 All right, welcome everyone to today's Google Webmaster Central Office Hours Hangouts. My name is John Mueller. I am a webmaster trends analyst here at Google in Switzerland. And part of what we do are these webmaster office hours hangouts, where webmasters and publishers can jump in and ask anything that's on their mind with regards to search and their website. As always, a bunch of questions were submitted. But if any of you want to get started, then feel free to jump on in now. Hi, John. I want a quick question. Yeah, so one of our clients in Michigan, they are actually working with a lot of products. And particularly, there are similar products. In terms of, for instance, we have one product of CBD water. And then we have one page of product for a pack of bottles. So the content is like somehow they are relevant. So what we are actually trying to focus is on the single product page. So is it a good approach to have an internal link with particular the keyword we are targeting for that particular product page for the caller to understand, or do we have to actually introduce canonical tag, which is going to be for our, I would say, primary page which we are trying to rank. So what will be, are the both things going to work? Or what do you recommend? I think both approaches should work. So ideally, we would be able to crawl your website and find those pages anyway. So that's, I think, almost a strong recommendation that we would always have. And pass that on the individual pages if you have different variations and you want one version to be kind of the main version, then you can still use the rel canonical to tell us about that. OK, so as far as my understanding goes, the internal linking thing will also work. Yeah. All right, perfect, thanks. Sure. All right, any other questions before we get started? Yes, I have a question. So I'm fighting with hijackers from maybe 20 days now. It's a long story. Maybe each day I have new information, so I try something new. And my idea is that I need to find a solution, because unfortunately it seems like they successfully pull Googlebots that a bunch of projects are not actually ours. So I have a solution here, which one programmer have advised me to try. And it's about implementing JavaScript in the footer that will say that if this content is represented to a domain name, which is not my domain name, so the content of the domain name of the hijacker, the meta tag will be served as no index, no follow. So basically this is somehow cloaking, but maybe white hat cloaking in a way that it does prevent someone from showing the content there. I've also did a very aggressive hotlink protection, so it does protect from taking our CSS, JS, or images, et cetera. And the Google cache is actually seeing the domain name. So when you check the cache for the hijacker, you see our domain name there. And that's the case for 20 other websites as well. And will some kind of cloaking like that work because I've seen other similar approaches. So when they find that, for example, someone is represented as a Google bot and the IP of the Google bot doesn't resolve to your servers, for example, means that that's the fact Google bot and probably it's used for scraping, et cetera. And then we can do a no index if we see that such kind of IP is that kind of white hat cloaking good for something that you will understand why we do it or it's not a good approach. In general, that's something you can do. So if you have other clients that you don't want to serve, you can serve them any other content. That's completely fine. The JavaScript, I think, is an interesting approach as well. I think that could help, especially since we do process the JavaScript. And from my point of view, that wouldn't be cloaking because you're just serving the content that's matching the URLs. So on your URLs, you have the full content. On the other URLs, with the JavaScript, you're turning that into no index. That's, from my point of view, that's perfectly fine. I don't know if it would work for these sites because sometimes the proxy sites, they remove some of the content from the pages. But if that works for those proxy sites, go for it. And a very quick follow-up question. If we try to remove the URLs, which are actually serving 404 to regular visitors, this is the URLs of the hijacker through Google Removal Tool. And it's a lot of work because we have to basically do this for 1,000 URLs, et cetera, one by one. And I'm trying to build a macro, basically, to do that repetitive job. I know this tool is not meant for that. But do these URLs actually hold power when they are removed from the service like that or just the visible part? Will they still have some page rank or some page flow that they can redirect to some other place and use that? So the URL removal tool would only hide them in the search results. It wouldn't change anything in indexing. But sometimes hiding them is enough for what you need to do. So that might be an option. I think there is also a Chrome plugin that does this for you. It works only on the search console part. It doesn't work on the public removal tool. OK. I mean, from our point of view, these tools aren't meant to be used at scale. But if that's something that you need to do and you can do it like that, I don't know. Cool. OK. One more question maybe before we get started with the submitted ones. Is there anything else on your mind? So John? Hi, here. Hi, how are you? Hi. So I have to say, I'm a little bit at my wit's end. I'm not sure what to do. If you search Angelina Jolie, for instance, since we have an entertainment site in the US right now, the fifth result is a story from Australia. It's a completely fictitious story about Angelina Jolie marrying an unnamed billionaire with details about her fictitious wedding. Frankly, you don't need to be a journalist to see that this is not true. It's completely made up. The sixth result is a story from the International Business Times in the edition, which completely rewrote my site's story linked to us. But again, the problem I'm having here is that, like, one, dealing with Angelina Jolie's people all day long, I know what's true and what's not true. The second part is that we're the only site that has the effect checks, and we're practically invisible now. Sometimes it's because fake stories are there. Sometimes people are taking our stories and knock us out of search. And I'm trying to figure out, at this point, what do we do? I mean, there's no question we're the only entertainment site that's transparent that lists its sources. It's not a quality thing. We've gone through everything with no thin content. I mean, we have devoted every resource. But we're at the end of the rope here, finally. I don't know. I'd need to double-check these queries. So I can kind of, we're just probably just on the side. So at least I have it there. But I think you can also send it to me directly when you see these things happening. It's always a tricky to double-check them from here, because I don't see the clean US results. So it's sometimes tricky for me to see what exactly is happening. I just don't want to inundate you constantly with things. But you know, there is, well, I feel bad about it. Because I hate to say it, but I have to say that I feel like we have gone from using the feedback section to being helpful to being annoying. And I'm not sure how Google has reacted to that, frankly. So I try to do it sparingly. But we're at a place now where this is untenable. OK, yeah. I'm happy to take that up with the team here. OK, I'll send you a screen grab right now. Thank you. Cool, fantastic. All right. And with that, let's jump into some of the submitted questions. Regarding the last big core update, how soon can we see site-wide changes after we fix all of the content that didn't satisfy the intent? Is that three months, or is that longer? So in general, the bigger updates that we do around Search like this core update, it's not a matter of us looking at sites and saying, this is bad and this needs to be fixed. But rather, us looking at these sites and saying, well, maybe this isn't as relevant as it used to be. So you can, of course, as a webmaster, take that and say, well, I need to fix things to make my site more relevant. But it's not, from our point of view, is something where we'd say something is broken and you just need to fix these five lines and then it'll be back to normal. But essentially, a matter of how can you show that you're relevant for these kind of queries? And these are changes that take quite a bit of time for algorithms to figure out. Like on the one hand, in general, if you're talking about a whole site, understanding how a whole site is relevant is one thing. Understanding how individual pages are relevant is one thing. All of these things take time. And there's no fixed timeline for that. It's essentially an ongoing process where we review the pages and we reprocess them, we re-index them, we reassign the signals that we have there. And this is kind of ongoing and ongoing. It's not that there's a fixed timeline, but we'd say, at this point, everything should be updated. You recently said if a web page is directly connected from the home page, it gets priority in crawling and indexing. Is that because the home page has the highest authority, the links from other sites? Or does Google always give more priority to the home page? There's a little bit of noise somewhere. Let me just mute all of you. So we don't have any artificial, like the home page is the most important page signal. That's essentially just kind of an effect that a lot of sites see, where we recognize that the home page is very important here. And a lot of times, that's within the website, that's also very clear, and that the home page is very visibly linked. Also, externally, we see that people link to various parts of the website, including a lot of times the home page. So all of these signals kind of add up. And in general, the home page is one place that we think is very relevant for a site. But it's not the case that we would artificially say the home page is the one place where everything needs to be linked in order for it to be findable. So that's kind of something that is more anecdotally the case rather than a fixed rule that we have in place. I'm a news publisher. On my site, I added a news article page linking with internal links pointing to several transparency sections. These links have the appropriate markup. These links are on every news page. So I applied a nofollow to that. However, I'm worried if applying the rel nofollow or follow prevents Google from understanding these pages, essentially, that's fine. You can do that. You can use a nofollow for links like that. For the most cases, you don't need to do that. So in general, we're pretty good at recognizing what kind of pages have which contexts within a website. So things like your terms of service or your contact page, all of those detailed pages that you have that are essentially linked across the whole website. These are things we're pretty good at recognizing what the natural structure of a website is and how they belong in there. So it's not the case that you would need to nofollow any links to the default pages that every website has or that your website uses in a lot of cases. So in this case, transparency section, if that's something that you're linking from all of your articles, that's perfectly fine. That's no problem. It's also fine to say, well, I want to use the nofollow for individual links here. From our point of view, I suspect the effect will be minimal to not visible at all. So a lot of times you're spending energy and time trying to set up this kind of nofollow thing for your internal pages when actually the effect is essentially nothing. So personally, I'd recommend just skipping that part and saying, well, these are normal links within my website. These are natural links that I have there. These are links to pages that I think are relevant for these individual pages. So I'll just link them normally. And by linking normally, you also don't have to worry about long-term maintenance, which links are nofollow, which links are followed. Why do I have them like this here? And if someone else takes on the website or works on the front end, will they do the nofollow thing as I intended them, you don't have to worry about that. If they're normal links, they're just normal links. I keep it simple. That's the best way to make sure that search engines have a chance to understand your pages and how they work together. I asked them previously out. I've been seeing image files show up as soft 404s in the new index coverage report. Why would that be? So from my point of view, that's perfectly normal. We show these as soft 404s because for web search, they're not suitable as landing pages. So in particular, for web search, we want to have HTML pages that we can link to. For image search, we have to have that combination of image plus landing page. But for web search, we need to have an HTML landing page. So if you have a URL, that's actually an image. And we think, oh, maybe we should crawl it for web search. And we notice it's an image. Then we'll say, well, it returns 200, but actually, we shouldn't index it. Therefore, we can just call it a soft 404. So that's probably what you're seeing here. For the most cases, we can recognize that something is an image just by looking at the extension of the URL. But there are lots of cases where sites use kind of non-standard URLs for image files. And from our point of view, that's perfectly fine. But we might pick them up as links somewhere and try to crawl them as web pages. And then we'll flag them and say, well, actually, these are not web pages. And you don't need to do anything special there. If they're not web pages, you don't need to have them indexed in web search. For the above-the-fold algorithm, it has historically been based on desktop pages, or were mobile pages evaluated there as well. And now with mobile-first indexing, will the above-the-fold apply to mobile pages once a site is moved to mobile-first indexing? I don't know. I actually contacted the team about this just recently. So maybe I'll have more on that in the future. I suspect for the most cases, it'll be similar in that the above-the-fold content will either be normal content, or it will be a bunch of different ads. But I don't know specifically what we'll be looking at here, or if there's something specific that we could even point out here. In the image guidelines, it says a null alt attribute is bad. However, on the W3C Web Accessibility Dock, it says a null value is fine. Do you have any comments on this conflict? I didn't check the W3C Web Accessibility Docks. I don't know exactly what the connection is there. From our point of view, we use the alt text as a very strong signal for image search. So that really helps us to better understand what these images are about and how we should show them in image search results. So if there are images that you want to have indexed, then I would take advantage of all of these signals and make sure to supply content there that's useful for us, for image search, as well as for users who might not have images enabled or who might be using a screen reader, for example. So that's kind of what I would aim for there. Obviously, if you have images that are more decorative, that are more placeholders, where you're just kind of like shifting text around by pixel by pixel, and they're not really meant for image indexing, then you don't need to put an alt text on those. Those are obviously images. They're not meant to rank in Google Images, so you don't really need to worry about what you're providing there. But again, if you do want these images shown in Google Images, then make sure you take advantage of kind of the different ways of giving us information about the image. If a website in India blocks other countries, that's a bad user experience. But what about a website in the US only allowing the US and blocking other countries? What's up with that? So I mean, that's also a bad user experience, so I think that's pretty obvious to anyone who has run across these kind of situations. It's not the case that I would say one is a good user experience and one is a bad one. The main difference there is mostly from a practical point of view from our side in that we crawl the web primarily from the US. So the Googlebot, the IP address of the Googlebot is generally geo-positioned to the US somewhere. So that's not something that we do on our side, but rather it's something that these geo-IP databases do, where they're saying, oh, well, these Google data center IP addresses, they're probably from the US. So that's generally how things are seen, and that Googlebot is primarily crawling from the US when websites look at the IP address. And based on that, if you're blocking users in the US, then Googlebot will be blocked as well. Whereas if you're blocking users outside of the US, then Googlebot might still have access to that content if you're not blocking users in the US. So it's not the case that we would say this is good or this is bad, but purely from a practical point of view, when we crawl from the US, if Googlebot can't see the content because it's blocked, then we can't index it. Whereas if Googlebot can see the content and it's not being blocked, then we can index it. And the difficulty, of course, with content that's allowed in the US or allowed for crawling and accessing in the US but blocked in other countries is that the users in those other countries will probably get upset and be unhappy. It'll be a bad user experience for them. And I suspect in the long term, that's something that those websites will kind of struggle with, where they might say, well, we never have any users from Europe accessing our website. Therefore, maybe we'll never, I don't know, have any subscription model for users in Europe. And that might be because they're blocking users in Europe. And then, obviously, users from Europe aren't going to go to that website. They're not going to recommend that website. It's going to be kind of, I don't know, sad for those users. And in the long run, probably suboptimal for the website in general. There's always talk about, well, Googlebot should just crawl from all other countries. Therefore, figure out which countries are available and which ones aren't. And I think purely from a practical point of view, that wouldn't work. On the one hand, from our side, having resources to crawl from every country, that's a lot of work. On the other hand, from your side, from the webmaster's side, if we were to crawl from, I don't know, was it 200 countries, that would mean we would create 200 times a load on your server. And probably that's not something that you would find very cool. So that's probably something that wouldn't make that much sense from your side, either. So I suspect in the long run, we'll find other approaches that might work well here. But at least for the time being, if Googlebot is blocked because you're blocking the country where Googlebot is currently crawling from, then that content can't be indexed. And it's not a matter of good user experience, bad user experience, but just purely, if we can't access it, we can't index it. How do you correct? Yeah, go for it. Yeah, so one of our clients in Netherlands, they are facing a problem with our, for instance, we were working on server-side rendering because whenever we were sending the URLs for indexing, so we were seeing the header and the content was showing a 404 because the crawler cannot access that content. So we fixed that issue with the server-side rendering. And now they are in a state where the page is cached properly. And I can see it in Google. But whenever I see the last time when it was like when I see like ViewService or when I see like what was the last date Google crawled that particular content, I'm seeing a 404. I can actually share the URL here. So we did all the fixes. And it was quite weird when you actually go to the Chrome and you inspect the element and you actually disable JavaScript for Google. And you go for checking the index, you can see the page properly. So I just want to show like, is the crawler seeing all the content or is there something we need to do? So you mean the cache page is returning a 404? Yes. So this is the link for the web cache for Google and it's showing a 404. So the particular URL. That can be completely normal. So that's more of an internal thing on our side rather than a sign that anything with crawling or indexing on your side is broken. In particular, sites that have shifted over to mobile-first indexing, they tend not to have a cache page at the moment. So that's possibly what you're seeing there. And it doesn't mean that we're not crawling or that we're not indexing these pages. It's just the cache page is not available. So what I can see is like how can I make sure of like, I'm delivering all the content for the crawler instead of it's not just getting the header and the footer? You can use the Fetch as Google tool in Search Console. That returns the static HTML response. So you can double check that that response matches what you're actually providing. Yeah, that actually has all the content inside. So it's cool. That's cool. Yeah, then you're all set. Sounds good. All right. Does the KML sitemap truly affect Google business listing ranking? So first off, I don't know how the Google My Business kind of maps listings are ranked. So I can't really tell you there. But we don't have KML sitemaps anymore. That's something that I think we turned off, I don't know, most of it maybe eight or nine years ago. So that was available away in the beginning. But these GeoSitemap files are something that we haven't used in a really long time. And KML files, in general, as you would use them for maps or for Google Earth, as far as I know, we don't use those for anything related to Search. But it might be that something on the map side is doing something differently there. But at least from my point of view, we don't use those. Does Google accept the mobile rel alternate links returned in the response header? And no, as far as I know, we don't use those in the header. We only have those in the head section of the page. But that should be clearly documented on the mobile-friendly developer docs that we have. It should be pretty clear there where those rel alternate links need to be. And ideally, of course, if you have a responsive design that you don't need to worry about the alternate URLs. What's your advice on syndicated content and hreflang? I have a network of sites, a US version and a UK version, same brand, and also have another US version that's a different brand. I want to syndicate site C to site B and have hreflang for the UK and the US markets. That's something you can do. In general, what you need to watch out for is that you have one clear canonical for each of the individual country versions. So if this is all English content and one is for the UK and one is for the US, then make sure that all of your US versions have the rel canonical pointing to your preferred US version. And all of the UK versions have the rel canonical set to the preferred UK version. And between the canonical UK and the canonical US version, you can have the hreflang links between them. So that's what I would recommend doing there. Make sure you don't have a canonical between the different language or country versions. So don't have a canonical from a UK version to a US page. Because what will happen then is we will just keep that one version. So make sure we can have a clear canonical per country and then the hreflang link between those versions. One tricky aspect here that might come into play is that we still might see these pages as being essentially identical. So that's something kind of to keep in mind in that for the most part, we will use the hreflang attribute to say, well, these are four different markets, so we'll show them differently. However, if the pages are really essentially the same, it doesn't make sense to index them separately. And we'll index them as one version. And we'll try to use the hreflang attributes to show the proper URL in search, but we'll index just one version. Does the amount of ad space on a site in a certain vertical directly impact ranking? So for example, should a help site or a nonprofit site generally have fewer ads than, say, an e-commerce site? I don't think we have any guidelines in that regard. So it's not as far as I know in our algorithms, there's nothing from the ranking side that would say, well, this is an e-commerce site. Therefore, it can have 10% ads. That's something that I could imagine if you're looking at the space overall, that you might find comparable things across different sites, but that's not something that would be coded into our algorithms. That's more something where probably indirectly the user experience is reflected in ranking in that people might like this one site because it's really fast, really sleek, and focused on the content rather than providing a lot of ads, but we're ranking it based on signals that we pick up indirectly, like people recommending that site. So that might be something where the direct connection is more accidental rather than on design. Does Google Local Business have any impact on local SEO ranking in search results? As far as I know, that's completely separate. We have kind of that Maps one box thing that we show when we think that someone is looking for a local business and we can tell that we have some information that we can show them there, but the normal organic results are essentially normal organic results. Does the click-through rate on an internal or external link increase relevancy or ranking of the URL they are pointing to? And no, on the one hand, I don't think we'd be able to measure that. On the other hand, that's a very noisy signal and very hard to even figure out what that could mean if people are clicking on one link. Does it mean it's a good site? Does it mean maybe it's a good ad? I think that's too hard to turn into anything useful from a ranking point of view. Is it bad if the canonical refers to a different URL or domain than the self-referencing hreflang? For example, the canonical of an AT URL points to the DE URL because of duplicate content, but hreflang for the DE AT version points to the AT URL. So essentially, you have two copies of the same thing and you want to have one version index, but you want to use hreflang to tell us about the different versions. I suspect what will happen here is we will index the canonical version and there's a chance that we can pick up the hreflang and use that, but it wouldn't say it's not guaranteed. I mean, hreflang is never guaranteed, but it's probably even trickier than a normal situation. So what you will definitely see in a case like this is we will only index the canonical version. Swapping out the URL is kind of something that we could do separately. So the visible effect here could be that if your page's title is something like, I don't know, best running shoes in Austria, and we indexed the Germany version of that, then the page's title might be best running shoes in Germany. And in the search results in Austria, we might show best running shoes in Germany and show the AT URL pointing to your Austrian site. So that could be a little bit confusing. On the other hand, if you're aware of this situation from ahead of time, then maybe you can just change your title to be best running shoes. And that way, it doesn't matter which kind of country version that we pick to show in the search results, that title will just work. So in general, these kind of situations where you have duplicated content available for multiple countries are really tricky to handle on our side because we really work hard to try to decrease the load on your side with regards to crawling and indexing, as well as to simplify things on our side when it comes to indexing. So if we can recognize that things are duplicate, then we'll try to fold that together. And we'll try to use something like the hreflang to kind of pull those apart again. But that's really tricky after we folded them together already. So if you really want to make sure that pages are indexed separately for individual countries, you should really make sure that there's actually really strongly unique content on these pages so that we don't kind of fall into this situation when we think, well, actually the content is the same. Why should we kind of cause all this extra fuss for the webmaster and for users by indexing that it's separately? So make it really clear what you want us to do. A lot of our competitors are using breadcrumbs, but in the back end, you can't see them anywhere on the front end. Is this considered cloaking? Even if Google shows your website structure hierarchy in breadcrumbs and search results, I don't know. I've never seen that happening in practice where people are cloaking breadcrumb markup. I think that's probably from kind of the, I don't know, return on the effort that you're putting into that. That's probably way too much. For the most part, making breadcrumbs visible on a page is not really a lot of work, and it helps users, too. So if you're in this situation where I really want to have breadcrumbs, then I'll just put them on the page. I don't think that's something that you really need to put cloaking or use cloaking techniques for. And similarly, I could imagine if you report this to the web spam team, they'll go like, cloaking breadcrumbs is like there's real hardcore spam out there to fight. I don't know if it's worth our time to manually take action on a site that is trying to sneak some breadcrumbs in to Google Search. So that's something where if I were in your situation where you were seeing this, I would just put breadcrumbs on the pages if you think that they provide value and not worry so much about what your competitors are doing in a case like this. In WordPress, usually the author puts their website or portfolio links. In our recent crawls, we found a lot of inbound links for a blog with few active authors. Should we remove those URLs? I don't know how you mean that here. In general, if you find links to your site and the reasonable links, even if you think that the quality of that website that's linking to you isn't fantastic, I wouldn't worry about it. I will just leave those links be and not worry about putting them into a disavow file or anything like that. John, just contributing in the same question, I would say we were facing the same situation like a couple of months ago. So we were actually blogging on WordPress and what we were facing was whenever we ran the site audits, we were seeing there were a lot of outbound links for some websites. For instance, even for me, I would say when I was blogging and I was just replying to people commenting, so when people click on my name, they will be redirected not to that particular website but to my website. So they were technically like more than 40 links pointing the same URL. So I don't know, that's a nice question. Like, how should we tackle this? Is it like? That sounds completely normal. I wouldn't worry about that. I think that's completely natural. I mean, if you're just dropping links on other people's blogs, that's something else. But if these are normal links to your website for normal conversations, that's completely natural link. That's not something I would worry about. OK. What's the difference between Search Console latest links and sample links? Then why does Google show different tabs? So I don't know how this is treated in the new Search Console. I forgot to double check there. But with regards to the old Search Console, the main difference there is that if you have a lot of links to your website, then the sample links will be a sample of all of the links, whereas the latest links will be primarily the most recent links that we found. So you can imagine the sample might be spread across over time, whereas the most recent ones will be just kind of the most recent ones there. If you don't have enough links that we would show in Search Console, then probably the sample will match the latest as well. So that's for a lot of sites that will have a strong overlap. Is it sufficient to tag paid articles, editorials with no follow, or should we tag both no index and no follow? The important part is really the links on those pages from our point of view. So if this is a paid article, essentially, it's an advertisement. So you need to make sure that those links are not passing patron, which you can do with a no follow. With regards to no index, that's totally up to you. If you want to have these pages indexed as a part of your website, then we can index them as a part of your website. Is it an issue if the Cache page in Google returns 404 when all pages, all tools report that Google can crawl and index the page OK? I think we touched on this briefly before. And yes, that's perfectly fine. I think it would be really nice to get this resolved, especially with regards to mobile-first indexing, where we're seeing that more and more. But at least at the moment, that's not something that you need to worry about. The Cache page is handled completely separate from the normal crawling indexing and ranking. So just because you don't have a Cache page for individual URLs doesn't mean that there's anything bad otherwise for your website in search. Is the meta author tag use? No, we don't use that at all. That used to be used for the authorship markup. I don't know if it was meta author or just like the link real author that you had there. But at the moment, I think for many years now, that's not used at all. If ads are not clearly differentiated from the main content, does that directly impact rankings? For example, if a text ad is the same color as the main content or does not have a different color, is that negative for ranking or not? Not necessarily. So what could happen there is that we see the ad as a part of the content. And if you want to have an index like that, that's kind of up to you. Some sites serve the ads directly in line, so we would pick that up as a part of the content. But that also means that your page might rank for something that's in an ad. So depending on how you deal with ads, that might be like no problem, because maybe all of these ads match fairly well what you provide on your pages. Or it could be kind of weird where maybe the ads are a completely different topic to the topic of your pages. And suddenly, your pages are ranking for those topics as well. In general, I'd recommend blocking the ads, definitely from passing page ranks, so that any links that you have in the ads, make sure that they're no follow. Or maybe using something like JavaScript and blocking those by robots text to prevent those from appearing at all. But ultimately, that's up to you. And sometimes that depends on the infrastructure that you have available where you can easily kind of make them in line, or you make them in line because of speed reasons or other reasons, or maybe you can't put them in line because you're using JavaScripts or iframes or some other technology that's totally up to you. John, one question. Is the pruning process in SEO really, really helpful, I would say? I was reading a lot about that. And I would say to start that thing, it will be really time consuming. So is it worth it in terms of SEO or making? So by pruning, you mean removing pages that you don't need anymore? I would say the pages which are not performing, anyhow, they're not getting any traffic. And the content there, I would say it's research content. But I would say they are not adding any kind of value for the users. Yeah, I think there are kind of two opinions and approaches there, even internally within Google, where when we talk with the search ranking leads on this topic, a lot of times they say, well, maybe they should just improve the content. If they have this content on the website, and it was initially there for a reason, and maybe it's not great content, maybe it's even bad content, but one approach is really just to say, OK, we will spend time to improve this content. And the other approach might be from a practical point of view where you say, well, I know this content is there and I put it out there for a reason, but it's really terrible content and I don't have time to kind of improve this. I don't have time to focus on this. Then maybe removing it is a good idea. Ultimately, it's something where the content that you have available on your website is how you present yourself to search engines. So if you're aware that this content is bad or low quality or thin, then that's still how you're presenting yourself to search engines. And you can say, I can handle that by removing the bad content or I can handle that by improving the bad content. And both are appropriate responses that you could take. Sometimes there are practical reasons to go one way or the other way. For example, if you have millions and millions of pages that are really thin content, maybe it's not practical to improve all of those. And maybe it's something where you say, well, in the long run I'll make sure that my new content is good and then you take all of those out. Or it might be that you try to find a middle ground and say, well, I'll improve some, but I'll also take a bunch out that I don't have time for or that I don't want to have it at all on my website anymore. Can I add something on that subject? Sure. For example, WebMD has a lot of articles which are very important, but not many people search for that topic at the moment. So there was an example from an SEO who said that maybe after five years some of these pages that are not getting any kind of visitors may include some very important information about a disease or a solution or something like that. So in that matter, what do you consider a terrible content, a content which is not getting a lot of visitors because they do not search for that topic at the moment? Yeah, I think that's a really important point. A page could be very, very important in terms of information there, but maybe it's not something trendy at the moment. Should I remove such kind of content? Or maybe I should wait in the time that we will become a relevance to the time frame. Yeah, that's a really important point, I think. Totally forgot to mention that. I would not use a metric like PageViews as the only way of recognizing low-quality content. You're kind of the expert of your website and you know what's good and what's bad. And sometimes a metric like PageViews helps you to find low-quality content, but I would not blindly say everything that gets a few PageViews is bad content and I need to remove it. So our algorithms don't look at the number of PageViews. They try to understand the value of the content instead. So just because it's rarely used doesn't mean it's a bad piece of content. OK, let's see. Why might a page be considered a soft 404? They started a thread with a bunch of examples. I took some of those examples and passed them on to routine, so hopefully I'll have some input there and I can post back in the forum thread for you there. For the most part, we do try to recognize soft 404s as really the situation where a page is returning a 200 OK result code, but actually the content is saying this page doesn't exist anymore. And sometimes that's really tricky to diagnose because sometimes there are also temporary issues on a server where maybe the server is returning like I'm overloaded. I can't serve your response. And they're serving that with a 200 and we say, well, that's kind of like a soft 404. And the next time you look at that page, it's like there is a full content and it looks like a normal page. So sometimes this kind of fluctuation happens on a server sign. But I'm double checking some of those sample URLs that you posted there to see if there is something, I don't know, maybe confusing on our side or maybe confusing on your side that's happening that we could point you at. Can you explain why Google doesn't add whether links are followed or no follow to Search Console? In general, we try to provide a kind of a representative sample of the links to your pages in Search Console. And we focus on the fact that there's a link to your page there. So it's not something where we purely focus on whether or not it's passing page rank because a lot of times that it might not be passing page rank, but it might still be interesting for you because you're getting traffic from that page. Or maybe that's something where you can take a look at these links and say, well, this is a reasonable source of traffic. And maybe they should have put a no follow on those links because it's actually an advertisement for our website. But essentially, the primary target for Search Console is not to kind of encourage you to go out and get more page rank passing links, but rather to get an understanding of how your website is connected within the rest of the way. And it's connected in ways that are no follow links as well. So that's something where we consciously decided not to pull that out separately so that we don't, I don't know, encourage sites too much to go down this route of like, I need to make all of my links followed, or I need to have the optimal no follow follow link ratio, which doesn't exist. You mentioned that many EDU links are ignored, are not given as much weight. Are there other TLDs that are also ignored? So I think it would be exaggerating to say that we ignore all EDU links. We don't treat them any differently than any other links from other sites. It's just the case that a lot of SEOs seem to believe that EDU links are what they really need, and they go off and spam a bunch of these EDU sites to try to get links. Or they try to do sneaky things with sponsorships or other ways of getting links from these sites because they think, well, you must be worth a little bit more. But essentially, from our point of view, these are normal domains. There's nothing special happening with regards to links from EDU domains, or from org domains, or .gov domains, or anything like that. So that's not something where we would blindly say, this TLD is good or better than other ones. It's really a matter of the web the way that it is. I get paid for writing articles from various sites, some links to my websites, and some don't link to my website. Those who don't link to my website link to my Google Plus profile. Would this help increasing the authority of my website? Or is it necessary that if people are talking about your brand that they must link to your website? So I think, first off, I would make sure that you're not doing guest posting on other people's sites just to get a link to your site. That's something that the web spam team does sometimes look at and see, are these links essentially links that you're placing yourself? And if you're placing all of these links yourself, they're not really natural links, right? They're links that you're putting out there. So that's something where you need to watch out for what you're actually doing there. And I would almost recommend using no follow links for a lot of these cases where you're guest posting on other people's sites. You can still get traffic to your site, and people can still come to your site and say, well, this is a fantastic site. I'll recommend it further to other people. And that's kind of what you want. You want to have this natural recommendation based on people discovering your content. So that's kind of what I would aim for there. I wouldn't focus so much on where specifically these sites link. And essentially, if you're placing those links yourself, then kind of treat them as links that you're placing yourself. And be honest in that regard. Let's see, mobile first question. There are different views around hiding content behind tabs and accordions and whatnot. Can you confirm if you hide content behind a tab of some sort that it still gets seen by Google as the same content that isn't hidden? Are there likely times when it isn't advisable to do this? So I'm not sure about that last part, what specifically you mean. But in particular, when it comes to mobile first indexing, we understand that mobile sites have a limited space on the pages. So that's something where we do treat content that's behind a tab or maybe invisible by default until a user navigates to specifically part of the page. That's something we treat as normal content on a page. You might see some differences with regards to how we show them in the snippet. But essentially, from a ranking point of view, they're completely equivalent to visible content. I would still make sure that from a usability point of view, you're giving users access to the content that they're looking for. So if this is a primary piece of information on your page, make sure it's visible. Otherwise, people will go to your pages after clicking in from the search results and are like, where do I find information about these fancy shoes that I was told about in the search results? And if they can't recognize, oh, it's behind this one tab that they have to click on, then obviously they'll go away and go somewhere else and buy their product somewhere else. So from a usability point of view, I would still recommend making sure that anything that's important to your pages is really visible on your pages. It seems kind of obvious, but sometimes designers like to make things really sleek and kind of clean on the page, and then it hides some of the important things that actually users are really keen on. Oh, wow. We just have two minutes left. Maybe I'll just open it up to questions from you all. If there's anything left on your side that you'd like to go into. I have a quick feedback. After the Google update lately, I've seen, and it's not only me, but many other people. For example, there are merchants that do not sell on Amazon or eBay. They sell only through their website. And their products are especially labeled not to be sold on Amazon and eBay. But for some reason, some of the customers decided that they do not want their product after they are fordered it. And they are selling it second-hand on eBay and Amazon. And what happened is that after that product is sold, maybe it's two or three items maximum, eBay and Amazon keep pages of these products, no matter of the fact that they are sold out. So if you search for the product name, just because of the power of the domain name of Amazon and eBay, we see tons of listings of items that are sold out. Basically, it's an empty page which said that the product is not there. And they will never be another item there on that page. And that's a problem because I think that Google, because of the power of the authority from these reputable brands like Amazon and eBay, does for some reason think that, well, maybe after some weeks or something like that, we will have the number of the products. But unfortunately, I think that's a very bad user experience because sometimes it even outranks the merchant website. So we basically open an empty page which says that there are no more items there. Can you send me some examples? Yes. That would be great because we try to recognize those kind of situations, kind of the out-of-stock situation where we know that the page is about this, but it says that none of this available. And we should be able to pick that up and show that I've seen other comments from other people around that topic. Yeah, I mean, if you can send me some examples, that would be great. And maybe this is something where we get it better in English and in other languages, we do a terrible job because we don't recognize the text. That's also really important. All right, cool. So we've reached the end. Thank you all for joining. We have next English Hangout lined up for Friday. Also, if you're in Zurich or in the area and you want to join in person, send me a note on Twitter. And we can try to get that set up. We also have a Google Dance lined up on Friday, so a small event here. And the German Hangout, of course, on Thursday. So if there's anything that comes up in between, feel free to drop those questions in there. Or, of course, drop by our help forum and let us know about things there. All right, thanks again, everyone. And I hope to see you again soon. Thanks a lot. Have a great day. Bye-bye. Bye.