy of exceptionalism continues to influence his work at Zippy.
• [26:19] – An unexpected twist in Cyrus’s disavow experiment: he reveals the tool’s surprising inefficacy—an insight that may change how you perceive link cleanup.
• [29:37] – Cyrus introduces a lesser-known metric increasingly pivotal in SEO—brand authority.
• [32:22] – Discover actionable steps to boost brand authority for small businesses.
• [34:29] – Stephan and Cyrus discuss the rising importance of personality and authority in content, a shift away from keyword-dominated SEO practices.
• [37:14] – Deep insights into how unlinked mentions can still bolster your SEO and brand positioning in an AI-driven search landscape.
• [45:56] – Contemplate the future of SEO with Cyrus’s outlook as he merges optimism with a call for strategic adaptation in an ever-evolving digital marketplace.
• [49:59] – Cyrus explains how to get in touch with him.

Jump to Links and Resources

Cyrus, it’s really good to have you on the show.

Stephan. Thank you very much for having me. Happy to be here.

Let’s talk a bit about Google’s algorithms and how you reverse engineer them and figure out what’s working and what’s not, as well as how that’s been done in the past. I know you were a quality rater for a while, and you learned some cool things from that, but there are plenty of other ways you reverse engineer Google, so let’s talk about that.

Yeah, so for years, I spent a long time working with the software company Moz, formerly SEOmoz. We used to run huge correlation studies across thousands and thousands of websites and URLs. Today, my ambition is much more focused and generally, what I do from a data point of view, especially with the last couple of years of Google updates, is Google has made major changes, and we see huge swings in winners and losers. So we have a database of a lot of sites.

Generally, what we do is we look at those sites that have one, the outliers that have done exceedingly well after a Google update, and then we do the opposite for the loser sites that have just been wiped off the map. And that’s a little different than how we did it at Moz because we’re just looking at the outliers, and then, we try to find statistically relevant correlations between those two sets and throw things against the wall. Let’s look at the title tags, schema, branded anchor text, etc. Most things don’t have a strong correlation.

Google isn't rewarding keywords anymore - they're rewarding brands. Share on X

So, then again, we’re looking for the outliers. What are those factors? Out of the hundreds of things that we looked at, what are the things that really stick out here? That’s the basis of how we do it.

Yep. Do you calculate standard deviation and..

Yeah, P values, strength of evidence, all those things that I didn’t quite understand when I had all the data sciences doing it at Moz. Today, these tools are much more available to people with any sort of strong data background. AI is a big help. I think the important thing to realize is that even when you find these correlations, you’re not saying this is a ranking factor. We’re using title tags as an example. If we find a huge correlation, we’re not saying that Google is suddenly honing in on this one factor. But what does this suggest about Google’s larger ambitions? We know Google is using a lot of machine learning to come up with its algorithm. How are they placing websites into buckets for evaluation? How are they putting websites into winners and losers? And generally, what we find, I mean, not to cut to the chase, what we’re finding is Google has become much more anti-SEO over the last couple of years.

So, things we used to talk about, such as optimizing your top tide links and optimizing your internal links when those are taken to an extreme and combined with certain business models, seem to be the sites that are losing more and more. Sites that aren’t aggressively attacking SEO like we’ve been advising for years seem to be doing a little better.

Yeah. So, are you alluding to the helpful content updates?

I think Google pulled a big magic trick with us by calling it a helpful content update. I don’t think it had anything to do with content. When you look at the winning and losing sites, it has to do with the aggressiveness of SEO. It was really helpful content. I just call it an anti-SEO update because that seems to be what it was, which hurts me because I’m part of a group of people who taught folks for years how to do better in Google’s algorithm. And all of a sudden, boom, those tactics stop working like they used to. And it’s been a big readjustment for me.

And I’m not sure the entire industry realizes what has happened yet, even today. They think the talk is about Google’s niche sites to the extent they do, but it’s more like Google is fighting back against SEO because we got too good at our jobs. Right.

So, what are some of the areas of SEO that seem to backfire the most when you get too good at it?

Yeah. So, one thing that hurts a lot is internal linking. I did a study a couple of years ago about anchor text. In that study, we found that the more anchor text variations in your internal linking, the fewer sites performed better in Google search than those with fewer internal anchor link variations.

More recently, when we studied that, after the helpful content update, that relationship flipped. And we haven’t published a lot of that. However, we found that sites with more internal anchors are actually on the downward slide. Sites with what we would traditionally consider poor internal linking or poor internal anchors were the sites that seemed to survive the helpful content update. So you combine that with a bunch of other signals, and again, it’s the sites that are doing it really well.

Pushing those limits were the sites that Google punished. And it kind of makes sense because Google is trying to eliminate unhelpful, aggressively SEO’d affiliate ad-supported spam, even if it’s not spam. So, it kind of makes sense that that’s what Google was targeting.

Tell us a bit more about your experience as a Google quality rater. How did you get that gig? How long were you there? What did you learn? What were you able to pull from that experience into your work today at Zyppy?

Yeah. So, I knew the helpful content update was coming. Google announced it in 2022. They had some really small updates, and I thought, “Yeah, it seems like something big is coming.” So how are they doing this? And I knew they were placing websites with their machine learning. They’re placing websites into buckets. And I’m like, how are they getting this data? Google admits that it uses quality rater data for machine learning purposes. And there’s evidence that that’s exactly how they use quality raters to get training data.

Google is trying to eliminate unhelpful, aggressively SEO’d affiliate ad-supported spam, even if it’s not spam.

So, I thought I should become a quality rater. If nothing else, this is going to make a really good story. So I went online. There are various companies you can apply to, and it’s not a great job if you’re trying to make a living as a quality rater. It’s not fun. And it took me two weeks, probably of unpaid effort, to land the job because you have to take all these tests, you have to do all this studying, there are the quality raters guidelines. But then the training materials dive deep into certain sections of that, and you have to know it. I do not know it perfectly, but I know how to apply it.

It was test after test. I actually failed one of the tests. I thought I’d been doing SEO for 15 years; I didn’t need to study. But it turns out I had to study really, really hard and eventually, I got it. There are all sorts of orientations. So finally, after two weeks, I got my $15 an hour job. And then once you start the work, it’s numbing, mind-killing, numbing work because you’re just evaluating websites for hours. You have your mobile phone.

That’s something a lot of people don’t realize. Google’s thousands of quality raters evaluate every site on a mobile phone. Then they do the work on a desktop, and they’re checking their phone, and there’s a little app where you send the results to your phone, and you’re scrolling through, and you’re, you’re checking the websites, and you’re scoring, scoring, scoring, trying to look up who the authors are, things like that. So, it’s not a great job, but it was very educational, and it really made me learn to look at sites the way Google does and what Google’s aspirations are.

The sites you rank high don’t necessarily rank well every time, but you can see where Google’s quality signals start to come in, what they’re trying to do, and what their ideas of a quality site are. So it was a really educational experience, and I got laid off because Google apparently laid off all North American workers because of the contract dispute. So I no longer do that. But it was good while it lasted.

At $15 an hour.

Quality scores attached to a website are really all about trust.

Yeah, not great. It is so interesting. Mark Williams-Cook, a great SEO working in the UK, recently discovered a Google exploit where he could see some of the code behind the scenes. He reported to Google that he had received a big bug bounty. But in that exploit, he found that for every website in Google search results behind the scenes, there’s a quality score attached to it. And it scored from zero to a hundred. And he was surprised because government websites, even if they were crappy, even if they looked terrible, didn’t necessarily answer the user question very well. Government websites almost inevitably had a very high-quality score. I thought that telling because, as a quality rater, quality scores are really all about trust.

Would you send your grandma to this website? Would you like to buy something from this site? Things like that. As a quality rater, Google instructs you to rate government websites extremely high by default. So when Mark revealed that in his Google exploit, government websites almost always had high-quality scores, it was not surprising to me at all, and it kind of validated when I thought about Google using quality rater data to train their quality algorithms.

Their quality rater guidelines are about 170 pages. What are some of the most critical nuggets of detail that you want to pull forward from that document for our listener to apply to their SEO tactics?

Yeah. So, I think what interests me the most is what Google changed with a helpful content update. The new directives they put in were about experience. Have the creators of the website experience with the subject matter before? In the old days, you could write an article about anything. And if you had enough domain authority on the Internet, you had enough people linking to you; you could kind of rank about any topic. You could review any topic. Wirecutter and The New York Times could review any sort of product they wanted. Now, quality raters are looking at the creator’s experience with a certain topic.

The sites you rank high don’t necessarily rank well every time, but you can see where Google’s quality signals start to come in.

So you want to demonstrate that on the page. If you’re reviewing a product, you should have pictures of yourself actually reviewing the product. Talk about your methodology, taking things apart, ranking and reviewing things. Do you write about this thing often? Is your subject a subject matter expert in these? So experience is definitely one thing. The other thing that I think people ignore, and this is huge in the quality raters guideline, is the idea of effort. How much effort have you put into this webpage? Did you write it with ChatGPT and just copy and paste it, or did you actually put some nice design into it? Did you actually do original research? Do you have information on there that no one else has? And I bring this up, and sometimes people say, “Well, Google is not a judge of effort.” Google can’t tell how much effort I actually put into the page. Well, they kind of can because they’re using machine learning.

They have a bucket of pages over here that people like me have scored as high effort and a bucket of pages over here that we’ve scored as low effort. Then, they use a machine learning algorithm to try to decide where your website lands. So they can’t actually tell how much effort you put in, but they can look at you and sort of guess. It’s an imperfect algorithm, but you want to look like those pages that have put more effort into it. So effort and experience are two things that they’re wishy-washy ranking factors. But I think they’re highly important and often overlooked.

Experience is something that makes sense when you think about it. You could ask ChatGPT about the implications of getting a brow lift, and it’s just going to give you some answer, but that answer could be wrong. It could have hallucinations in it. And then your health is at stake, or at least your looks after you get the brow lift from some faulty advice generated by whatever LLM.

So having verifiable expertise, and let’s say it’s a plastic surgeon who’s writing the article about the implications of having a brow lift that can be verified by a quality rater looking at the page, clicking on the link on the author’s bio section there to go to a larger, longer page about the author, see that the person has certifications and diplomas and a certain number of years of experience, et cetera, et cetera. So, how far does a quality rater go in checking that sort of information? Will he or she go to the about page or go to the author page to try and collect that information about things like diplomas and certifications and all that?

So, I would say it’s more superficial than what people think. And you have to realize the quality raters they’re rating, they’re working fast. When you get a search result, you get 10 blue links and need to move on to the next task. And you’re just doing that over and over again every day. One couple of checks that every quality rater will do is they will do an independent Google search about the website.

So you can list it in the quality raters guidelines as a number of queries, x.com reviews, -x.com, and things like that. There’s also, if you go to the search results, there’s a little thing called about these Results, and you can see more about these websites, and these are mentions and links about the website that you can do a quick reputation check. So those are basic things that every quality rater will do.

But it’s also really just kind of looking at the website. If I’m evaluating a query like bicycle tours in France, I am looking at a bunch of results. I’m going to look at it. What does this company do? Does this company specialize in bike tours in France? I’m going to look at the navigation, or do they specialize? Do you know if they are? Are they just an article site that is trying to rank for this term? So, things like navigation and the photos you put are huge in this process, especially when evaluating experience. And it’s something I’ve worked with. I’ve changed something about my SEO and my SEO process in the last year. Working with clients, we emphasize photos so much more and display that firsthand experience by emphasizing pictures of people. And we seem to have positive results for it.

I know you want a checklist, but putting things in these buckets is really more psychological. But I think certain things you can do are really quick and easy. Just showing people. I am an expert in this subject matter.

So, what were some of the things that you missed the most at Moz? You’ve moved on from Moz. I don’t remember how long you were there, but it was quite a while. What do you miss the most about technology, culture, studies, leadership and so forth? And are you incorporating any of that into Zyppy?

Experience and effort are highly important and often overlooked ranking factors.

Moz was a very magical experience for me. I came up in my career through Moz. I was working as a freelance SEO in Seattle, and I am just starting my career. I thought I was getting underpaid. So I googled how much an SEO in Seattle should make, and this job listing came up for Moz. And it was a customer service rep. I’m like, “Well, I’m not really customer service, but I love the company and know SEO.” So yeah, I can answer calls about the tool and things like that.

So I started. They put me on the morning shift answering calls, which meant my first job in the morning was unloading the dishwasher. It was a very small company then, so I’d be there at 6 am unloading the dishwasher. Like, “Yeah, I’m working my dream job now,” but I caught a few lucky breaks, and I was very quickly promoted to the role of lead SEO at the company. And it was a magical time. It felt like Rand Fishkin, the founder, great leader, great person, provocateur of the Internet. It felt like we were inventing the Internet for a while, making up our own rules. It was an age of discovery, doing things that were just amazing.

Now it feels like everybody’s copying everybody else on the Internet, and there’s less originality, less daring. But that was a magical time. One thing that felt great was that Moz had this philosophy called TAGFEE, which was transparency, authenticity, generosity, and things like that. One of them was exceptionalism, which means that we produce exceptional work. That was a great philosophy to live by because that meant we could put extra effort into everything we did and use that as a justification within the company. No, we need to make this better. So, if I’m releasing a ranking factors report or a tool, we would spend all the extra time on it to make sure it went well when we released it. And that’s what it felt like during the golden age of Moz. We were just releasing something wow every few months, which was great.

I’m trying to do that work with Zyppy now that it’s a wow moment, and we have put something out there. Something we feel great about, that’s original and new and fun. We’ve got some things in the works that I can talk about in a few weeks that I think are going to be exciting, but that lesson of exceptionalism and generosity and creativity, I think that’s missing a little bit in today’s SEO world, and I really miss that.

I bet you miss Rand as well.

Yeah. Fortunately, we’ve remained good friends. We used to be practically neighbors in Seattle. I only see him in person a couple of times a year now, but he and his wife, Geraldine. However, friendships with them have been one of the most valuable things that came out of Moz’s experience. So I’m one of his biggest fans.

Yeah. That’s awesome. You know how I co-authored a book with Rand, The Art of SEO. 

I remember that. Yeah, tell me, tell me.

So he and I were on a panel together called Give it Up at SMX Advanced.

Yeah, yeah.

And I blew his socks off. He was just really impressed with all the ninja stuff I came up with in my presentation, and he didn’t say much to me then. However, a few weeks later, we happened to be at another conference called SES Toronto. And in the speaker room, he came up to me and gave me a hug.

Oh, wow.

Hardly knew the guy. And here he came up and gave me a hug. So I was like, “Wow.” He told me, “You really brought it. That was incredible. I loved that presentation you gave.” Because I’ve reversed engineered the group to Google results. There are positions 1 and 2, but it was actually a position, let’s say 1 and 9, but you didn’t know it was 1 and 9. It just looked like 1 and 2 because Google was artificially inflating the position visually.

I showed how you could figure out the true position by adding num 9 into the URL and num 8 and so forth until it finally dropped an indented result.

Yeah, I just want to acknowledge the awkwardness I’m in that I really should just be interviewing you. People look at me doing SEO for 15 years. But you are. You’ve been there since the beginning doing, you know, expert-level stuff. A lot of it is something that a lot of people today don’t realize. Some of the foundational knowledge that you started sharing back in the day. So, my hat is off to you, sir. And you know, in that perspective, I’m sure.

Yeah. Thinking about what I have done is impressive. But man, you are the original OG, so kudos to that.

Thank you. Yeah, it’s been a long haul. Now it’s 30 years.

You have to go above and beyond with comparisons, data, and analysis - do the things ChatGPT can't do. Share on X

Damn.

Years ago. Yeah, I know.

Damn.

I built my first websites in 1994, and they were database-driven, real-time, updatable websites by the users. They would update the database. Yeah.

Damn.

Anyway, that situation was where he came up, gave me a hug, and did everything. It’s just really special. And we had a really wonderful conversation. And in that room, in that speaker lounge, we decided to do a book together. And it just so happened that Danny Sullivan, a founder of all these different things, Search Engine Land, SMX, and everything, Right? So he happened to be at O’Reilly‘s Unconference at that very moment. So Rand texted or called. I think he just called Danny.

And Danny was at Foo Camp, friends of O’Reilly conference. “Oh, you want to do a book together and look for a publisher? Well, I happen to be at O’Reilly’s Foo Camp. Do you want me to ask them if they want to do your book?” And went, “Yeah.” So, within minutes, we had a book deal. It was verbal, but it turned into an official contract and everything. It was going to be the SEO Cookbook. Then, we teamed up with Jessie Stricchiola to do The Art of SEO together. So that whole thing came about because Rand came up and gave me a hug.

That book has had such longevity. What edition is out now?

4th edition. It’s a few feet away. I’m not going to go grab it, but yeah, it’s 770 pages now. It was almost 1,000 pages in the third edition. And. And O’Reilly was like, “You can’t make this any bigger. Please make it smaller. Because every time you add several hundred pages to it, it sells fewer copies.”

Oh.

That makes sense, right? Like, who wants a big brick to have to read to learn SEO?

So much of that fundamental information is just so relevant. Despite all the changes in Google and everything else, the basics of information retrieval and how Google works are so relevant today to people that they would do well to read chapters in that book.

Yeah, well, thank you. But, you know, some stuff had to go. Even though we loved it, there’s no glossary in the fourth edition. That was a sad removal, but it’s a very up-to-date book. You can get plenty of definitions from Wikipedia, but the core knowledge is in the book.

So, yeah, yeah, I had that book on my desk my entire time at Moz.

That’s awesome. Okay, so enough about me and about Rand. What inspired you to start Zyppy? Because that’s quite a different deal than working at an agency where there’s a cult of personality and a whole ethos and all those brand identity and stuff.

Had a few stumbles. It’s been an educational journey learning what Zyppy is. So Moz was acquired by J2 Global Martech, a very large publishing company that owns a lot of online brands like PC Magazine and Ookla, which are things like that large company. Good people had a good experience with all of them. But it was time to move on. It was time for me and a lot of people to move on. It felt like a very different company at that point.

We must shift our businesses from a keyword information focus to a brand personality authority focus. Share on X

It was just not the same thing. I have nothing bad to say about J2 Martech. Very positive things to say. But it wasn’t Moz. It wasn’t Moz, the TAGFEE, the inventing the Internet, the being out there and making some controversy kind of company that I grew up with. So it was time to move on. I started my own software company, and that didn’t last long.

I learned some entrepreneurial lessons. My partner and I just couldn’t make it work. We were fighting every day, which was sad. So I learned I wasn’t very good at that. So, I’ve been consulting and doing my research ever since. I feel like my consulting just pays the bills so I can do my research. And that’s a horrible business model.

If I can share insights and secrets with them, that helps a lot. So, I’m trying to transition a little bit. I’ve always enjoyed connecting. A lot of people come to me. I’m in a fortunate position as a consultant. I get far more leads than I can handle. So, one part of my career that I’ve always enjoyed is connecting with people. People looking for specific SEO help.

If they’re looking for local SEO or link building, outreach or PR campaigns, connecting people to the right thing. So, I’m scaling that out a little bit. Part of my career has always been social media. Talking about things. That’s been an interesting journey. And I’m not quite sure what’s happening there. That’s going through some transitions..

So we’re having fun. But that’s where Zyppy’s at. It’s a consultancy that’s exploring a lot of interesting things.

So, let’s talk a bit about Disavow. I know you did a recent blog post or at least updated it recently. I disavowed every link to my website. Here’s what happened. Cause I love your research. You just have a hypothesis, run out, and test it. That’s really cool.

So, great question. I’ve updated this, but I haven’t published the results of the study. So, a little background disavow. There are questions about how Google’s Disavow tool works and what it’s used for. So I thought, “Well, let’s just disavow every link I can find to my site.” So I went through Google Search console Webmaster tools. I looked at Ahrefs Semrush. I just combined everything into a huge disavow file and disavowed at the domain level, something like several thousand websites. And I waited a couple of months, and nothing happened.

If anything, my rankings went up a marginal amount during that time, and I publishers. I don’t think the Google Disavow tool is doing anything. And I published the results, and many people came back, which was really interesting. Oh, you should have left it in longer. Another theory was that Google ignored your disavow file because no one disavows every link, and Google says they may ignore directives in the disavow file. So, I did another test. I disavowed a third of my links, but they weren’t necessarily my best links. They’re medium-tier links.

I submitted that and left it there. And I waited three months, four months. Nothing happened again, maybe a little down at this point, but I hadn’t published much content, so I couldn’t tell. So I re-uploaded just to see. I re-uploaded the original link disavow file, and I waited again. And just like I think a week and a half ago, I remembered I had this experiment running, and I looked at my traffic and literally over the course of a year, this entire experiment, nothing has happened. Google ignored every file I put up, and maybe there were little flutters, but I can’t read anything statistically significant about that. So it’s more like voodoo at this point.

The disavow tool, at this point, is just busy work that Google put into webmaster tools.

So maybe the disavow file does something; maybe it doesn’t. But I can’t seem to prove it either way, no matter what happens. The most interesting theory that I heard that I almost suspect might be true is that these days, Google disavows files. We know it works well if you’re hit with a manual link penalty, which is exceedingly rare, but they do happen to show that you’ve disavowed links. The other theory is that Google’s disavow tool only works on negative, spammy links that have been hurting you. If you have those, it will remove the pressing effect of those spammy links or bad links. Otherwise, it’s not going to do anything. I don’t know if that’s true or not, but I think it’s just the disavow tool at this point is just busy work that Google put into webmaster tools to get people to click on things because I can’t see that it does anything anymore.

Yeah, I still use it, though, because I’m superstitious.

That’s funny. Do you also throw salt behind your back while doing it?

Yeah, yeah. I have a general disavow file with a lot of Blogspot block domains and known spam sites and things like that.

That doesn’t hurt, clearly. What are your favorite metrics? And there’s the obvious like ahrefs has their authority metrics, DR and UR and so forth, and a number of keywords that are ranking in the top 10, et cetera, et cetera. Any kind of like space spam rank or brand rank type of metrics or link velocity. Anything that’s kind of not the obvious generally accepted metrics.

Yeah. So the new one that I think we can’t ignore these days is that it happens to be a mods metric. It’s brand authority. And brand authority is roughly an estimation of the search volume of people who are looking for your brand or your brand-related terms. And it’s a very imperfect metric because it’s kind of hard to determine. Some brands just find it really hard to determine those branded terms. However, what we saw after the helpful content update, and Tom Kapramas has an excellent study on this, is that websites with low branded volume search terms did poorly after the helpful content update. Sites with a lot of branded search volume seemed to be insulated from those results and thrived.

I’ve also done my studies looking at branded anchor text and things like that. And those relationships survive. Every single evaluation that we do that people searching for your brand insulates you from negative Google updates. But I think other tools, such as Ahrefs, Semrush, would be great. I’d love to see Ahrefs, Semrush develop its brand authority metric to make it more accessible for folks. But you also know they’re now imperfect tools. So you kind of have to look into your own data and see how many people are searching for you. I’ve been working on how to increase this brand search with clients. And you don’t. From my data, you don’t need a lot.

You don’t need to become IBM or Amazon. You just need a little bit. I don’t have a ton of people searching for Zyppy SEO, but I have some. When we look at two sites that have been impacted, you just need to have some, and you can leverage that. And that’s a huge insulating factor if you can get people looking for your brand and searching for it. I’ve made mistakes. A great example of a mistake I made when I wrote my terribly inactive newsletter. It is called the SEO Tips Newsletter. I did that, and everybody knows why I did it.

I was choosing a keyword. I wanted SEO Tips. I thought that would be a great SEO Tips newsletter. I think that was the wrong mistake. In this age, because it’s not branded, anybody can have an SEO Tips newsletter. So I’m rebranding it to the Zyppy, something, or another. I want that brand because when people are searching for the newsletter, I want them to use Zyppy. I want to get Zyppy out there.

I want people to remember Zyppy, even If it’s just 1% or 2% of them getting in there. I think leveraging that brand recognition is important today because that’s what Google rewards. Google isn’t rewarding keywords; they’re rewarding brands. So that’s the direction we’re going. Brand authority. Great question.

So, how do you move the needle? You have very low brand authority if you’ve got a really small company. What’s actionable from that metric for you? Like, what do you do with that?

That’s an excellent question. And the answer is whatever you can. I had one client who didn’t have a logo on their site. The irony is they were so customer-focused; they were customer-focused, and they didn’t want anything to distract from the information. So go on to their website. The article just starts: here’s how you solve your problem. And their brand information was just a little thing at the bottom of the footer to find out who was actually on this website. You have to bring your branding to the top of the website so at least people know where they are.

And maybe if they want to search again, they can find you. So we did that, and they’ve seen some positive results, but now, at least, we know their big, beautiful brand and their name at the top of the results. The other thing that I encourage people to do is brand everything. My newsletter example. If you have an online calculator, tool, or free digital product, make sure you’re branding it. Make sure you’re making it about you and your company. Do everything that you can to put yourself front and center. And I think as marketers, as SEOs, we haven’t been good about that historically because we’re all about the keyword.

We want to rank for this keyword. We got to start introducing ourselves. I have another client who’s done exceedingly well. He works in a formula space like Excel formulas and things like that. And it used to be just content. Well, he started putting himself front and center in every piece of content where he’s got a picture of himself, his name, and the company’s name. It’s not just information at this point; anybody can do Excel formulas, but if you’re attaching your name to it, you’re attaching your picture, you’re doing videos of yourself, then it’s your Excel formulas, it’s Cyrus Shepard’s Excel formulas, and just leverage it just a little bit.

If you have an online calculator, tool, or free digital product, make sure you’re branding it.

But we have to start putting ourselves and our company into our content. So, it’s not just information; Google has plenty of information. If you want to rank for the best things to do in New York City, they have something like 100,000 articles with that in the title. They don’t need information about the best things to do in New York City, but they might be interested in what you are doing in New York City if you can establish yourself as an authority in this subject. So it’s all about getting ourselves into the content. It’s not easy. I wish I had easier answers, but it’s hard work, and that’s why building a brand is hard.

Especially more important as AI overviews and ChatGPT are taking more front and center positions for the searcher, the user. And a lot of times, they’re not getting to those 10 blue links. They’re getting a full answer right from the algorithm. And that algorithm is essentially stealing from all these thought leaders who provided all the research and had all the experience going to all those sites and rating them and grading them and all that. And it’s just very frustrating for a content creator to face this new reality. What do you recommend we do about that?

This is tough because it’s the reality. And I think the AI answers are getting scarily good. I am often googling things, and as a searcher, I’m just not clicking into search results because Gemini is summarizing the articles for me, and I’m just getting that answer. But I am doing a lot more searches that way. They’re more rapid. I’m still clicking on search results, and I think a lot of people, if we’re looking into our analytics right now, might see a little bit of a decline in traffic. But people, at least for now, I don’t know how long it’ll last. People are still clicking on search results.

I think the answer is again going back to that shift of ChatGPT & Gemini. They’re giving information. Everybody has so much information. So, we need to shift from a keyword information focus to a brand personality authority focus. One thing I work with my clients on is, traditionally, if you have an article on the best mattresses for side sleepers, you try to get links to the best mattresses for side sleepers in the anchor text or other things like that. And I’m telling my clients to stop focusing on those keyword-rich keyword-rich links they don’t want. Let’s shift to about you. Why are you a mattress expert, and why are you an authority in the mattress space? So, let’s go out and get some articles and links about why you are an authority.

Getting yourself mentioned in those AI overviews as an authority, as a reference link, is going to become increasingly important.

I don’t want links about the content; I want links about you. So, if I can get my clients onto a podcast or do an interview where they’re talking about their expertise, I think that’s much more valuable today than it was five years ago when brand building was a fuzzy metric. So, getting yourself mentioned in those AI overviews as an authority, as a reference link, is going to become increasingly important.

Yeah. And a mention that doesn’t even have a link necessarily can still be valuable.

Yeah, absolutely. I tell people we’re just getting to know Google’s knowledge graph. I don’t know what you call it with ChatGPT and large language models. We’re just trying to get out there. We’re just trying to get into those graphs of knowledge so that we’re referenced more and seen as an authority. It’s all kind of the same, right?

Yeah. So, there are different ways of referring to it. Unlinked mentions are one way to refer to it. But not every appearance of, let’s say, Zyppy needs to be a link to the Zyppy website, right?

Exactly. It’s more fun when it’s a link, and it’s more fun when it’s a follow-up link. 

Or two links.

Yep.  Seriously, I do a lot of podcasts like this, and sometimes the podcast will, you know, link to a certain article, but sometimes they don’t. It doesn’t really matter to me. Doing marketing for my own company, getting the name out there, and getting Zyppy mentioned as much as possible are important.

Makes sense. Now, as part of your offerings, do you do audits?

Oh, yeah. It’s weird because when Google came out with the helpful content update, I was auditing many sites that had seen steep traffic declines, which was great because I was learning and studying. But it’s been a real struggle working with sites hit by these Google updates because recovering them requires a fantastically huge amount of work. And sometimes, you have a completely different website or a completely different business model by the time you’re done. Not all clients have the resources or the will to do that. I’ve had a couple of clients slog through, and they’re seeing great results. Some clients have just had to shut down because, you know, they’ve got a month of runway and six months of work to do. It’s been an incredibly educational learning experience, but doing those audits has been a struggle in many ways.

A mention that doesn’t even have a link necessarily can still be valuable.

I am always happy to do audits, but it’s tough when you’re working with folks who have seen their entire business wiped out, and Google doesn’t want anything to do with them.

Yeah. Is it something that you normally do for a new client? You’ll audit their site to get land and determine the biggest weaknesses and missed opportunities. Or oftentimes, you’re just jumping right in?

Oftentimes, yeah. I love working with the big 10 million page sites that find something in their sitemap files or Twitter tweaking some technical issues at a scale that can get a million more pages indexed within a month. And we see that that huge hockey stick growth is really fun. It takes a while, but I love doing that kind of work. I also tend to work with a lot of early-stage startups. Just being from the Seattle tech scene, I had a lot of connections there. So, many of my clients come to me with no website. Now we start from scratch, and like, “Okay, let’s build a two-year content strategy and go from there.”

Those are fun, too. So I enjoy doing that. It is weird signing a client with no website, and there is no audit to do, and you’re like, “Okay, let’s start.”

Yeah, that’s fun. That’s a grassroots kind of ground-floor opportunity.

Yeah, yeah.

How do you use AI to assist you with audits and content creation, optimization, and link building?

That’s a great question. I’m probably behind a little bit. Every AI solution I’ve tried for content creation has. I’ve been disappointed in the results, and I prompt everything to craft it exactly how I want it. I think it’s fine to outline, at least in my content creation stage. It’s fine for brainstorming and outlining, but I just can’t have it do the writing at this point. I love using AI to help me with data analysis. I am doing things in spreadsheets these days that I couldn’t dream of doing five years ago with formulas that are so overly complicated and work that it’s truly amazing.

My front-end coding skills are amazing now, thanks to ChatGPT. I could fool you into thinking I’m a CSS expert at this point. So those things are fun. But the answer is I’m a little disappointed with ChatGPT for replacing most of my content work and my SEO work. I still find it easier to do many of those things manually, and I’m not a big fan of some of those automation workflows. The next 6 to 12 months may change my mind as agents and prompts improve. But I’m a little behind.

Oh, I don’t know that you’re behind. I think you’re cautious, and I think that’s a smart thing to do. There are tools. I’m not going to name one, but I just looked at one that I was really unimpressed by, and they’re kind of cagey about giving examples of sites where they’ve done the hockey stick or whatever. So I did find one, or actually, one of my team members found one mentioned as kind of like an example site that the founder of the tool set up as a test and showing, you know, up into the rain. It’s doing really well. Well, it’s tanked as of at least two weeks ago. It’s nowhere near what it was before, so it doesn’t seem to be working.

Going back to Ramaz’s days, one of Rand’s philosophies was to do things that don’t scale. That’s the only way to differentiate yourself long-term from the competition. Google has Gemini. They can create articles and answers at a scale much better than you. And so that’s not going to be your competitive advantage. You have to do things that don’t scale. You have to do that, get the data and the information that others don’t. So, I think a lot of these strategies work, and they may work for a while, but I think the long-term prospects are not looking good for these scalable AI strategies because the technology is cheap, and anybody can do it. So, what’s your differentiator?

Great point. So, what would be an example of a great differentiator? How do you put that into practice of creating something that’s not scalable and really differentiates?

Yeah. So, working with the HCU updates, I was working with many affiliate sites, including one I remember looking at that never seemed to lose any updates. I’m having trouble remembering its name. Something like, oh, I was, I don’t want to put them, but I will. It was called NapLab. I haven’t checked their traffic recently. They may have tanked by then, but they survived every Google update, and they reviewed mattresses. It seemed to me that when you look at their reviews, they were doing such hard work. There were videos of them bouncing on the mattresses.

Data points are the most detailed data points about mattresses you’ve ever seen in your life. They were doing many things wrong. They were way over-optimized, they were targeting, and they were doing all the things that we’ve talked about in this podcast. They were doing them wrong. But they had such. Obviously, they were spending weeks on each mattress, getting everything down right, including the videos and everything like that. That’s something that doesn’t scale. You can’t ask ChatGPT to review a mattress for you.

You actually have to do the work if you want it to be original. So that’s one example. If you’re just writing articles, sometimes you have to go above and beyond with comparisons, data, and analysis. If you’re just doing informational content, you have to go a little bit beyond the informational content and start making comparisons, adding data, adding unique interview people, and things like that. You have to do the things that ChatGPT can’t do and bring that real-world experience into the article. And you have to do that every single freaking time. And that’s how you do things that don’t scale.

Yeah. So, what’s the thing you’ve done most exemplifies that strategy?

Oh, great question. I would love to scale all my content. I think for me at least. Zyppy. We’re very proud of the data studies we publish, and there are studies nobody else has. I don’t have a lot of content on my marketing website, but title tag length is a topic I was fascinated with a few years ago because we had a tool that did that. You could write an article about the optimal title and title tag length, or you could do a study, research it yourself, and publish proprietary data.

You have to do the things that ChatGPT can’t do and bring that real-world experience into the article.

And that article, I think, has brought me tens and tens and tens of thousands of visits and survived. I haven’t updated it in years, but it survives because it is so unique and has a lot of links. You know, it answers people’s questions, and it has proprietary data. It has something in it that doesn’t scale.

I know we’re getting close to time, so I’ve got a couple more questions, and then we’ll wrap. So you said earlier that prospects weren’t looking good just for creating AI-based content and so forth. I’m curious about what the prospects are looking like from your perspective in terms of SEO. We get the SEO is dead kind of mantra happening every year, and it’s not dead yet. What do you think is the shelf life of SEO as an industry?

It was interesting when AI first came out when ChatGPT was released to the world. I thought that traditional SEO had a two or three-year shelf life. I think that was overly pessimistic. Nobody knows what’s going to happen, but we have a few dynamics. Google and all these AI crawlers need content, and they’re striking deals with publishers for access to that content. If you’re a small publisher, most of us are small, medium publishers, and we don’t have access to those deals. But Google’s and OpenAI’s ecosystems need an ecosystem where people constantly feed them new information, so they’re incentivized to send some of that traffic to us. I think it’s much less than the traffic the SEO world receives now.

I think it’s a fraction. I think the profit margins of publishing companies will get squeezed much further than they have in the last 10 years as people rely more and more on AI answers. I don’t think SEO is going away. Companies still need marketing. SEOs are. I think it was Michael King who I saw talk to the other day. SEOs are uniquely positioned in our skill set to help people appear in AI overviews more than any other industry. So I think we’re going to see a little bit of a shift to that.

When I work with a lot of SEOs, a lot of them are getting into AI optimization, pivoting to that as a service. Some software companies trying to explore this space, getting brands into AI overviews, are reaching out more and more and looking for feedback. So, I think SEO survives, but it’s going to look a little different in two or three years because people are still going to need our services. The number of SEO firms will probably decline, but I don’t think marketing firms will decline any. But services are going to look a lot different in a while. I’m not sure what it will look like, but it will be different.

Yep. Well, the only constant is change, right?

Yes. Yes, we have to change. Adopt early.

Yes. And it’s not the survival of the fittest; it’s the survival of the most adaptable. That’s really Darwin‘s thesis. It’s not that you have to be fit; you just have to be adaptable.

Yeah. And with AI, we’re still in the early adopter phase. There is far more land to stake now. Now is the perfect opportunity to get your foot in the door, start learning, and even offer services if you can, rather than waiting for it to reach mass market appeal. So yeah, get in early if you can.

Yep. Okay. Last quick question. This is a branding question, actually. Why the yellow glasses? Were you inspired by Seth Godin?

I would. Funny story. It was COVID-19, and when you went to a glasses shop, you had to make an appointment, and there was one person in the store at a time. This was in Seattle, following health code regulations, and you had to make an appointment; one person in the store at the time, you weren’t allowed to touch the glasses. They had to hand them to you and wipe them down. I’m like, I don’t have time for this. My wife went through it, and she got a pair. She wasn’t really happy.

I’m like, here’s what I’m going to do. I went to a bunch of online services. Online glasses are the cheapest. You know, you get a pair for like $40. And I just took my budget, which was like two or $300 that I’d spend in a glasses store, and I just bought seven or eight of the cheapest glasses I could find, and I got them, and most of them didn’t fit. If you’ve ever shopped online for glasses, it’s not always the best experience. Most of them didn’t fit. Whatever.

This pair resonated with me, and I started wearing it. And people. Everywhere I went, people were like, oh, nice glasses, Nice glasses. I can never have these now, so it’s time to order another pair. Fortunately, they still offer them online. I have to do these. Unfortunately, I think Seth stole my look, but he has more brand authority than I do, so people think I stole it from him.

Well, for now. For now, yeah.

Yeah.

I’ll try to bring that back, you know.

Right.

So, I’m brand authority shifting.

Right? Yes. Awesome.

Okay, so how does our listener or viewer find you? Work with you? Learn from you?

Yeah, there’s zyppy.com, and I’m active on most social networks, but that’s a confusing place. LinkedIn‘s always good.

Zyppy.

Zyppy, yeah.

For our listeners who do not see anything important on the screen, please Google Zyppy SEO.

Z-Y-P-P-Y-S-E-O.

Yeah, Google. Great. Awesome. Well, thank you so much, Cyrus. This was a lot of fun, illuminating, and inspiring. I hope our listener or viewer takes this information and does something valuable with it because it isn’t just edutainment. It’s for actionable insights to be applied in your business. Thank you so much, and we’ll catch up with you in the next episode. I’m your host, Stephan Spencer, signing off.

Important Links

Connect with Cyrus Shepard

 

Apps/Tools

 

Book

 

Businesses/Organizations

 

People

 

YouTube Videos

 

Your Checklist of Actions to Take

Demonstrate real experience. Include photos of myself actually testing/reviewing products, document my methodology, and show my hands-on expertise.

Prioritize branded content. To build recognition, rename generic content assets to include my brand name (e.g., change “SEO Tips Newsletter” to “Zyppy SEO Newsletter”) and prominently feature my brand identity across all content.

Place brand identity front and center. Add my logo, company name, and branding prominently at the top of my website rather than hiding it in the footer.

Shift link-building focus by getting mentioned in interviews, podcasts, and articles that establish my expertise and authority in my field. Prioritize building my reputation over keyword optimization.

Emphasize personal authority. Show that real experts with verifiable credentials are behind the information I’m sharing.

Invest in non-scalable content, such as detailed product comparisons with original data points, hands-on testing videos, and expert interviews. Document processes that demonstrate genuine expertise.

Monitor brand authority. Track the search volume for my brand and brand-related terms as a key metric. Focus on increasing branded searches through awareness campaigns and establishing myself as an authority in my space.

Optimize internal linking naturally. Focus on natural, contextual internal links that enhance user experience.

Demonstrate content effort. Show clear evidence of the work put into content creation through original research, unique data points, custom visuals, and detailed analysis that goes beyond basic information.

Connect with Cyrus Shepard. Reach out through zyppy.com for consulting services focused on SEO strategy, particularly for startups and businesses looking to build long-term organic search presence. 

About Cyrus Shepard

Cyrus Shepard runs Zyppy SEO, a global online marketing consultancy based in Astoria, Oregon – home of the Goonies. He writes about Google’s algorithm and frequently publishes new studies and experiments. Cyrus has worked as a Google Quality Rater and previously led SEO and content at Moz.