SEO Like A Google Pro with John Mueller | Mediavine On Air Episode 34

Mediavine On Air Logo

You’ve heard of SEO Like a CEO.

Today we’re bringing you SEO Like a Google Pro. As in THE Pro.

On the next episode of Teal Talk, Senior Director of Marketing Jenny Guy is joined by Google Search Advocate John Mueller. Yes. That John Mueller. SEO expert translator extraordinaire.

Listen in as Jenny asks John questions about updating existing content, images and videos for your site, the best way to structure your site and general algorithm questions. You don’t want to miss this episode!

Helpful Resources


[MUSIC PLAYING] JENNY GUY: Hello. Hello, everybody. And welcome to a very special edition of Teal Talk, the show about the business of content creation. I’m with you, as always, as your host, Mediavine’s Jenny Guy. And today you might notice it is Wednesday, November 17, as opposed to our plan of Tuesday, November 16. But that is all right. It is OK that we are so grateful that our guest was able to reschedule. Sometimes the internet is an unpredictable place. And that is part of why we are so thrilled to have our guest here today.

But as you know, next week is Thanksgiving in the US. And ordinarily, I would ask a question or two about your plans for the holiday, maybe get into a healthy debate about stuffing versus dressing, really important questions. But today, we don’t even have time for the briefest of bantering. That is because of my guest, who is joining us live from his home in Switzerland.

I have to say that we have been extremely fortunate over the course of four seasons on this show to talk with some of the most influential and knowledgeable people in the industry. But I’ve got to say, even with all of our incredible past guests, today is special. I already told him that I would try not to geek out, so I am working on myself.

As a result, the format of today’s show is slightly different. Just a little housekeeping. We asked you to submit questions in advance, and you guys have showed up. In fact, if I asked everything that was sent, we would be here for five hours.

Instead, myself and my team have put together a list of questions that we hope covers a wide range of important current CEO topics. So in order to get to as many of those topics as we can within our time restrictions, I won’t be taking questions live. Please keep commenting with your questions and feedback, though, as they may show up in a future episode if we are able to convince John to return.

So without any further ado, let us get SEO advice from the source. Today, I am joined by the one and only John Mueller. I’m going to read his bio real quick just to wow us all.

John Mueller joined Google in 2007 and works in Zurich, Switzerland, where he coordinates Google Search Relations efforts as a Search Advocate. He and his team connect the Google internal world of search engineering to those who create and optimize public websites. Together, they help to bring an understanding of the external web ecosystem to internal teams and help publishers, like us, make awesome websites that work well for users and search engines.

OK, here we go. Our first question. John, welcome to Teal Talk for the first time.

JOHN MUELLER: Hi. Nice to be here. So cool.

JENNY GUY: Wonderful to have you. Thank you. OK. Christine from Taste of Maroc. Do we run– we’re starting with a section on updating existing content, which is something we preach a lot of here. All right. Do we run the risk of losing rank on a top post if we make a minor update, such as adding an internal or external link, correcting a typo, adding a custom excerpt, breaking a long paragraph into two paragraphs, et cetera? How about adding additional photos to a post?

JOHN MUELLER: No problems. So all of these things sounds like great things to do with regards to content, kind of keeping things fresh, making sure that you’re fixing small issues when they pop up. And there’s definitely no downside from a search point of view with regards to any of these changes.

I think when you’re adding photos, the only thing there is, of course, we find another photo that we could index, which is almost like adding another way that people can find your website in search. So it seems like a good thing overall.

JENNY GUY: All positive things. Love hearing that. OK, so there’s no reason to be afraid that we’re going to ruin our rankings by updating existing content.

JOHN MUELLER: I mean, as long as the primary content is the same and that you’re not removing something that people are finding at the moment. So if, for example, you have this one mention of a word on a page, and it’s the only reason why this page ranks for that word, and you remove that word, then, obviously, yeah, it’s going to be hard for us to show that page for that query.

JENNY GUY: So before you do that, if it’s a high ranking post, make sure to look into where the Google– look into your Google Analytics. Dive in, find how people are finding you, and then go ahead.

JOHN MUELLER: Yeah, I mean, small stuff like fixing typos, I don’t see that as being an issue ever. But if you’re significantly rewarding things, if you’re changing a title completely, if you’re swapping out all of mentions of one word, then, obviously, that can be a bigger thing.

JENNY GUY: Love it. OK. Fantastic. So we’re going to open up a couple of cans of worms here. While we’re on the subject of updating older posts, here we go. Let’s talk about republishing.

When substantially– this is from Sherry Smothermon-Short of Cub Scout Ideas. And this is not only her, though. This is everyone. When substantially updating a blog post, can-should we publish it with a new date? Our CEO published a blog post a few weeks ago advocating for changing the updated or modified date, but leaving the published date alone, as per Google’s policy. Where are you on this?

JOHN MUELLER: I think that’s a great idea, yeah. One of the things that I notice with websites is some sites get a lot of traffic from fresh content. And that’s something where, especially if you’re a blog, if you’re regularly publishing content, or if you’re a news site, then essentially through the new content you keep appearing in search again.

So if you have something that you’re significantly updating and you’re saying, well, this is not just kind of like fixing a typo, but you’re actually revamping this page, then that feels like something where you put a new date on it. You make sure that Google understands this is actually a new piece of content and republish it like that. And I think that’s essentially a good idea to kind of stay on top of things.

Obviously, you don’t want to overdo it and just republish everything across a website. Because your users will notice that too. If they keep coming back, and they’re like, oh, yeah, I saw this blog post last week. And it just has a different date on it. Then that doesn’t really encourage them to keep coming back.

JENNY GUY: Definitely. OK, so it sounds like if you’re doing a huge revamp of the post, essentially changing major components of it, not just a couple of photos or switching typos, then yes. Where do you fall on the published or modified date? So we have the original published date. And then we usually– so the ideal would be to have a modified date, an updated date, but rather not changing the particular specific published date? Or where are we? We get very in the weeds on this, John, all of us.

JOHN MUELLER: I suspect from a web search point of view, there is no big difference. I could imagine that if this content shows up in Google News, then maybe there are some Google News guidelines in that area.

JENNY GUY: They do have that, yeah.

JOHN MUELLER: I don’t know. I don’t know too much about the Google News guidelines. But if it’s essentially a blog post, if you’re essentially looking at web search, then I suspect there isn’t a big difference, which is probably why this debate keeps going, because it’s like, if there’s no big difference, then both sides see that it’s working. So that’s kind of my take there.

JENNY GUY: OK. Thank you for sharing that. All right, so because we love controversial topics around here, let us hit another one, which is deleting old content. We understand that it can definitely make us feel better as content creators, getting rid of old posts that no longer apply to our niche. But is there a benefit as far as Google is concerned? Does it impact, for instance, a specific crawl budget for a lifestyle website?

We talk about– or our CEO has talked about is not linking to posts, as opposed to deleting. But if you are going to delete, verify the post isn’t getting traffic, and then remove all links. So where are you on this?

JOHN MUELLER: Deleting old stuff. So I think from a crawl budget point of view, for pretty much all blogs and that-sized websites, you’re not going to see any impact there. Usually crawl budget is something that you would impact if you’re creating a multiple of the pages that you have.

If you have a website that already has a million pages, and you’re creating versions– I don’t know– or you’re changing a sort order around. And suddenly, instead of one page or like millions of one page, you have 10 variants of all of that. Then it’s like going from 1 million to 10 million. That’s a significant change. But going from even saying like a large blog where you’re 100,000 pages going down to 99,999 pages, that’s not going to change anything.

So crawl budget I would ignore completely here. With regards to removing content, I think sometimes that makes sense. When I talk to the search quality teams, usually they say it’s better to improve content than to remove it. But obviously, when it’s a larger website, sometimes you’re looking at just the scale of things. Oh, I have 1,000 things I should be improving. And I have time to do maybe one a week. Then at some point, you might say, well, it’s not worth my time to focus on improving all of these 1,000 things. I’ll instead make a bunch of new stuff and just delete all of the old stuff.

And from my point of view, I think that’s a perfectly fine decision. It is something where you need to be careful that you’re not deleting anything that is of value. So like you mentioned, looking at the traffic to those pages makes a lot of sense. Making sure that when you delete something, you really delete it properly, like removing the links like you said. I think all of that kind of matches in there.

And sometimes people also are not sure if they want to remove something. And they’ll mark it as no index, for example. And the advantage there is that people within your website, they can still find the content. Because it’s still technically on your website, it’s just no longer shown in search. And because people can still find it within your website, you can still track the visits to that page.

So you can play around with things and try to improve that page a little bit. If you see the visits go up significantly, then maybe it’s worthwhile to let it be indexed again.

JENNY GUY: I love that idea of a wait and see approach. What I think we emphasize the most as Mediavine is focusing on creating new content or really doing great updates on your existing content, as opposed to finding things you should throw out. That’s what we focus on. With the limited time– because you’re right. We’ve got a lot of things happening. And worrying about which posts to get rid of–

JOHN MUELLER: I think removing things is really something I would try to limit to the absolute necessity, where you’re really saying, this is something I can’t stand behind at all anymore. And you want to get rid of it. Whereas if you’re just seeing, this post is not getting a lot of traffic, then it’s like, well, you have lots of other posts. And maybe they are getting traffic. So it’s like, don’t worry too much about these individual posts that only get traffic maybe once or twice a year, because maybe they’re unique enough that they have value once or twice a year. And that’s still valuable.

So just kind of like going in and saying, oh, I need to purge 10% of my website every month and taking the ones with the lowest page views, I don’t think that makes sense.

JENNY GUY: Thank you. We’ve heard a lot of advice from different sources where people are saying, it’s great to prune your trees and prune your content trees. And it’s better for Google. And thank you for clarifying on that. We just say, create more good content.

All right. Katalin from Spatula Deserts. Here is your question. Hope you’re watching. Why are none of my new posts getting indexed since the end of September? How long does crawling really take for new content? There are some rumors about Google bugs and a lot of us feeling like we’ve been impacted.

JOHN MUELLER: Yeah. I don’t know. So I think this is tricky. And it’s something that we talk about with the team every now and then. And I think one aspect here is that we generally don’t index everything from a website. And I think pretty much as far as I can think back, that was never the case. And the tricky part is within Search Console, we present a lot of things as almost like where you can look at them and say, well, there’s an error.

So we’ll say, oh, we know about this page, but we didn’t index it. And then you as a site owner will say, well, Google knows about 1,000 pages. And they’re not indexing it. I have to fix this.

And from our point of view, we’re like, well, that’s kind of normal. We just don’t index everything. And I think that kind of first step of seeing the information in Search Console and taking a step back and saying, well, it’s like, this looks like an opportunity. It doesn’t look like a problem. I think that’s really important. And yeah.

JENNY GUY: No, keep going, please.

JOHN MUELLER: I mean, it’s just one of those things where almost through the transparency that we have in Search Console, people get confused. And they think, oh, I need to fix it. So that’s kind of the one angle there.

The other thing that I see a lot with websites in general is, at least at the moment from a technical point of view, most websites are actually pretty good. They use a common CMS. They use WordPress or whatever system that is out there. And where in the past, we would say, well, it’s not indexed because of these five technical things. You look at it now, and it’s like, well, it’s kind of perfect from a technical point of view. They’re doing everything right. Because it’s kind of forcing you to do everything right. So at least that part is taken care of.

But that still leaves the overall aspect of Google maybe just not– I don’t know– being interested in indexing everything on a website. Which I think as a site owner is probably really hard to take, because it’s some weird algorithm that’s saying, oh, you have 100 pages, but I only like 70 of them. But it’s kind of I think the practical reality that we have a limited capacity for indexing things. And we–

JENNY GUY: You’re saying Google has a limited capacity, John?

JOHN MUELLER: Yeah. Yeah. It’s like, we have to buy hardware like everyone else. We don’t have this magic– I don’t know– machine learning machine that just does everything for us. It’s like, people have computers. And I don’t know. Yeah, have to store the data somewhere.

JENNY GUY: Why, that is going to probably be the most revelatory thing that you say, which is Google isn’t magic. So I’m going to follow up that– [LAUGHS] you’ve heard it here, folks. I’m going to follow up that question with, do you recommend manually going into Search Console and submitting a post or indexing if you’re wanting to see quicker results?

JOHN MUELLER: I think on an individual level, sometimes that makes sense to kind of speed things up. But if you’re doing this for every post, then I would assume that it’s just– I don’t know, like you’re pushing five new ones in. And Google takes five out somewhere else. And it’s almost like, maybe there are other things that you can do on your website to make it clear that actually what you’re publishing is really important.

And that could be something along the lines of maybe not posting 10 articles a day, but rather posting one, kind of concentrating your energy into one really good thing. So that when we come by, we see, oh, there’s actually this one really good thing that we can pick up. Or making sure that within the internal linking of your website, you’re really making it clear what you care about.

And for example, if you’re a blog, and you have individual categories, then maybe really highlighting the categories that you do care about and kind of de-emphasizing the ones that you don’t care about that much. And that can be categories where maybe the competition is really hard, where you have the highest income, or where you just– I don’t know– feel the most opportunity or whatever. It’s essentially up to you.

But making it so that Google doesn’t look at your website and say, oh, there are 500,000 pages here. But rather, they look at it and say, oh, it’s a big website, but this is the really important part. And I’ll focus my energy on here and making sure that everything from here is properly indexed and shown in search.

JENNY GUY: Love that. Good on site navigation. We have some more questions on site navigation and site layout coming up next. We’re going to switch a little bit, change gears to images and video. Lin said, I know Google recommends using images that are at least 1,200 pixels wide. But does Google have a preferred height? Is there any benefit to one orientation over another– square, portrait, landscape, particularly for the SERPs?

JOHN MUELLER: So I think the tricky part here is the recommendations that we have depend a lot on how we show those images. So if you’re just interested in image search, like people search visually for your content, which might be the case for things like recipes or– I don’t know– fashion content, obviously people search in a visual way. For that kind of visual content in image search, it doesn’t matter. It’s essentially whatever you want.

For the content that we show in the rich results– so if you do a normal web search, and you sometimes see images there, which could be like a recipe thumbnail. Or maybe you see other kinds of thumbnails in the search results. Most of the rich result types have recommendations with regards to things like aspect ratio of the images.

So that’s kind of where I would go. Rather than just blindly saying, oh, it must be this size and this aspect ratio, kind of think about, well, is it a recipe? And if so, double-check the recipe guidelines and look at what kind of aspect ratios they recommend there. And then taking that width– I think the 1,200 pixel width is for the large image preview in discover, that recommendation. Kind of taking that width, and then using the appropriate aspect ratios to make sure that your images are appropriate for whatever type of rich result that you want to appear in.

JENNY GUY: Awesome. OK. Julia from Happy Foods Tube. Why is Google cropping the search result images on desktop browsers? It’s really hard to understand what’s in the picture. Is that a glitch?

JOHN MUELLER: It might be. I don’t know. Sometimes things go wrong on our side. So it’s really hard to say just based on this question. Because we have the recommendations for the different aspect ratios, it might be that the images that we’re getting are the wrong aspect ratio. And we have to kind of crop them for that. It might also be that something on our side is going wrong. And I’ve seen that specifically with regards to images in the past where we end up cropping images that we don’t need to crop.

But it’s kind of hard to say without examples. So if you can send me some examples, maybe on Twitter post a screenshot, that’s always super helpful. We pass these on to the team. And the team tries to figure out what went wrong in this particular case. And having a screenshot essentially shows us exactly what you’re seeing. Whereas if people come and say, oh, my images are showing incorrectly. And I can’t give you my site, because I don’t want to tell anyone what my site is, then we can’t do anything with that. We can’t go to the engineering teams, like, oh, some website on the internet has the wrong image aspect ratio. I think–

JENNY GUY: I also have a feeling that Google might know your website, probably knows a lot about it, maybe. And that would be a good thing. That would be a positive thing. You want Google to know about your website.

Another thing that I need to know is when things go wrong, do you have a bat signal in Switzerland that you can push that’s like the Google G that shines out to the team? That is how I picture it happening.

JOHN MUELLER: When something goes wrong and I see it, usually I end up pinging Gary and sending a bunch of emails. But no big light. Where I live, we have a lot of fog in the winter. So having a bat signal wouldn’t be that useful. It’d be– I don’t know. I could see it, but it wouldn’t go very far.

JENNY GUY: Fair. Fine. All right, John. We’ll let that go. OK. Marly from Namely Marly. And we had some questions about this during the live. She said, some people say having original video content on the page is an SEO ranking factor. Does Google consider this to be the case, or is it just a coincidence, because good video may increase dwell time?

JOHN MUELLER: It’s probably more of a coincidence. So from purely a technical point of view, when we look at a web page for web search, we recognize there’s a video there, but we don’t look at the contents of the video. So we wouldn’t know if it’s a Rickroll video or something useful. Essentially, we’d see, oh, there is this video embedded. And we would process the rest of the web page as we normally would. That means we don’t actually look at the content of the video for web search.

For video search, we do look at the video. And I don’t know what exactly video search all looks for, but that’s something that could be taken into account. I suspect the bigger part is really just that you’re providing something of value for your users. And they appreciate that. And they come back. They recommend you to other people. Maybe they send a link your way. And through that providing of high quality information and giving them something useful and unique, you’re kind of building up the value of your website overall.

So it’s not so much that Google recognizes, oh, this is a unique video, therefore, the website is better. It’s more that we recognize all of these signals all around where it’s like, well, people think this website is good. Maybe we should consider it as good as well.

JENNY GUY: Yes. And then a follow up on that, does Google prioritize YouTube over other video players? I suspect I know the answer for this, but still would like to say it out loud to dispel the rumor.

JOHN MUELLER: No, we don’t.

JENNY GUY: Thank you.

JOHN MUELLER: We don’t. And in fact, there’s a lot of work on Google side, within video search in particular, to make sure that, essentially, all video players have the opportunity to appear. I think YouTube is a particularly easy-to-use platform, which makes it really popular. They do a really good job with their landing pages as well. But essentially, it also means that you have a separate YouTube landing page for your videos. And for some sites, that’s OK. You kind of have that double visibility. For other sites, you really want the video only on your website. And for those, maybe you find your own player or find a different solution for hosting that video.

JENNY GUY: Excellent. OK. We’re going to pivot to site structure. I told you I promised you this was coming. It’s coming. So Sandi Gaertner from Fearless Dining said, I often link to the relevant category in my recipe posts. Is there value in this for search? Are category pages relevant for SEO? And how can we improve the user experience for categories? It’s a lot.

JOHN MUELLER: Yeah. I think that makes a lot of sense. What I see sites doing by kind of linking to the categories is giving a little bit more information about the structure of their website and making it a little bit easier for us to recognize, oh, this is a part of the website in this part of the website. And here are some pages that are part of the other part of the website. And kind of understanding how things are connected. What is the context of the individual pages there?

The tricky part is that, depending on how you have that set up on your website, maybe you already have those cross links in place. And by adding another version of the same link, you’re essentially not giving us any more information. So that’s kind of the, I think, the tricky part there.

Because usually in the navigation you’d have things like links to all of the categories. And then we would look at your posts and see, oh, all of the categories are linked up here. So from that post to the categories, that’s probably trickier. Whereas the other way around, that works a lot clearer. You go to one category landing page. And from there, you see a list of all of the posts that are in there. That connection works really well.

And category pages I think are really useful. Because it’s one of those places where, on the one hand, you have a persistent URL. So over time, that page builds up value. Even if individual posts don’t end up getting indexed– kind of like with that first question or a while back. If those posts don’t all end up getting indexed, that category page will still have the snippet of the post, some information about that post maybe, and at least be able to appear for kind of that broader category of search terms.

So that, from my point of view, makes it, I think, super relevant and super helpful for people. If they’re searching for a broader kind of category of things, and you have a category page, it’s a perfect match. So I would definitely make sure that they’re at least indexable. And making sure that you have a clean site structure helps make sure that we understand how things are set up on your website.

JENNY GUY: I’m going to ask a follow up question on that that I did not have. So is there such a thing as too many categories for one website?

JOHN MUELLER: I don’t think so. I think, at least from an SEO point of view, if you’re doing something that works well for users, then probably you’re on the right track. Which I realize a lot of SEOs just want a number. Like, oh, I want 17 categories maximum. But I would really look at it more from a user’s point of view. Is it something where they can still understand, what is the structure of this website?

And usually if users can understand it, then we can figure that out too. I think having too many category pages where maybe you run into hundreds of categories, that just makes it a lot harder for us to understand the overall structure of the website. So instead of having this Christmas tree layout where you have one small part and then more parts kind of going down, you end up having a really flat navigational or flat structure of a website. That makes it really hard for us to understand where individual parts fit in.

So having some kind of a reasonable structure I think makes sense. And reasonable is like– there’s no absolute number for that. For some, it’s like– I don’t know– five categories. Other people have maybe 20. I don’t know. Somewhere in between there I think is probably a good range.

JENNY GUY: Excellent. OK. Another can of worms, here it comes. Do jump buttons help your SEO or impact your rankings?

JOHN MUELLER: I don’t think so. So I mean, jump buttons are basically links that you have on a long page where you jump to a section further down or further up on a page. And from our point of view, those are essentially links to the same page. So they don’t really have kind of this link effect.

And I think the only aspect where sometimes we take that into account is when we have– I don’t know. I don’t know if they have a name. Kind of like special site links, where if you search for something, sometimes you’ll have– within the snippet, you’ll have links to individual parts of that page. Sometimes we can take it into account there. But that’s not a sign that this page will rank better or that it will perform better in search. It’s just, we recognize it’s a long page. You have links to individual sections. Maybe we can show those directly into the search results for your users.

JENNY GUY: Excellent. OK. Melissa asked, is it better to write ultimate guides covering a vast amount of information on a topic or better to write smaller posts that hit specific user intent?

JOHN MUELLER: I will go with the traditional SEO answer. It depends. Because I think it’s something where you kind of have to know your audience. And you have to know what they’re actually interested in. And it might be that you want to target a broader audience, and you have more of like a broad ultimate guide thing. It might be that you’re targeting something very specific, and you want something kind of more specific.

The one thing I would caution from, though, is starting off by doing everything in a very specific way. At least in the beginning when you’re starting off with a website, make sure that you have some kind of broad structure almost that is available. Pages that can rank for a wide variety of queries where you know people are interested enough in that topic to actually make it to your website, to avoid the situation that you create this fantastic article on this one very niche topic, and Google doesn’t notice and sends nobody your way.

So that’s kind of the thing that you want to avoid. And usually by starting off with something broad, you recognize which of these broad topics hits the nerve of the time where people are actually interested, and you have enough people coming. And it’s worthwhile for you to actually create more content in that direction. And based on that, you can kind of refine things and say, well, here’s my broad guide. And here are these five individual kind of more niche specific guides that I also create.

JENNY GUY: We have a whole bunch of stuff about that called parboiling where we use Google Search Console to find the specific topics where people are getting the hits and then creating supplemental content, because people are clearly interested in that.

JOHN MUELLER: Oh, perfect.

JENNY GUY: Yes. I also–

JOHN MUELLER: You have this covered. What do you need me for?

JENNY GUY: Well, no, hearing it from you is– I also have to say that, I think you said that it depends was an SEO answer, which is also a lawyer answer. And that [INAUDIBLE].


JENNY GUY: So the SEO equals lawyer answer. So we’re learning a lot, John. All right.


JENNY GUY: We’re going to go to the general algorithm now. And we’ve got a couple here that are a little hot button and have been asked in the comments. So we have several people saying that there has been an onslaught of websites written by AI tools that tend to be inaccurate and have poor quality content, yet get millions of views. The question is, how does Google view this new AI content?

JOHN MUELLER: Yeah. I don’t know. I think there are two aspects here. On the one hand, I think there’s a bigger fear of this content than there is actually an issue in large cases, where people publish some AI generated content or show some AI generated artwork. And they pick the ones out that are particularly compelling. And based on that, they present it in a way like, well, machines can just take over and create all of this content.

When actually, when you work with a lot of these machine learning algorithms, you realize they create millions of things. And you can find 10 really nice ones in there if you’re kind of really picky and work to figure out what’s happening there. So that, I think, is the one angle. I think over time that will evolve and will probably get better.

The angle that I see happening there is, on the one hand, the lower quality AI generated content, we’ve seen this for years and years. I think, at least before I started at Google, when I created my test sites to try to figure out what is Google’s algorithms doing, I had little scripts that created websites that essentially looked like normal text. But if you read them in detail, you’d recognize, oh, these sentences don’t really fit together, or they don’t make as much sense. And that’s the kind of content we’ve seen for years and years.

And I’d say the kind of– I don’t know– almost like lazy or smart SEOs, they would use this kind of content. And sometimes it works. Sometimes it doesn’t. That part I don’t really see going away.

What I see more is that our algorithms work hard to understand the quality of the piece of content. And that’s something we’ve been working on for a really long time. And ultimately, what will happen is people will create AI generated low quality content. We’ll see, oh, this is low quality content. We don’t have to recognize that it’s AI generated. We just recognize, well, this is kind of gibberish. It doesn’t make a lot of sense. Or it makes a little bit of sense, but it’s not actually useful content.

And we will treat it based on that. We’ll say, well, it’s like– we don’t really know if someone created it who’s just not that fluent in English or if some computer program created it. We just see it’s not great content. So we’re just not going to show that that well.

What I do kind of hope to see happen with all of this AI generated content is more that people use it as a tool. And they use it in a way to create better content on their own. So just like people often have copywriters who double check your content to make sure that it’s written well, more and more there are tools like Grammarly that go off and kind of help you to create a better page. And I think that mixture of having experts on a topic and all of these tools that are smarter and smarter, I think that’ll help us kind of move forward.

And that’s kind of more the direction I see happening where at least the more serious publishers, they’ll go off more in that direction. And obviously, there’ll always be people who are like, oh, I’m super smart. And I don’t want to spend a lot of time on writing something. Therefore, I’ll just have my computer do it for me. And maybe every now and then, it’ll be OK. But overall, I expect compared to someone who’s actually a subject matter expert, who’s working with these tools to create something fantastic, it’s not going to be comparable. And our systems will hopefully be able to pick that up more and more as well.

JENNY GUY: OK, I’m going to ask just, what is the algorithm? I’m just going to go off script. Tell us about what the algorithm is. How many people are working on the algorithm?

JOHN MUELLER: Lots of people. So I think the tricky part is, it’s not like one algorithm that figures everything out. That’s kind of the first thing. And the other part is, it’s not a magic box. It’s not something that looks at the web and understands what this page is about and how it should be shown. There are just lots and lots of really small technical aspects in there.

And that starts off with easy things like understanding where the word boundaries are, understanding what the words actually on a page are, what synonyms are, and figuring out kind of what synonyms evolve over time, and abbreviations and things like that. And all of those kind of small, more technical systems, they kind of come together. And we kind of frame it as the algorithm. But it’s not one long formula that processes everything. It’s just all of these small individual tools that do a small step.

So that’s I think– at least before I joined Google, I always thought like, oh, it’s like Google understands the web and knows what this page is about. And I saw it as some mystical thing. But it’s just a lot of really small, technical, nitty gritty details that anyone kind of would have to do.

JENNY GUY: I’m not going to lie. For me, the algorithm has like a capital A. And it’s maybe kept in a room like a Rosetta Stone or potentially like The Da Vinci Code. I have a lot of those images happening. I don’t know if I’m alone, but that’s how I’m seeing it.

JOHN MUELLER: So we have a tool internally that lets us look at some of the debug information in the search results pages, kind of to pull out the individual algorithms. And when you have 10 results on a search results page, usually the debug output, at least the parts that I see– and I don’t see everything– it’s maybe 100, 200 pages long. So it’s just all of these small systems. They’re all working together, trying to figure out– oh, it’s like this query, actually it means this and this and this.

And starting at the query, I think that was already a big aha moment for me, that we don’t just take what people write, but rather we turn it into a bunch of synonyms. And saying, well, probably this and this and this. And also it looks like they’re looking for something near them. So we should add some idea of the location to that as well. Kind of expanding what people type and turning that into a bunch of really small things and then looking at an index for all of those small things and compiling that back up.

So it’s just– I don’t know, just lots and lots of technical details that are kind of added up in the end. And we have to show them as a small list. And I think what makes it really interesting is we present it in such a simple way that when you look at it, you’re like, oh, there are 10 results here. And there are five images. It’s like, I could have made this page myself. But actually there’s just so much happening behind the scenes.

And all of that complexity hidden away I think makes it– I don’t know. It’s almost like part of the secret sauce there that we hide that complexity. That you look at it, and it’s like, oh, I can deal with this. It’s like a text box. I can just enter something. I don’t have to work with some complex UI to figure it out.

JENNY GUY: Making it look easy.

JOHN MUELLER: Yeah. But it’s really hard.

JENNY GUY: Oh, yeah. There’s a lot. There’s a lot going on. But that’s part of the magic. So I guess it is magic, the magic of lots and lots of people working really hard over the course of many, many years.

JOHN MUELLER: I mean, you see that when you’re creating content as well. It’s very easy to create a guide that is long and long and long, really long thing. And it takes a lot of work to refine that into something that is actually readable, where it’s like, oh, well, it’s a five minute read.

And I think that part of the magic is– when someone first looks at that five minute guide, they’re like, well, five minute guide. I could write that. But they don’t realize actually it’s this giant thing that you started with and you had to refine to that five minute guide.

JENNY GUY: Yep. Love that analogy. OK. We have a couple of questions here. So Norman from Nimble Needles and Laura from Your Guardian Chef ask– I smashed up their questions. We’ve often heard that Google has an interest in showing the user the best possible content. But as bloggers, we often find the same household names in the top three SERPs with content that is either thin, outdated, sometimes even recipes that are wrong. For example, Googling risotto, and the first result shows a best risotto recipe made with sushi rice from Bon Appetit.

Does the Google algorithm not recognize the value and authenticity of food bloggers versus large publications? And furthermore, is acquiring a lot of backlinks the only way to hit those top positions for high volume keywords?

JOHN MUELLER: I think that’s always tricky. Because we do look at individual pages, but we also look at the website overall when it comes to search. And sometimes a website that we think is overall a really fantastic website might have individual things that are not that great. And kind of that mixture of, well, some things are good, some things are bad. Where does that leave us? That’s sometimes really hard to figure out in our systems.

It is sometimes also an advantage, though. Because it means that as a site owner, you don’t have to do everything perfect. You don’t have to do everything exactly the same way as other people do. It’s like, you might get these five things wrong and these 20 things right. And someone else gets those five things wrong and the other 20 things right. And we have to show those sites essentially on the same search results page.

We have to figure out which order to put them in. But we kind of need to balance the good parts and the bad parts and figure that out. And I do think sometimes we get it wrong. And sometimes we show bad results on top.

What usually happens here is that if someone sends something like this our way where it’s a generic query, and we can tell lots of people are running into this, and the top results are really bad, or something really fantastic is hidden away, we can send that to a team internally. And they’ll figure out what went wrong here and try to figure out which part of the algorithm needs to be adjusted. And usually that’s something– they love digging into those details. It’s almost like– I don’t know. They’re looking at the algorithm overall, but they’re actually focusing just on their small parts. And based on that, we try to improve things over time.

And a lot of these cases where we see things go wrong, we tend not to announce, well, we made another change and yet another change. Because ultimately, we make– I don’t know, what is– thousands of changes every year in the search ranking algorithms. And all of these small things that are just adding up. And hopefully, when people try this out over time they’ll see that the results get better.

I think it’s always tricky, because you tend not to notice when things work well, and you notice when things don’t work that well. And you’re like, oh, I found another query where Google got it wrong. It’s like, Google’s always getting it wrong. But we fixed all of the other ones. So it’s like, yeah. It happens.

JENNY GUY: You don’t typically call customer service to tell them how amazing something was or that you had zero problems. So there’s a little bit of attention on that, yes. But that’s true. OK, so we’ve had multiple comments. And the elephant in the room, speaking of algorithm shifts, we had won today, John.

So other than just a couple of people who would like me to say why, tell us what happens with these major algorithm shifts, a core update. And what if your ranks drop overnight? What do you do?

JOHN MUELLER: I think it’s– yeah, I don’t know. I understand the anxiety around it a little bit. Because Google announces a big change. And you don’t really know what that actually means for your business. So you see all of the people externally like, oh, it’s like, everything is breaking. It’s the end of the world.

But we make these bigger changes every now and then, I think every couple of months. And the core algorithm changes are usually based on what we try to figure out with regards to relevance for individual queries. So it’s more about us trying to figure out, what are people actually searching for when they enter this thing? And how can we recognize that this is actually a great result for those queries?

And it’s less about– I don’t know– spam or flagging bad things or anything like that. It’s more about, well, we need to update our systems based on what we see people actually doing and what people actually expect. And that essentially leads to these core algorithm updates. For most updates, we tend not to announce them, because we just make so many of them. But the core updates tend to be something that– at least the individual SEO tracking tools they recognize– and some websites do see stronger changes in the search results based on them. So we decided at some point to announce these, just so that people know that, hey, this is happening. If you see something changing, probably it’s from here.

So that’s kind of our angle there. It’s something where I suspect most sites would not see any changes. So it’s not something that everyone needs to worry about. But it is, I think, something to keep in mind. Google does revisit what they’re doing and what they’re trying to show to users every now and then. And it does try to rethink, what is actually a good experience in the search results for these things?

One of the ways to look at this is to look at the quality raters guide that we have. The quality raters are basically people outside of Google who tell us which version of an algorithm is better or worse. And we essentially send them a collection of search results that we’ve seen and show them one version of search results with a change and one version without a change, or maybe multiple versions. And we ask them, based on the guide that we give them, which of these is better and why?

And in the quality rater guidelines, we have– it’s really long. So I think like 170 pages, something like that. We have a lot of detail on what we think they should be watching out for. And usually, they bring back really useful feedback for the algorithm team to kind of improve that.

So if you want to kind of be ahead of the curve with regards to these bigger changes, if you’re thinking that maybe your site is kind of like on the edge somehow, I’d strongly recommend kind of watching out for these quality rater guides and maybe working your way through them.

There are also a number of SEO bloggers that regularly go through them, say, oh, these things changed. And just following along with them usually gives you a little bit of insight as well.

JENNY GUY: OK, fantastic.

JOHN MUELLER: One of the things– I think with regards to figuring out what to fix, I think it’s something with regards to these core algorithm updates– it’s not something technical usually that’s kind of the matter. So it’s not something that you can just run off and fix and just say, oh, I’ll install another plugin. And that’ll make my website relevant again.

It’s really often a case of, well, maybe you have to rethink what you’re publishing and kind of rethink how you’re presenting that to people. And that’s not something that you can do overnight. Usually that takes a while to figure out.

JENNY GUY: All right, thank you for that. OK. Core Web Vitals. What impact do they actually have on rankings?

JOHN MUELLER: Core Web Vitals. OK. Yeah. So Core Web Vitals are essentially measures that we have for understanding the quality of a page, more or less, with regards to speed, interactiveness, responsiveness. I think those are the three main measures that we look at. And we have kind of a scoring system where you say, well, this is a good metric. This is somewhere in between. This is a bad metric.

And from a ranking point of view, we kind of take everything into account that is not a bad metric. And we say, well, there’s kind of a sliding scale here. This is almost good. Or this is in the good area completely. And based on that, we do apply that as a ranking factor.

But it’s a small factor. It’s not going to catapult your site to number one. It’s something that, for the most part, people would see in areas where there’s a lot of competition, where there’s no clear number one answer. So for example, if someone is looking for your company name, then your company’s website is kind of– it’s like the clear answer. Or if you’re a local business, and someone is looking for a pizzeria, and you’re a local pizzeria, it’s like you are kind of the clear answer for that query.

But if someone is looking for– I don’t know– a turkey recipe, and there are five million versions. That’s something where we might say, well, all of these are pretty good turkey recipes, at least as far as we can tell from a technical point of view. Googlebot is not going to try the recipes. But we can tell these are good pages–

JENNY GUY: Googlebot is–

JOHN MUELLER: –reasonable pages. And then we might figure out, well, this one has a better score here. So maybe we should show it a little bit higher.

JENNY GUY: I was going to say the Googlebot is probably vegan. So can’t eat turkey. Have some tofurky or something like that for the Googlebot.

JOHN MUELLER: I don’t know. I don’t know if vegan recipes would work better. I think that’s kind of– I don’t know. Now that I think about it, it is kind of this weird thing where even from a technical point of view if we could understand English well enough to read a recipe, understanding that this is actually a good recipe or not is probably not something that computers will be able to do, at least not in the near future.

JENNY GUY: So breathe easier all of our lifestyle bloggers. Computers still cannot tell you if something is yummy. So you are a very necessary part of this process, for many reasons, but that one specifically.

I wanted to– so we’re almost out of time. And as I said, I wish we could stay all day. But I wanted to ask your opinion on SEO plugins in general. Is there something that– is there one that everyone should definitely have? Is there something that could solve a lot of issues for people with regard to search?

JOHN MUELLER: I don’t know. So I think on the one hand, we try not to make recommendations with regards to different third party providers. So we wouldn’t want to say this one is better than the other one. I do see a lot of SEO plugins making your life a lot easier. So from that point of view, I’d say it’s probably worthwhile to look into the options.

I think with all of the technical things around SEO, they can do only so much. But if they can take some of the technical details out of your hands and take care of it for you, then sometimes that’s just worthwhile from a time point of view. So that’s kind of the angle I would take there, is look at what you want to do and think about the plugins that do it for you, so that you don’t have to spend time on it. And make sure that whatever you pick is something that is well-supported, where they regularly do updates, where they help you if something goes wrong.

And the other aspect I would watch out for with plugins in general is try to minimize them. Because every plugin adds a little bit of latency usually with regards to websites, makes things a little bit slower, adds a little bit of room for maybe security issues along the way. And the more you can concentrate into maybe one really strong plugin that is really well supported, I think that can make things a little bit easier.

So instead of going out and saying, well, there are five SEO plugins. I’ll install them all. Kind of pick what you really need and just install that.

JENNY GUY: So with that being said, I want to make sure I drill in on it. Page speed is important. [LAUGHS] Basically what you’re saying is page speed is important.

JOHN MUELLER: Yeah. I mean, it’s important for Google. But I think for the bigger aspect is actually for your users, where if you have a page that is really slow, they’re going to think twice about clicking around and browsing around on that website. Whereas if you have a page that is really fast and snappy, then of course they’ll be like, oh, maybe this article isn’t what I was looking for, but actually it looks like there’s other stuff here that is what I’m looking for.

And they’ll go and browse around until they find something that they actually like. And they’ll come back. Whereas if something is really slow, then all of those things kind of fall away. It makes it a lot harder.

JENNY GUY: It’s the Back button rule. How long would you personally sit there and wait before you click the Back button and go back to the search results.

JOHN MUELLER: Yeah. And especially when it’s about something where there’s almost like– I don’t know– commodity content and that. Like with recipes, there are lots of different recipes. And you find one recipe you click on, and it’s a really annoying website. You just go back and look for different recipes.

There are different variations. Maybe the other recipe isn’t as tasty. But it loads fast. You can read it easily. It’s like, that’s worth a lot too.

JENNY GUY: Totally agree. OK. I wanted to circle back on the question we had talked about with some of the big legacy sites or big name sites having the great results in the SERPs. I’d asked about backlinks. How important are backlinks to a content creation strategy?

JOHN MUELLER: I don’t know. It’s always a tricky question, because when we say we do use links in our systems, people go, oh, well, if you use five links, then I’ll buy 500 links. And then I’ll be on the good side. And that’s really not what we want people to do.

So we do use links. We use them and in some ways to understand the overall context of a website. But for the most part, it’s not something that you need to go off and kind of do these fancy schemes to get all of those links.

One way I like to think about this is, of all the teams that we have working on Google, as far as I know there’s only one team that works on understanding links better. And that’s the web spam team. And all of the other search quality teams, they’re kind of working on kind of understanding your content better, understanding the context of your pages better, how we can show that better in the search results. I don’t think there’s any team that’s working on, how can we use links more with regards to search results.

So there is I think that one aspect, especially for a new website of– well, it has to be found. We have to understand what the context is. But at some point, we found your website. We’re crawling it regularly. It’s not something where you need to regularly go off and buy links or organize links or do anything to keep getting them.

I think it’s super important that you make it easy for people to link to your website. So if they want to recommend it to other people, make it as easy as possible. But that doesn’t have to be a link that Google actually sees. Those recommendations can be make it easy for people to share on WhatsApp or kind of all of these other networks where people recommend things to other people.

And it’s not something that we would necessarily see. So we wouldn’t use that for search. But it’s kind of the same thing as you would do for search.

JENNY GUY: Awesome. OK, well, we are almost out of time. Right before we go, I wanted to ask, is there a specific article or two that you would highly recommend to our content creators who are wanting to perform better in SEO and in the SERPs to go read that is Google content?

JOHN MUELLER: I would actually point at the blog post we have on core updates, not because it’s timely or anything, but because it includes a lot of details on things that content creators usually try to suppress. So things like if someone new comes to your website, would you give them your credit card? And those are the kind of things where you would probably want to go to someone who’s unassociated with your website and ask them directly.

It’s kind of that– I don’t know– honest advice, where when you look at your website, you’re like, of course my website is trustworthy. It’s the best website ever. But if you present it to someone who doesn’t know your business, that kind of honest advice is really priceless. And the blog post on core updates has a bunch of those questions in there where you can really kind of go to someone who’s not associated with your website and ask them these questions.

And maybe present different websites from your competitors. Ask the same questions. And then really get that honest feedback of, well, your website looks outdated. Or the design is not nice. It’s not pleasant. Where you’d say, well, probably like an SEO aspect– it’s not really directly an issue. But indirectly, it is. It is something that people look at and they think about.

JENNY GUY: That is great. We are adding that link. So we are handing you guys– everyone, we have our one-pager that we always have that has all of the links that John has been talking about, including some of the links that I mentioned for Mediavine. We’re going to stick it in there. It’s going to be everywhere. You don’t have to worry about it.

John, this has been so much fun and so amazing. Would you come back in 2022? Is it possible?

JOHN MUELLER: Sure, sure.

JENNY GUY: That would be wonderful. Thank you so much for coming. Thank you guys for watching. I wanted to let you guys know that on the next episode of Teal Talk, it is Tuesday, December 14. We’re going to get through the Thanksgiving holiday. We’re having Paul Bakaus of Creator Relations at Google. He’s the head of it. We’re going to talk about Google for content creators. We’re going to talk Web Stories, Discover, and how to navigate all of those resources that they provide for content creators.

It’s going to be an amazing month. And we could not be more grateful. This has been all that we could have dreamed of and more. So thank you, John.

JOHN MUELLER: Thank you. Thanks for having me.

Related Posts

Publishers First in the Privacy Sandbox

Publishers First in the Privacy Sandbox

4 min read

When Google finally kicked off 1% of cookie deprecation at the start of 2024, this opened the door for real-time auction testing of Google’s Privacy Sandbox. Thus far, the industry …

Read More