 That's fine. All right. Welcome, everyone, to today's Webmaster Central Office Hours Hangouts. My name is John Mueller. I'm a Webmaster Trends Analyst here at Google in Switzerland. And part of what I do is these Office Hours Hangouts. And for some reason, the URL changed when I had to restart the Hangout. So it looks like we have Daniel who managed to find a new link. I've got exclusive access. This is brilliant. So if you have any questions to start off, feel free to jump on in. Yeah. OK. Yeah, I've got a couple of questions, actually. I suppose while we're waiting for other people to join. So on desktop, if you hide content, then it's considered as a lower value. It gets a lower waiting for ranking. Now, with the mobile-first index, where, for example, crawling an index in content directly for mobile, is that going to change now in any way? Because I suppose with mobile and mobile interface, you have to hide content. You have to have content in tabs. So is there going to be some sort of reli- you know, is there going to be a more relaxed approach from Google side on mobile with hiding content? Or is it, you know, could you kind of try and explain how that's now going to impact mobile? Yeah, with the mobile-first index, we decided that sometimes with the UI limitations on mobile, you kind of have to use these kind of, I don't know, design patterns where you hide. That's fine from our side. So on desktop, it's something where we think if it's really important content, it should be visible. On mobile, it's a bit trickier. Obviously, I think if it's critical content, it should be visible, but that's more kind of between you and your users in the end then. So again, does it kind of stem down to user interface? It's I suppose kind of what would you recommend as the better option at a show that make visible all the content on mobile, but then the page becomes insanely huge or interface. Now, as I said, interface, but I think there's that, there's that worry that, you know, there's that risk of losing and ranking or not ranking as well because we're not showing a Navinal content visible. From an SEO point of view, both are legitimate options. It's totally up to you. Both are legitimate options. And you will, are you treating, so are you treating content on mobile differently to desktop now then? When we index the mobile content, yes. So at the moment, we index the desktop content. That's a canonical. And switch to the mobile version, then we'll be treating it differently. You'll be treating hidden content differently. Yeah. Oh, fantastic. Okay, I suppose I'll just kind of like slide into my next question then. I think it's just to get your advice a little bit or any kind of tips on this. And it's where you have hyphens or spaces or forward slashes in between words. So an example I've got here is DC hyphen DC, which is kind of the plug, you know, the kind of the port, sorry. But obviously you can write that in hyphen, DC hyphen DC, DC forward slash DC, DC with the space DC. And generally, I suppose we would write that depending obviously how, you know, I suppose what the industry kind of, how it kind of sits with the industry maybe or what we would write it probably in a certain consistent way. But the problem is they rank differently, you know, if we were to type DC DC with hyphen or space, it ranks very differently. And so obviously it seems like the option is to write them differently on the page because they're a different rank, the different ranks for different ways to write it. But I suppose it's kind of just to get your tips and suppose ideas on that because we want a consistent way of writing things like that. But it's very difficult when we're not quite sure how to approach it. In general, I suppose we would approach it based on okay, what generates more search volume. But Google now and keyword plan are more or less merges these different variations into one. So then of course we don't have that visibility either. So I suppose if we get some tips from you, that'd be great. So I'd recommend just writing naturally and trying to figure out what your users are writing, the language of your users and using that. And that's probably where you'll have the most success there. So for the most part, these kind of variations, we try to recognize the synonyms. So we'd fold them together. But sometimes like you said, we still recognize subtle differences between them and we'll try to rank them differently. So I look into your own logs, look into search console, see what people are searching for. I've seen some people, for example, run AdWords on these kinds of questions. And also to see what are people searching for, what would people click on in the search results with regards to titles, snippets, those kind of things. So it's something where, or basically your experience comes into play and you kind of have to think like a marketing person rather than just an SEO. Okay, I'll leave it there for a while. I might have more questions later, but I'll give someone else an opportunity. I appreciate that. Sure. Any questions from your other folks? Not at the moment. Okay, all right, here's one in the chat. How many canonical tags are allowed on Google and the website page? So we recommend just having one canonical tag on a page for a website, like across all of the pages on the website, that's obviously on a per page basis, but per page, one canonical tag so that we can really have a clean signal and say this is the preferred canonical that you want to have for that specific page. Oh wow, okay. And a list of canonicals to different, like country and language versions. So what I would do there is not have the different canonical tags on one page to the different country and language versions, but rather use the hreflang link between the individual URLs. So let me just write the word in the chat here. We have a lot of information about that in the help center. And with that, you can link between these different country and language versions to tell us which one you want to have used for which users. All right. Let me run through some of the submitted questions. My voice is a bit shot, but we'll see how far we go. Otherwise we'll have to kind of limit ourselves to yes, no questions. How would you say that online users tend to behave when a site has great unique selling proposition in comparison to other websites? Wow, I have no idea where to even start. So from my point of view, this is not really a SEO question. It's something where you probably get more input from general online marketing type people. So that's kind of who I would talk to for something like this. So how users behave. Is it a good practice for every page to have a real canonical pack pointing to itself, even if there's only one page? Yes, I recommend doing this kind of self-referential rel canonical because it really makes it clear for us which page you want to have index or what this URL should be when it's indexed. Even if you have one page, sometimes there are different variations of the URL that can pull that page up. I have an example with parameters in the end, perhaps with like upper or lower case or dub dub dub, non dub dub dub and all of these things can be kind of cleaned up with a rel canonical tag. Multiple product links pointing to the same product using the canonical to avoid duplicate content but in the HTML improvements in search console, there's still displayed. Is that something we need to fix or not? That's something you can essentially ignore if you're sure that you set up the rel canonical on these pages, then that's fine. We show that information in search console when we initially crawl and index those pages and the first time we see a rel canonical we have to index that page first before we can follow the rel canonical and kind of process that. So that's probably what you're seeing there, this kind of subtle skew between the first time we index a page and afterwards. What's the ideal content length for a page? Totally up to you. Some pages are short, some pages are long, there's no word count, there's nothing where Google has an algorithm that says this page must be this long to rank optimally. That's not the case, it's really totally up to you. Why do m.pages rank on desktop? So let's see. This is usually a sign that the m. and desktop page link between the mobile and the desktop URLs is not set up completely correctly. So perhaps the rel alternate link is wrong or the rel canonical link is wrong. Or perhaps we've crawled the m.version without crawling the desktop version first and it's not a sign that anything is changing on our side. This is really mostly a matter of the website being set up in a way that we think these are URLs that we need to index separately. So it's not that also the question goes on like shouldn't these be desktop optimized pages? So essentially every mobile page works on desktop. So from our point of view, it's not the case that mobile oriented pages are worse on desktop. We have a lot of filtered links, kind of dynamic categories and filters. And we use a rel canonical to avoid duplicate content. Is that okay? That's a good approach. There are various approaches to this kind of faceted navigation is the term that you're probably looking for. And I double check our help center or blog but there are also lots of other blog posts out there that have tips on how to deal with faceted navigation. So I check those out. Rel canonical is definitely a great way to start. Internal links, can the position of a link on a page affect its weight? Do image links get treated the same as text links? So position on a page for internal links is pretty much irrelevant from our point of view. We crawl, we use these mostly for crawling within a website for understanding the context of individual pages within the website. So if it's in the header or in the footer or within the primary content is totally more up to you than anything SEO wise that I would worry about. Image links versus text links. One thing to think about with image links is if you don't have an alt text for that then we don't have any anchor text for that link. So I definitely make sure that your images have alt text so that we can use those for an anchor for links within your website. Also, if you're using image links for navigation make sure that there's some kind of a fallback for usability reasons for users who can't view the images. Also, specifically around images in general. So not like specific to image links, but images in general, especially on mobile and make sure that your mobile pages also have an alt text. Some predefined links to like about us policy pages. Do we need to know follow them or do anything? No, I would just link to them normally. See, bunch of stuff here, skip. We're running an external fashion blog where we post blogs and give a followed anchor back to the e-commerce website. Is there anything wrong with this? It depends on how you have that kind of set up or how you deal with that. So if those links are in response to anything that you're getting from this website then you should be using no follow there. For example, if they give you these products to try or if they give these products to you in general, again, in exchange against a review, then all of that is kind of a relationship between you and that other website. And then those links would need to have a no follow attached just so that it's clear to us that this link is not there solely because you thought this was a good link to place, but it's there because there's kind of this background relationship that's happening. Thank you, I'll miss you, Dad. Okay, okay, well, he's gone now. Okay, yeah. So again, like with a fashion blog or in general any kind of a website that you run, if there's a background reason behind the content that they were posting and the links that you're placing, just make sure to have a no follow attached to those. In general, this doesn't mean that the content is worse or that these links are bad. It's really just a sign for us that this link is there because of some relationship in the background and that can be totally fine and still be relevant for your users. If there's a disparity between geo-coordinates and schema markup and embedded map on a website, can this affect rankings? I can't think of any way that that would affect rankings. So not that I'm aware of. I think the strongest aspect is really just you're Google my business localistic so that we understand exactly where your business is located and we can show it in the local search results. And that's really the strongest way for us to kind of understand that your business is located here and this is your website that's attached to your business and we can kind of link and rank those separately. On our category pages, we display 25 products with their titles beneath the image. Can we make them H3 or H4? Will that help? You can do that. I don't think that would necessarily change anything for SEO reasons. So kind of up to you. What social signals impact SEO rankings? As far as I know, no social signals affect SEO rankings. So especially with regards to social media websites, usually those links are no follow. So they don't really have any effect there. They can of course indirectly affect your website in that if you have a lot of traffic to your website and they love your website and they recommend it to other people, that's something we could pick up on. But there is nothing directly like just tied one to one to social signals where you can say, I will have my friends plus one my website a thousand times and then my website's ranking will go up or I'll have my friends kind of share my link on Facebook four thousand times and then my website rankings go up. That's not something that would happen there. That's kind of also very easy to manipulate. I keep having a ton of non-related websites, education language websites, linking to my tool rental website with two follow links. Is that affecting my website in any way? I don't see how that would be affecting your website. So lots of websites get weird links that they might not expect. A really common one is forum links from another language forum where you don't really understand what these people are talking about but they're linking to your website and they're going to your website. And just because it's not directly related doesn't necessarily mean that it's a bad kind of link. So from that point of view, I wouldn't focus so much on whether or not these are related sites or not but try to think about is this a natural link or not? Is it possible that maybe your previous SEO placed these links and you're just not finding them and they've been like link spamming for your website? And then those are the kind of links where you might say, okay, I'll just put them in my disavow file and have them taken care of that way. So in general, if you feel that links, specific links to your site are problematic and you don't want to be associated with them, I will just go ahead and put the domain in the disavow file and move on from there. Can I publish 100,000 pages at once? Sure, no problem. I think the biggest impact, the biggest effect you'll see is that Googlebot will try to crawl a lot of pages, especially if it's a larger well-known site where we already know that crawling pages and indexing pages from that site makes a lot of sense. Then we'll probably kind of try to crawl a lot of pages on your server and it will be good if your server had the capacity to deal with that but from an SEO point of view, that's generally not an issue. Some sites put their archive online all at once. I think kind of artificially introducing a kind of a trickle into the index is something that often causes more problems than it solves anything. So obviously putting 100,000 pages online all at once doesn't mean that these pages will rank all at once immediately. It's still going to take time for us to actually rank those pages properly in the search results but I don't see anything against putting them all online. E-commerce sites with 50,000 pages, what would the internal links count be? Really depends on your site structure. Some sites have a very flat hierarchy where they link very quickly to a lot of the leaf nodes. Some sites have a kind of a narrow hierarchy and kind of a taller hierarchy. There are usability studies around let's say like the pages should be maximum, so many links from the homepage. From our point of view, both are options. It's something that you kind of have to look at and think about what works well for your website. The thing I just kind of watch out for is if there are things on your site that are really, really important that are really critical for your website either where you think this is like your primary product, the one thing that people really, really want to get from your website or it might even be something where you say, well, this is the product I make the most money from, therefore I want to promote it the most. Then I'd make sure those kind of links are placed fairly high up in the hierarchy so that Googlebot can get there quickly so that we can pass page rank to those fairly quickly but also of course so that users can get there fairly quickly. So not from Hangout Links. Does content mismatch between the web and deep link of an app trigger in algorithm or harm SEO ranking? I don't know for sure what the current status is there but it used to be that when an app page was seen as having a mismatch of content with the desktop page, then we wouldn't be sure if we could actually link those pages together. So the background here is that with app indexing you can specify this is my web page, the one page from the website that I want to put in the app and this is the URL that have within the app where people can go directly. So the idea being that when we rank your web page in the search results, we can swap that link out and point directly to your app with the app deep link and we can really only do that if we're kind of sure that the app page is actually the same content as your web page. So that's what we're checking there. Obviously that's really hard to determine properly because apps are structured in very different ways than web pages but that's kind of what we were looking for there. If you're unsure about these errors in Search Console, I recommend checking out Firebase. They have a support feature on their site where you can submit, I believe, a handful of support tickets directly to the Firebase team to have them double check the app side of things. Do interstitials impact mobile ranks only or desktop ranks as well? At the moment, the mobile interstitial is specific to mobile. So that's something where on mobile devices, when we rank a page that has an interstitial that we've recognized, we'll kind of treat that page as being not so mobile friendly as the other pages. When we remove products, we 301 the links back to relevant category pages. Is that okay? You can do that. You can return 404. You can have like a friendly 404 page, of course. That's kind of up to you. If we see too many links to the category page, we'll assume that it's a soft 404 and treat it like a 404 internally. So I primarily watch out for usability in a case like this because we'll try to figure out which one is a 404 and which one is actually a 301 independently anyway. I usually recommend using a 301 more if you have one product that replaces a different product. So not that you say, well, this product is gone but we have lots of similarish products here too but more like this product is obsolete but we have this one that kind of replaces the previous product. That's the ideal situation for a 301. Should I implement AMP for e-commerce pages in the Netherlands? Sure. Go for it. I think there's more and more functionality for AMP specifically for e-commerce. I would keep up with the AMP team what they're doing and see which aspects make sense on your website and which ones don't. AMP is a per page set up so you can pick and choose and you can say, well, my category pages work great on AMP because I have all of the functionality available but maybe my product detail pages are not so great on AMP yet. So in a case like that you might say, I'll just put my category pages on AMP and start like that. You could also think about things like, well, maybe you have a blog attached to your e-commerce site and switching the blog to AMP might be just a matter of installing the WordPress plugin, for example. So that's an option there too. What should I do about someone in another country linking to my website on his social profiles? Whatever you want. So usually this is a non-issue. If you really don't want to be associated with those links then obviously use a bit of alphabets. How does Google treat websites that only give no follow external links? We treat them the same as any other website so we don't do anything special there. Is a product page considered duplicate content because there are many sites that display the same content? It can be or it can be that part of the product page is considered duplicate content and we fold those together when we show their search results. So if you can provide something unique and compelling on your product pages then I would definitely go for that. Sometimes there are also things like your own reviews, maybe your users' reviews, your people who have bought this product, similar related products that they bought, all of this can add additional value to your pages. So when John said on that, so when you say you fold them together obviously you're going to show one page out of say maybe five duplicates so you'll just show the one. What would be the best way to go about that? I've got something similar, I've got a client who has multiple of the same product but it's based more or less, the only difference is the color. So they've got individual SEO-friendly pages but the only change really is the color. Would you recommend to have one page where they could just change the color themselves or keep it as it is? I think my understanding is we've got five products, five different colors, but only one will be used depending on obviously the search. So what would be your recommendation on that? I think about what the primary aspect of this page is, is it primarily the product and the color is kind of an attribute of that product or is that color kind of the unique selling proposition for that product that they have in there? So you could imagine that maybe you have something like I don't know, a diamond plated laptop and you have the same laptop available in black then obviously the laptop itself is kind of the same thing but the kind of the color, the style is something that's very unique and that's something where you'd say, well, I probably want separate pages for that. On the other hand, if it's like a laptop and you have it in green, gray and black and people kind of randomly choose which one they want then that color is not the primary aspect of that page. Yeah, would you go back to the search volume and would you then kind of go back and say, well actually you would kind of go back to basics and have a look at the types of searches for that product using a Q&A monitor, for example. Sure. Say for example, if we use Diary so we've got a pink Diary with black down people that might actually search for that then again, just optimize based on the color because that obviously is probably its main USP in comparison to the rest of the duplicates. Yeah, yeah. I mean, that's something where you know your products best and you can kind of figure out like, are people coming to your site because of the color or are they coming to the site because of the product? Right, yeah. Okay, now it makes sense, thank you John. Hey John. Hi. Hi Bharat here. Actually I have a question related to some 500 error. So what is happening, we were having old URLs which were not SEO friendly, search engine friendly, but now we have come up with the new URLs, the canonical URLs which will be replacing those URLs. So instead of redirecting 301, we have just inserted the canonical URLs. Does it harm in SEO by any chance? So we use a number of factors to try to figure out which one of these URLs is the right one to show that includes the canonical, it also includes redirects and internal linkings, for example, those are probably the strongest signals that we use for that. So if you can kind of align as many signals as possible to your new URLs that you want to have used, then we'll try to use those. If you're just changing the canonical and the internal linking still goes through the old version, for example, then probably we might choose one or the other, but it sounds like you have a strong preference which one you want to have shown. So I would try to align everything as much as possible to point directly to your preferred version, which includes redirects, if you can set them up, definitely internal links, sitemap file, all of that pointing to the new URLs that you want to use. Yeah, so the canonical URL in the page and this sitemap URLs, those are the URLs which we want to, we want Google to be indexed. So those are clear, but my concern was related to 301 URLs which were throwing 5.00. So should we redirect the old URLs to the relevant new page URLs? I would do that if you can do that, yes. I think that's... Okay, okay, thank you. Okay, thank you very much. All right, let me run through a handful more questions that were submitted and then we can move over to more live questions for you all. Is it an issue if we have a large number of nofollow on internal links? No, it's not an issue. I think, especially with regards to internal links within your website, it's probably wasted effort to try to do like page rank sculpting, those kind of things there. I would mostly use nofollow in situations internally where it's really a matter of kind of the crawler getting lost with it goes in that direction too far. For example, endless calendar links is a good approach there. You might want to nofollow things that go too far out in both directions. Perhaps faceted navigation pagination, you might at one point say, well, I'll just nofollow the links here so that Googlebot doesn't dig its way even deeper and deeper into my website. But otherwise, for things like your about us page, your terms of service, all of those things, I will just link to them naturally. We have a lot of practice with normal websites. It's not going to be the case that we will say your terms of service page is the most linked page on your website, therefore we'll rank it number one. That's definitely not going to happen. I'm making a Hindi language blog in India. How can I tell Google my language is Hindi? Is there any way I can tell? No, there's no way specifically to tell us for one language version which language that is in, but we do pick that up automatically. So the one thing I would recommend doing there is really making sure that your whole page is in Hindi. So one thing that you should avoid, for example, is having parallel translations where you have English here, Hindi here, and they're both on the same page because then we have trouble understanding what is the primary language of this page. But if we can understand the primary language as Hindi, then that's fine, that's all we need to do. And I think it's fantastic that you're doing a Hindi language blog and I wish you luck with that. Sounds like a fun project. We have a branch finder page with our branch links behind tabs, are these links being discounted? Should we have one page with all the branch links visible as well? So in the current state, we do kind of discount hidden content on a page. So if this is primary content for your page, then that's something we probably won't be able to bubble up as well in search at the moment. That will change with the mobile first indexing. But in a case like this, it sounds like you have primary content on your page and in the tab, you have kind of cross links to different other branches that have this product. And in those kind of situations, that's not something where I'd say this is a problem because that branch information within that tab is probably not the primary information from that page. So that can be perfectly fine. Can I just ask on that? You said it's going to change. Could you explain how it's going to change? I.e., it's not going to be an applicable algorithm for mobile anymore? So with the mobile first indexing, we'll index the mobile version of the page. And on the mobile version of the page, it can be that you have these kind of tabs and folders and things like that, which we will still treat as normal content on the page, even if it's on the initial view. Great. Thank you. I have some, a subdomain URL with the same content as the main website. We can't delete the subdomain, but we've implemented a 301 redirect to the main website. Is that okay? That's perfect. Yeah. If we set up a redirect, then users go to the main version. Search engines go to the main version. Perfect. The question, I think, for the Hindi language blog goes on, some of the articles are in English. That's also fine. So the primary thing is that on a per page basis, we can clearly understand which languages page is in. And that's what we need to know. If some pages on the website are in different languages, that's perfectly fine, as long as there's always one clear language on a per page basis. Hi, John. Hi. Hi. My question is, if my site is penalized, so how to recover that one? It is based on the content, duplicate content, or it is based on the URL, or based on the user navigation. What are the major factors penalized in my site now? Okay. So, I'm not sure if I got everything right, but it sounds like your website dropped in rankings, essentially. No, it is manual penalized for the last June. Okay. So if I'm going to resubmit, what are the minimum rectification required? That depends on the issue that was found on your website. So in the manual action section, there should be some information on what we discovered on the website and what the website, what web spam team took action on. And some of these might be issues like you have unnatural links on your website or unnatural links to your website that you can clean up. Some of these can be based on the content itself. For example, we have one manual action that's kind of for the extreme situations that we call pure spam, where essentially the web spam team looks at the website and says there's nothing really valuable here that we would miss if we didn't have this website index. And those are the situations where you really need to kind of rethink what you're doing and think about what you can provide on your website that really has value and remove everything else. So for example, I did one common thing that we see a lot is people take these scripts that essentially show YouTube videos on your pages. And you kind of create this video site with a collection of YouTube videos and nothing else. And from our point of view, that's just aggregating content from other websites that's not providing something unique and valuable on your website specifically. Or for example, one thing we sometimes see is affiliate sites that just take an affiliate feed and they publish it one-to-one on their website and they don't provide anything of value at all, then the web spam team might look at that and say, well, this is just a duplicate of Amazon. There's no reason for us to actually index this content directly or maybe they'll do one of these things and run it through a tool that rewords the articles or that takes different news articles and just rewords them subtly. And that's also something that the web spam team would say, there's nothing of value here that we need to have index. So we'll remove it as pure spam. And if the webmaster says, okay, I really recreated my website, I have something that is unique on my website that is not available anywhere else, then that's the kind of thing that the web spam team would say, well, it makes sense to reconsider what we have there from manual acting and kind of take that back. Yeah, most of the aggregator site, it is content is aggregated by the users, correct? Yeah, I mean, if the content is aggregated and you're not providing anything additional, then that's kind of one of those situations where we say, well, we already have this content index as well because by definition, if it's aggregated, we already have it somewhere. So you really need to provide significant additional value there so that it makes sense for us to index your website and show your website instead of all of the other websites. So you just have to write the content again and again, but they are just the bookmarking the content, correct? It really has to be that your website stands on its own so that there's a situation where we say, if someone is searching for this specific topic, we should clearly show your website instead of any of the other ones. So instead of aiming to be kind of as good as the others, I would really aim to be significantly better so that it's really clear to us and to users that going to your website has a unique value of its own. All right, let's move on. I have to know index that all the URL which are generated by the user. Depends on your website, I don't know. So some websites, when they make bigger changes like this, they know index the content, some of them remove the content completely and say, I don't trust this old content, some of them work to improve the content, which I think is a good approach as well. What is your suggestion in the UGC site? Because I have no control on the user. I think for UGC websites, the important part to keep in mind is that you're hosting this website and you're providing it to the rest of the world. So just because the content originates from users and it's like, I didn't write this content, doesn't mean that you're not the one who's kind of held responsible for it because you're publishing this content. You're showing that content there so that when we crawl and index your website, we see this content and we don't really make a judgment like, where does this content come from and how should we treat it differently because it's not you who personally who wrote it, but someone else, but rather we see you're providing this website with this content on it. So that's how we will treat your website. All right, let me double check the other questions that were submitted. If there's something short that I can jump into is with labels for reuse answer, you really need to work out with content that you find online yourself. Okay, John, we can't hear it keeps going in and out. I thought I just mentioned that it's getting quiet when you talk. I don't know if it's the internet connection or not. I can hear you now. Okay, maybe I screwed the other microphone. We're good now, we're good. Sorry. Okay, perfect. Okay, Google is indexing pages that I have a no index on. What's up with this? I don't know. So if we see a no index on a page, normally we will drop that page from our index. There can be situations where it takes some time for us to see no index, for example, for large sites, or if a URL is blocked by robots text. But as soon as we've indexed the page with a no index, we will drop it from the search results. And if you see that not happening, then I'd love to see examples for that because that would be a bug on our side. That's not something that should be happening. Let's see. I think most of the other questions we've kind of gone through. Yeah. What else is on your mind? Oh, here's one. What's the best operator for figuring out if a URL is indexed or not? So what I would do there is use an info colon query with the URL to see what comes back there because that's almost always a sign of what we've indexed like that. So I'd recommend doing that. If you're searching for keywords or if you're doing something complicated with the in URL operator, then that can be a bit tricky. But the info operator is usually the best one there. With regards to the cache, the cache page doesn't necessarily reflect what we actually indexed. So I just keep that in mind. And sometimes we don't have a cache page for things that we do already have indexed. All right. What other questions are on your mind? What can I help with? Can I jump in and just a quick question around my brain, please? Okay. Okay, so you've got these different Google algorithms. Is it RankBrain's job to decide what algorithms are used based on the query? So for example, if I search for something like shoes, RankBrain might decide actually the metal type and the metal title tag would be more important. Whereas if I search for something like health or health care or something like that, then RankBrain would decide actually links, page rank is more important. Is that what RankBrain does is, does it depend on the search? And I suppose RankBrain's learning and understanding of different searches and different industries and things like that. I don't know. Where can I start? So. I started with the question. Sorry. Yeah. So from my understanding, RankBrain is more about understanding the query. It's specifically understanding queries that we don't really know how to handle. So things like shoes or health care or something generic like that is less of a problem. But we see roughly 15% of our queries every day are completely new that we haven't seen before. So those are the kind of situations where we say, we have no idea what we can show here. We have to kind of figure this out on our own. And we have to understand what the user might actually be meaning with this query. So maybe they're mixing things up or maybe they're using synonyms in a weird way that we haven't seen specifically like that. Then that's something that RankBrain will try to figure out. It's like, what is this person actually searching for? So I guess one example I've heard in the past is, for example, if you have not in your query somewhere where you're kind of saying, well, I want shoes that are not in blue, where if you just look for the words there, then you're saying like shoes, not blue, and you'll probably show them a page with blue shoes on. So I have no idea what kind of results this query brings now. So maybe that's a bad example, but that's the kind of thing where understanding what this query means makes a big difference in the type of search results that we can bring. To say it's more focused on language rather than all the multiple different core algorithms is tied in together. As if I say that it's got nothing to do with that, but actually it's just understanding language and using it. Thanks, yeah. I mean, obviously we don't just have one algorithm that tries to figure this out. We have a lot of different things that try to figure out. Like what is this user actually trying to do? Are they trying to get more information about a product or do they want to buy it now? And those are very different kind of approaches with regards to the type of results that we should bring. Or sometimes it's also the case that we say, well, we don't know. It could be this. It could be that. And we should show a diverse mix of search results for this specific query in this specific case. We're just saying just standard search and standard search results that you would use the same number of algorithms. It doesn't mean that. So for example, if I'm looking for cars versus shoes, for example, each algorithm that you use is actually being applied to each website in the same way. It's not applying some algorithms in one case and some algorithms in another. I think it's always a mix. So that's something where because we have so many different signals that we use for crawling, indexing, or ranking, it's something where some sites have this a little bit stronger, some sites have this a little bit stronger. And we still have to find out how do we weight these for this specific case? And that's kind of, I think what brings additional value into search there as well is that you don't have to do exactly the same thing as other people in your niche. You can focus on your own thing and kind of work on your website in your own way. And we'll still try to show that appropriately in the search results. So if your competitor has a link from CNN, you don't need to go off to CNN to try to get a link there too. You can work on other ways to kind of differentiate yourself. And we'll try to show that in search. So one site might have a lot of links. The other site might have a lot of really awesome content. It's like, well, we'll try to figure out which one is the right one to show in this specific case. OK, that makes sense. Thank you, John. All right. More questions from you all. Anything else on your mind? Hi, John. Hi. Just a quick question regarding the API. I don't know any on the API team, so I guess I'm just going to ask you about it. So a few users reported a bug that's been occurring with pagination. It seems like there are multiple pages of results, even if the first page doesn't reach the maximum 5,000 rows. So some people have been getting like 4,000 and something. But they notice there's an additional page if they query with a paging parameter. So I don't know if that's new or. But I've seen that in the past week or so. Yeah, I've seen that for a while. I thought there's something that they were fixing there. So I'm not sure what the current status is there. But I know that the API team is aware of that. And that's something that they're trying to improve. OK, because I only check for more than one page if I get 5,000 results, so I don't query every time. So yeah. Yeah, I think it's a matter of the sorting there. And especially, I don't know. I'd have to double check what I saw there. I thought it was something like if it has the same number of clicks as other queries or you're out there, then sometimes with the order we would get that wrong with pagination. OK, I'm good. Yeah. Well, if you get any news from people. No working on that. So I hope that will be improved. OK, awesome. Thanks. Cool. All right. So I guess if there are no other questions, we can take a break here. Thank you all for joining. I believe the next couple of Hangouts will be skipping since I'm out. And probably back in, I don't know, three weeks, something around that time. So I'll set up the next Hangout just afterwards so that you have one place to kind of pull your questions into. And otherwise, of course, going to the Webmaster Help Forum is a great place. There are lots of great people also active on Twitter. If you have any questions. So in the meantime, lots of places to also get help. And until then, I hope to see you all again. And have a good time. All right. Enjoy your holiday, John. Enjoy your holiday. Enjoy it while you can. Yeah, thanks. Bye, everyone. Have a good weekend. Thank you. Bye, John.