 All right, welcome everyone to today's Google 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 we do is talk with webmasters and publishers like the ones here in the Hangout today. And of course, the ones that submitted a bunch of questions already. As always, if any of you want to jump in with the first questions, feel free to go ahead now. Yeah, I'll go for it. Go for it. Can you hear me OK? Yes. Perfect. So it's my first time here. Welcome, everyone. Thank you for having me. And I submitted a question on Twitter, and you've partially answered. And thank you very much for that, because I imagine you get hundreds of those every day. We work with a big international fashion brand. What happened is that sometime ago, the knowledge graph has disappeared. And at the same time, I noticed that on Google Trends, when you search for the brand's name, it doesn't show as a brand anymore, although it used to show as a brand. Now it shows as a topic in a category in a Google Trends when you look up their name. And it's a cute brand. It's international. Basically, they just suddenly disappeared. And we tried fixing everything. Bro, why are you looking at yourself? So back to the topic. And basically, they suddenly disappeared. We tried to fix everything we caught, but it just seems to be extremely odd that the brand of the size wouldn't trigger a knowledge graph. Would you mind, do you know anything on that topic? What else could we try? We tried fixing Google Class accounts. We do have a huge Wikipedia listing that's been there for years and years. And the brand has been in existence for many years as well. So it's just a very, very weird case for us. I don't know. I don't know what specifically that could be. So in general, the knowledge graph entries are kind of algorithmic in that they can come and go over time. But I don't know if any of that kind of plays a role in your specific case. If you want, you can maybe send me the name on Twitter or on Google+, and I can double check with the team here. I don't know if they'll be able to change anything there, because this is essentially all algorithmic. And they wouldn't manually tweak things like that. But I'm happy to take a look. Thank you. I'm just afraid that it's a bug of some sort, so that maybe we have a problem in our schema code somewhere, or something just isn't linked up correctly. Because for the size of the brand that it is, I mean, I saw way smaller brands, like local electricians show you the knowledge class, so kind of why they wouldn't. But yeah, thank you for that. I'll send it today. Sure. That's great. All right. Any other questions from any of you before I get started with the list that was submitted? Nothing special? OK. If you have a question, is that OK? Very quickly. Question, this is something you've covered a bit before, but it's OK. So we have a PR company, and they basically do our PR for us. Most of it's offline PR. It's not really online PR, so they get us into newspapers, our products into newspapers, and magazines, and so forth. And what they do is that they have a image library designed for journalists, so the journalists register and they go on to their image library site. It's quite a well-known website, press website. And they've uploaded all our images from there so that they can be downloaded by journalists. What we found out very recently is that they haven't just uploaded all of our images. They've also uploaded all of our product descriptions. So what we effectively have is the journalist database, which is probably quite a well-respected database, that has content very, very similar to our website. Now, I appreciate that this happens quite a lot, and Google is generally very good at working out who's the master and who's the slave kind of site, and it doesn't kind of filter the wrong one out of the SERPs. But we have historically had a very, very major problem with Panda. And I know that there are people in the past who've said that sometimes spam sites that might kind of scrape their content, they kind of get shown first. Now, we've done some very, very good work in recovery from Panda, but we've never really seen any recovery from Panda. We've got some very well-known chap on board who I think you probably know from Twitter and so forth who's helped us a lot and made the site absolutely fantastic. But we've never seen any real recovery. And so the question is, are there situations where if a very respected, excellent kind of website innocently uses your content in this PR kind of way and your site has been penalized or whatever you're going to call it in some way, is it possible that that might cause a delay or a prevention of recovery from something like Panda? That shouldn't be the case. So I think in the worst case, what would happen is the other content would rank as well in the same search results, which it doesn't sound like is happening, or at least you didn't mention there. You're correct, yeah. It's not that it's ranking instead. It's more whether or not Google is looking at the two sites and kind of saying, well, hang on a minute. We still see that this, as an outside, seems to still be doing something. We think that maybe they're taking something from the other site and therefore we're not going to allow any kind of recovery or anything like that. No, that shouldn't be the case. OK. So it's perfectly OK to just leave the PR site doing what it does. It's not going to cause any issues. I mean, if it hasn't been ranking for your content in the meantime, then I don't think that's going to cause any issues in the future. OK. So again, it's not that Google will see the content of both. And it's not so much about filtering one versus the other. It's that Google is kind of thinking, hang on a minute, our site hasn't done enough to recover yet. Yeah, I don't see any problem in that regard. OK, thank you very much. All right, let me run through some of the submitted questions. And as always, if you have questions along the way, feel free to jump on in. And we can see where we go. A question from Ramesh, linking to one section of the site from the same page with different anchors will it pass value to both anchor texts? Essentially, this is a situation that we have a lot of practice with. So this shouldn't be something where I'd really worry about which link is actually passing anchor texts or which link is passing any value. Lots of sites have this kind of scenario where multiple links from one page go to another page. And we need to understand the context based on something that I would worry about. When do you generate site links for non-branded queries? I see lots of results for competitive keywords, ranking number one with site links. So you only show site links for pages that are considered really high quality. So from my point of view, this isn't something where I see anything special happening in the sense that we try to show site links in situations where we think this provides another extra value for the user to figure out where they want to go within this website. And oftentimes, that's for branded queries. So if someone is searching for your brand name, but actually they're looking for a specific part of your website, that's something that we would sometimes show there. But that can also be for other situations as well. So it's not really tied to branded or non-branded queries. John, so in such case, sometimes I have seen that second position or third position website ranking with site links in non-branded queries, while first position website is ranking on first but does not have a site link. What would you say for such kind of case? That can happen. That's something where algorithms have either figured that out based on what we've seen so far, or it might just be that the structure of the website is just clear enough that we can actually do that for the second one. We can't really do that for the first one. So if crawling and indexing is really hard for a website, if the links aren't really clear, if there isn't a clear hierarchy on a website, then that's something that makes generating site links really hard for us. So that might be a sign that maybe you need to work on improving the quality of the internal navigation. Our business operates from 50 stores. We have a branch finder showing separate pages for each of them with NAP, individual services offered, et cetera, et cetera. But we'd like to show a link to these stores in the footer on our website. Will this be OK or can it cause SEO problems? So in general, you can link to individual store listing pages that you might have on your website. From a practical point of view, it seems linking out to 50 different stores is kind of a bit over the top. So I would aim for something like maybe a handful or 10 kind of the important stores for your area and maybe focus on those. Or just point to your store locator page and link to that from the footer of your web page. I think having 50 links on the bottom to individual parts of your site, that just makes everything a bit more confusing. Can low quality pages on a site affect the overall authority? Yes. Well, I guess it all depends on how you define authority. But in general, when our algorithms look at a website across the board overall, then they do look at individual pages as well. And if there are a bunch of really low quality, bad pages on a site, then that does affect how we view the site overall. On the other hand, if it's a really large website and there are like a handful of pages that are kind of bad, then we also kind of understand that in the bigger scheme of things, those individual pages aren't really like the main issue for this website. So that's something where, from my point of view, if you're aware of low quality pages on your website, then that's something I'd try to fix. And find a solution to be that either removing those pages, if you really can't change them, or in the best case, kind of finding a way to make them less low quality and actually make them useful, good pages on your site. Are fonts used in any way to evaluate a page's content? For example, handwritten fonts might suggest the content is not to be taken as seriously. No. As far as I know, we don't do that at all. So if we can crawl the page and see the content, then that works perfectly fine for us. The only situation where I could think of this being a bit tricky is if you're using images instead of text on a page. Then obviously, us doing like OCR over the images is hard on its own. And as far as I know, we don't do that at the moment. On the other hand, if we have to do OCR over handwritten images, then that essentially makes it impossible for us to find out anything. So if you're using normal text, then you're just swapping out a fancy font because that's kind of your corporate style, or you just feel like using a fun kind of handwritten style font, then that's totally up to you. That's not something that we would kind of demote a website for or assume it's to be taken less seriously. So feel free to use Comic Sans if you want. Feel free to use a kind of a more corporate type font if you prefer. That's totally up to you. How do you find the balance between showing up for keywords you want and keeping a site's content fresh? Google Value and content and keywords that are prominently displayed at the top of the page makes me hesitant to update the top of a site with news and new features. From my point of view, these aren't really items that are completely separate. So that's something where you can update your site's pages if you think that there's something new to say and you can keep evergreen content on your site if that's something that you think hasn't changed in a long time. And it's still valuable like that. So essentially, these are different aspects of making websites. And it's not the case that you need to do one or the other, you can do a bit of both. You can find the mix that works for your website and work out what works for your users and use that. I noticed in analytics, refer traffic, a large number of spam bots or bad sites are accessing our site with 100% bounce and zero session time. Can these affect our Panda score or other rankings? What's up with these? Should we block them? If so, how? From our point of view, this is something that we essentially don't see at all when it comes to search. So we don't use analytics at all for search, for crawling, indexing, and ranking. So this is something we wouldn't even be able to take into account. The other aspect here is that, as far as I know, a lot of these kind of analytics spam bots, they're purely accessing the analytics code. They're not even going to your server. So that's not something where you'd be able to kind of block them from accessing your server. They're just pinging analytics and analytics is tracking that as a view or whatever. And that's not something that we would even see from a search point of view. Is it important to take note of which pages has the most internal links? We're seeing that Contact Us page has the most, but it's not the most important page on a site. Should it always be the most important page that has the most internal links? No, it doesn't need to be that way. So we have a lot of practice with crawling and indexing websites. And it's totally normal for lots of sites to have maybe a Contact Us page, or About Us page, or even a Terms of Service page as being one of the pages that has the most internal links. Because that's something maybe you're always linking to from the footer. Maybe that's always visible in the sidebar. That's completely normal. That doesn't mean for us that this is the most important page on your website. So that's not something where I'd say you need to manually interfere and hide those links to the About Us page or to the Terms of Service so that Google doesn't know about that. That said, if there are important things on your website, I would definitely make sure that there are also links from important pages within your website. So if you're launching a new product or if you have some news that you want to share that you want to rank for, then maybe put that link on your home page so that people can find it fairly quickly, and that search engines understand that this is actually something that they should focus on as well. So it's not the case that we completely ignore internal links, but I just wouldn't worry about the normal website structure where you have a lot of internal links to pages that are probably not the most important ones on your site. How should we tackle low quality pages? I think I've kind of gone through this a few times. Essentially, if you're aware that they're low quality pages, then the best thing to do is to kind of improve them, to make them so that they're not low quality pages anymore. And if you can do that, then that's probably the best approach. If you can't do that, if you think this is really something terrible that you put up on a weekend when you're feeling bad and it's like terrible content and nobody really wants to see it and you feel ashamed for having it on your website, then you can just take it down. Remove it from your site, no index it, redirect it, whatever you want to do with it. If it's really something you don't care about at all, then by all means, remove it. You said that cruft in the head section can cause meta tags to be ignored. We see a lot of websites where metadata is pushed down into the body of the page, but sometimes Google is able to process them. Can you give more details on what tags cause the head section to end? Are there any ways to test this in Google? So as far as I know, there is no simple way to test this within Google's tools. That's something we're working on to try to find some solutions for that, or at least so that you can recognize the situation a little bit better. Essentially, we have a normal HTML parser and we take into account which tags are valid in the head and which tags are only valid in the body. And if we see tags that are only valid in the body, then we assume that the body has silently started. So that's something that is essentially what we're looking for there. And what you could do is maybe run this through a kind of a headless browser and take the HTML that comes out of that and then double check that with the W3C validator for HTML. And then usually you can see fairly quickly which tags are flagged as being invalid and where the head is kind of inherently closed when that page is processed. So that's kind of what I would do there. Obviously, this isn't trivial. So if this is all really crazy talk to you, then one thing you can do is put some of these tags in your sitemap file to make sure that they really don't cause any problems there, in particular the hreflang links. You can put those in the sitemap file and we can pick those up there regardless of whatever crazy stuff you do on your pages directly. What are your best recommendations about thin content? Like how to identify it? We have an adult site that was one of the biggest in Mexico and we received a manual action. We've worked hard to remove that without positive results. So thin content in this regard is probably based on a manual action that, somewhat from the web spam team, has placed on the website. And usually that's based on content where we can recognize that it's not really worth our time or effort to actually index this content. That means either it's really low quality content, it's content maybe that's fun. It's perhaps content that's just great from other sites that's kind of compiled from other sites, the kind of content where essentially we have enough copies of this in our index already and we don't need even more websites that just include the same content. So a common scenario for this is, for example, all of the YouTube scrapers, the video scraper sites that are out there where they take a bunch of videos, maybe from YouTube, maybe from other sources, just jumble them together again and put them on their website and say, hey, this is my website. Essentially, it's just a copy of other people's websites. But you're saying this is your website. So from our point of view, we already have these videos indexed. We already have all of this content available for search. There's no reason for our resources to kind of be used to index the same content again. That's kind of where this manual action comes from. So that's something where I would try to figure out where you might be running into this situation and perhaps get advice from other webmasters who have run websites like that in the past as well, who can point you out and say, well, this is something you really need to fix. This is exactly the same as this other website, or this content is exactly the same, or this is just fun content. It doesn't make any sense when I try to read it. And that's the kind of advice that you'd need to kind of heed. Yeah, someone had a comment. Well, yeah, it was me. I was just going to ask where you draw the line between, we've seen this before, and therefore we won't index it versus there's a real problem here you're getting a manual penalty. Because tons and tons of normal, large e-commerce sites have descriptions that everyone else has. And you can put them in the omitted results, or you can just say, we're not going to bother indexing it more than once a year. But there's a difference between that and here's a penalty, surely. Yeah, so usually the main difference is that there's really no additional value add and that the whole website is like this. So if you're an e-commerce site, then there's some, usually some specific twist that you're offered that's kind of unique to your e-commerce site. You're not just like exactly one out of 1,000 similar websites that have exactly the same product listing, exactly the same kind of experience. Usually there is some unique aspect that you have there. And if the website is really just a complete copy of other websites, if there's absolutely no additional value for us to actually index the whole website, then that's something where the manual action would come in. So it's not something that a normal website would run into if they just have a bunch of products that are the same as other websites. It's really something where the whole website is built up on reusing other people's content. And oftentimes what also comes into play here is that it's not just one website, but you have like a whole set of websites that are set up in the same way. So kind of a. That's what suggests there's no web neutral. You're either adding value and therefore you get some benefit or you're getting a penalty. There's nothing in between. Well, where's the neutral ground? I think there is a lot of neutral ground there. It's usually pretty clear when you look at these websites that there is actually no additional value add here. And it's just kind of a remix of the existing content. So that's something where it's not something you would accidentally run into. And then someone from the website team would think, oh, these five products are the same as on Amazon or somewhere else. And therefore we'll give you a manual action. It's really kind of a quite high bar. So the subjectiveness of this is a manual. Everything you've been talking about, John, is a manual action. And therefore someone's making a judgmental call rather than it being automatic, where everything is automatic. All of us on this side of the conference worry about us. We don't know where the line is. Hopefully that helps the other contributor. So you'd like more information on the automatic side or more information from the manual side? I think I speak for myself. I always want more information on the automatic side. For instance, you can have websites which look at my website and said, oh, you've got thin or duplicated content. And that's just their opinion. And you can adjust your content where I have lots of indexed pages with artists and songs and so on. It's difficult to differentiate them. But we don't know how your index or other indexes view them. So on the automated side, the manual action is clear. And it's subjective. And it's probably good in the instance you're talking about. It's always on the automated side. The closer the more detail we know, which is why we keep poking you for information. It's not that sailing close to the width. It's just not so frustrating not knowing. So where are you running into problems in this regard? Well, I mentioned once a few weeks ago in Marypaw. I've been suffering from panda, as they say, for years. And I really don't know what to do about it. I have worked on a whole lot of aspects of my site. And I've seen some great improvements. And it's very rewarding doing that. But I used to get the same sort of data. I used to get 10 times the amount of traffic before panda was put into place over the years. So it's very, very frustrating for me to know what to do next. So that answers your question. The next bit she's been able to help me here. That's my reply. Yeah, I agree that that's always a tough situation to be in. So that's I don't have a perfect answer for your case. I'm not completely sure which site you mean. But it is something where the quality of a site overall is sometimes really hard to figure out where Google or where users or where algorithms are seeing issues across the board. So I totally agree that's sometimes tricky. Well, I can help you. My site's fab. No, it's great scores. But anyway, I understand your answer. OK. If you can send me a note on Google Plus, I can take a look there again as well. Kind of double check what's happening. I appreciate that. Thank you. With regards to duplicated content, when it comes to the automated side, usually it's less a matter of us indexing or not indexing the content, but a matter of us filtering that out in the search results when we show the search results. So for example, if you have the same product with the same description on two e-commerce sites and we would show exactly the same snippet in the search results, then for us, that's a kind of a bad user experience because users would see the same thing twice. They wouldn't kind of have a variety of search results to click on. And that's where we would try to filter out one of those search results. So that's really the kind of situation where we would index both of these versions. We're fine with knowing that there's a unique spin to each of these websites. And this is something that from our point of view, we might show one a bit higher once and the other one a bit higher once, depending maybe on the user's location or other settings or the query specifically. But if we recognize that, for example, the snippet would be exactly the same, then it makes sense for us to say, well, we'll just show you one of these. And you can click to the end of the search results and see the unfiltered version. But essentially, it doesn't make sense to show both of them to the user because it's both essentially the same thing. I don't know if that applies to your website specifically, but that's kind of the general approach that we take when it comes to duplicate content on unique websites. John, just jumping in for a second. What about things like marketplaces? They're a place somewhere in between where they will have potentially products that would appear duplicated from someone else's website, but there's nothing dodgy going on because it's the original merchant or whatever you're going to call it would be giving permission. We'd often really want people to find their products on that site because the merchant themselves might have a tiny little website. They don't want to have to invest in SEO or promotion or whatever it is. But in that case, how would Google work? I mean, again, if there were too many examples of content that appear to be copied or scraped from someone else's site, might Google jump in there and think, hang on a minute, there's a Panda type issue or a quality type of issue? Or would it simply filter? For the most part, I think we would just filter that. In the search results, when it comes to showing the rankings. So that's not something where I'd see us going in there and saying, well, this description is the same across a bunch of different websites. Therefore, it must be a low quality website. Essentially, if we started doing that, then we would have to treat all of these as low quality websites. And then the whole rankings would be adjusted appropriately, and nothing would really change in the search results. But that's something where we would try to index these pages individually and just filter them out in search if the snippet is really exactly the same as what we would show. But you're saying that a marketplace, which, again, is more likely to have very large amounts of duplicate or what might appear to be scraped content, that Google isn't going to just, it will purely filter. It's not going to take a view that, hang on a minute, we think that there's something more dodgy going on here. We'll actually apply some kind of punitive side as well. Yeah. So I could imagine situations where it is actually scraped content, and that's a bit different. But the normal situation where you have a bunch of users and there's so many user-generated content and some of it is copied on a bunch of different market places because they're trying to sell, I don't know, their old guitars or whatever, that's totally normal. That's not something where I would worry about. But you often have situations, I mean, again, our site's a bit like this, where you'll have a number of sellers and they will have 50, 100, 200 products and they will list those brand new products on a number of market places, one might be Amazon, one might be this one, one might be that one. And it was more, and that content will have been supplied by the seller. So it would appear genuinely duplicate, it might even appear scraped content, but of course there is nothing illegitimate about it. It is kind of genuine content. So I can appreciate that Google might want to make a quantum filter, because it doesn't want to show the same thing lots of times, but it's more whether Google will take the view when it looks at one of those market places that hang on a minute, this looks like duplicate or scraped content, we're actually going to penalize for this, because that's a very different situation to a spam site, scraping content without the permission of the person because they're trying to artificially rank and that's a very, very different situation, so it's just more to understand how Google would look at that. Yeah, I think that's something where we would have to take a look at that manually. So that's something where the manual web spam team would take a look and say, is this really like a scraper scenario or is this essentially just like a marketplace with user generated content and that's completely fine. And by having someone manually take a look at that, usually that's pretty obvious if this is like scraped content or spun content or if it's actually a normal marketplace. So if you can submit something new and it's listed there, it's like these simple things. So that's not something algorithmically would be hit. That's something where... I can't imagine that we would do something specific for that case, so. OK, so anything that happened in that kind of scenario would be a manual action as in someone would have to have a look and would determine that this isn't a genuine marketplace. This is a Samsung kind of spam site and if they saw it was a genuine one, they wouldn't apply the manual action. Exactly, and with manual actions, we would notify you in Search Console. So if you're like, well, they totally misunderstood my website, it's actually like this, then that's something you could reply to or if you recognize that, oh, they looked at, I don't know, a hacked part of my website and I can fix that and do a reconsideration request, that's something you could also do. So that's not something that would silently affect your website without you having any recourse for that. Yeah, I mean, I think there's always a benefit to manual actions. It's more, at least if you get a manual action, you kind of know what you've done wrong. It's more people not really knowing if something's happening algorithmically, that's kind of where the concern is. No. OK, thanks. All right, let me run through some more of these questions here. You said recently that if a site is doing much on topic X, it won't affect the rankings on topic Y. On the same site, will Google consider quality pages for X and not for topic Y? Essentially, we look at a site overall and we look at the pages individually. So we can recognize that some pages are important to rank and it's not so much a matter of which topic these pages belong to and they all need to be on the same topic. But we look at the site overall and we look at the pages individually. So it's kind of a mix of both sides. Which one are your best recommendations about thin content? I think this is the one we just looked at. If a forum produces tons of low quality pages, should we consider removing the forum? Maybe. Maybe. I mean, that's a hard judgment call to make. But that's something that you might consider. So if you think that your forum really doesn't provide value to the rest of your website, then maybe at some point it's worth making that call and saying, well, this was good while it lasted, but the internet has since moved on. Sorry, John, one question about that. The forum could be set to no index. So it would still be available for the users. I wouldn't cut it off and remove it. You can do that as well. That's totally up to you. I think sometimes it makes it harder to maintain if it's no index because you don't have a way of searching across the whole site. You can't easily do a site query and look for spammy keywords and find those spammy posts that someone dropped there. But that's totally an option. That's something you can do as well. I've seen forums also do a mix where they say, these are trusted users. I will allow their content to be indexed. And these are totally new users that I don't know that well that maybe I'll put a no index on by default. Yeah, I don't want the internet to close all of its forums. I think forums are a great place to share knowledge and share ideas, meet other people. But sometimes it's something where a forum might have been a really great idea. It might have been really popular at its time. But in the meantime, nobody really goes there. The content there is really bad. It's essentially just a bunch of spammers that are dropping their links. Then that's the situation where maybe you make this kind of hard judgment call. Google has unofficially brought emojis back to the search results pages. We tried a couple that were relevant to our topic, but they don't seem to work. What's up with that? How do we optimize for emojis? I don't have any input on how to SEO for emojis. This is something where I trust on you all trying crazy things and tweeting about it so that we see what's happening there. I don't have any special insight there on what specifically we were looking for, which exact emojis were picking up and showing in search. But it's always neat to see what people have come up with. Does Google use the linking text or ID of internal anchors to evaluate the content of a page? Should very long articles be broken into multiple pages? Let's see. So we do try to use the links as a way to understand the context of individual pages. And based on that, we try to understand what's the connection between these different pages within a website. And for that, we take into account the anchors on one page that are linking to another page. What we don't take into account is kind of this everything after the hash. So if you have a hash and then some keywords or some extra text that jumps to a specific part of a page, then that's something we wouldn't take into account. I have a website that's using a content fade in the viewport scroll feature. All the content is present in the DOM on page load, but Fetch as Google doesn't see all of the content. The main content area gets rendered, but supporting content like links to other pages don't appear. Similar things happen on article pages for the article posts. Does this hurt my organic visibility? The developer wants to detect the user agent and set the elements to visible for bots. But I'm assuming this is a no-no. Yeah, so I wouldn't do cloaking like recognizing the user agent and doing something special for that. So that's kind of the first comment there with regards to showing the content and kind of fading in the content. That's something where you might want to experiment with different things to see what works best for your users, where they don't get too lost and what works well for search as well. One thing you can do there to kind of double check that we can actually pick up the content on the pages is to do a site query and some of the keywords that fade in. And if we can recognize those, then we can recognize those. So that's kind of the baseline check that I would do there. With regards to not being able to see that in Fetch is Google, that's something where you might want to consider maybe setting the time out to be a little bit lower so that it displays a little bit faster, just so that it's easier for you to diagnose other issues on your website. So not so much that you need to do that for Google, but more to make it easier for you to diagnose other issues that you might run into over time. My question is about publishing videos and associating them with the site. I upload videos for clients on Google Drive and send them authorized links to it. Then the client download the videos. Is there a way to credit the website for the volume of data my company produces, like by having a pass through to the video sorts via link or even storing the video on the web's domain server directly? I'm not completely sure what you mean in that regard. So that's something if you can perhaps elaborate or send me some more details on Google+, I'm happy to kind of take a look to see what might be happening there or what specifically you're looking for. But I'm not really sure what you're trying to achieve there. How much time does Google take in removing no index pages from its index? Usually, once we've processed pages and seen the no index meta tag appropriately, like in the head section of the page, then that's something that we would drop from our index more or less immediately. One thing that might play a role here in the background of this question is that sometimes we crawl pages and it takes a while for us to actually process them, which is more of a technical issue on our side rather than anything else that's happening. So that's something where if we can recognize the no index on a page and we've processed that in our index, then we will remove that page essentially immediately once we see that. If you need the pages removed faster, for example, if you're not sure when it will actually be crawled, then what you can do is also use the URL removal tool in Search Console. Essentially, it removes the page within, I'd say, less than a day around that time frame. Let's see. There is a general question around mobile page speed. I don't see the question here. So let me see. What else? Does Google still count backlinks to sections of our site that have been disallowed by robots.txt? Yes. So what can happen in a case like this is we find a link to a URL on your website. We can't crawl that URL because it's blocked by robots.txt. Therefore, what can happen is we will index that URL without having crawled it based on the link to that URL. So what you might see is that the title is the anchor text of the link that was pointing there. And instead of a snippet, we'll show this placeholder text that we couldn't crawl this page so we don't really know. And that page can collect page rank like that. It can show up in the search results based on that external information. So essentially, that would still count. Depending on what you're trying to do there, you might find a different approach would work better. For example, if you're saying this page is private and you don't want it to be indexed at all, maybe it makes sense to show a login page with a no index on that page instead. Or if you're saying that perhaps you're trying to clear up duplicate content and someone links to the wrong version of your page, then instead of blocking that by robots text, I would let us crawl that page and use a rel canonical to let us know about your preferred version. Yes, I think Deborah, you have a question. Oh, yes, John. Can you hear me? Yes, go for it. Thank you so much for taking my question. I'm trying to take myself off the mute here. OK, I'm seeing a situation with a website that's pretty old and has been doing well. It's meeting most of the Google guidelines as much as one can. It's responsive, blah, blah, blah. I'm seeing an issue around mobile page speed. And I'm getting conflicting information when I'm running it through Google Insights, through the Google Webmaster Tools, or if I'm running it as a full test. And I meet all the criteria responsibly. The site's coming up at a 95 score. For mobile page speed, it's coming up with like a 65 score. And all of the line items that have come through in the reporting are kind of conflicting between Google Analytics showing me and what I'm seeing from the reporting. And my site is definitely the pages are around. I've read a lot of your work on keeping two to three seconds as your guideline. And what's really on my page is a lot of YouTube videos. And I'm wondering, is that something that could be weighing down the page enough to affect the mobile page speed? Maybe, maybe. So from our point of view, we try not to make any exceptions for Google services. So whether or not you're using YouTube or AdSense or anything else on those pages, if that's slowing your pages down, then that's something we will report. So it doesn't matter if it's a Google service or if it's some third party service or you're hosting the videos yourself. Essentially, we would treat that all the same. Is there a way to optimize the video to condense it from the YouTube snippet? I don't know. I don't think so. So usually, the YouTube players are pretty well optimized in the sense that just loading the page won't necessarily slow your page down just because it has YouTube videos embedded as well. It uses a player where it loads the dynamic parts later so that you can click on it and then play the video so that just by loading the page that has the videos embedded, it shouldn't be slowing things down too much. That said, I haven't looked into this specific setup where you have a bunch of YouTube videos on a page and you want to see what the mobile speed is there. How much emphasis, one more question to you. How much emphasis is mobile page speed in the ranking factors right now? If we were to scale that one to 10, is it a five? Is it a 10? At the moment, we don't use it at all. The mobile page speed is not an index ranking factor, is what you're saying? Exactly. It is not a ranking factor. Interesting. I really appreciate you taking my question. Thank you. Sure. OK. So I think with regards to mobile speed, that's something where you'd see more of an effect with regards to what users actually do on your site and less with regards to how we would rank that. So it's a good thing to keep in mind and to focus on so that users stay on your site and you have a fast site. They click around on your site. They can convert in whatever way you're trying to convert them. And that's always a good thing to aim for. But from an SEO point of view, from a ranking point of view, at the moment, we don't take that into account at all. I assume that at some point in the future, we will be taking that into account. Thank you. Sure. All right. Let's see. In the chat, there's something more. What would be the best tool to see how Google is rendering a page? Will Search Console have this option soon for fetch and render? I don't know about soon, but we have been talking about this with the Search Console team to try to find a solution for having the kind of rendered HTML version available in Search Console. So that's something we've been looking at. I really have no idea on the timing there. So probably not going to happen today. Can I have a quick follow up on this one? Sure. Until we see the code. So if there is any delayed content, how does Google sees that? So I don't know if a pop-up gets after a few seconds or a feature of some content is displayed after a few seconds, how Google does treat that? I think there are two aspects there. On the one hand, if you have to load the content later, then probably we won't see that. So if you have an Ajax call, and you're pulling in content from the server, and you're displaying that after some time period, like 20, 30 seconds, then probably we won't see that content at all. On the other hand, if the content is already loaded on the page, and it's just a matter of making it visible or making it invisible, then that's something we'd probably be able to see and take into account. But, John, how does that work with interstitials then? Does that mean that if someone decided to do their interstitial by Ajaxing it loading in afterwards, that Google wouldn't know the interstitial was there, or does Google have some extra special criteria for looking for that kind of cheeky stuff on mobile? Yes, we do. So I guess this is specifically with regards to actually making content visible so that we can index it. But with regards to weird behavior, sometimes we have to go a bit further. OK, thank you very much. Let's see. With regards to 301 or 404, when removing thin content, essentially this is kind of the same. What would probably happen if 301, a lot of the thin content to one main page on your website, is we would pick that up as a soft 404, and we would treat that the same way. So that's something where you probably wouldn't even treat them in a different way. But is there any downside to have a lot of 301? I understand the fact that you might treat most of them as soft 404, but there is a good chance. And in the past, there was actually a lot of them remained as 301s. Is there any downside to have lots of 301s, like the million or so, or even more than that? Million 301s. Go for it. I don't see any problem with that. I think from a user's point of view, it can be confusing if you click on a page or a search result and you're like, I want to see this shoe, and it redirects you to some random apparel site. That's kind of confusing. So that would often be easier. That's going to be clean out anyway at some point, right? So it's not going to be just too much. And the last part of the same question was that let's assume that you cannot delete this content. You cannot remove it. What would be the best way to still keep it, but don't show it to Google because it doesn't make sense? Not to do segregation, I mean, but what would be the best way to basically, I don't know, save crawling budget and save users for not having that in the Google's index? Would that be a no-index? Would it be like if you can group it, would it be a robot 360? You can do robots TXT if you kind of need to save the crawling capacity there, but for the most part, I would just no-index it. And no-index follow would be a golden rule, right? That's fine, yeah. To still flow everything and have discovery, but by the way, and again, there is no limit to how many pages we can have on knowing it. No. Just to make sure. Cool, thanks a lot. Can I ask a question? Sure, go for it. OK, I know that all the answers aren't there yet on the mobile first indexing, but one of the questions I had is for companies with separate mobile sites, it sounds like going forward, those pages are going to be the ones that have all the ranking signals coming into them versus the desktop sites, is that correct? Yeah, more or less. OK, so for companies that have all the ranking signals going to their desktop site, are they going to be at a disadvantage now because now Google has to forward all those ranking signals to the mobile site once it becomes a mobile first index? No. Our goal is really that even if you're set up at the moment with separate mobile URLs where you have maybe an MDOT and your desktop version of your site, that we would just switch to indexing the mobile version of your site, but we would be able to process everything in exactly the same way as your desktop site. So all of those existing signals they apply to your mobile site as well. Just like at the moment, all of your mobile signals apply to your desktop site. OK, so I can't help but be concerned that it's not going to go as well as you just laid it out. Because forwarding signals, for example, with 301s has always been tricky for people who come on these videos and complain that forwarding signals has been hard for them or going to HTTPS has been tricky or redoing their URL structure. So I can't help but be really concerned that it's not going to go as smoothly as we wanted to. Is that a concern that's kind of brewing in the background? Definitely. I mean, that's why we're going so slow with this and doing so many tests. So that's a big concern from our side as well. And it's something where we really need to make sure that no matter how these sites have been set up in the past, we need to be able to keep them in our index. We need to be able to kind of rank them in the same way. Otherwise, we're doing our users a disfavor as well. So just because we kind of have this idea that switching to a mobile first index makes sense at the moment doesn't mean that we can live with throwing away half of the index because they're technically doing something in a weird way. So that's kind of why we're taking this so slow, that we're doing really small tests. We're doing kind of step by step kind of testing things and from smaller to larger ways. And we really want to make sure that this kind of thing doesn't cause any bigger problems. Right. And one other thing that I've been hearing on the videos, the theme from you is that responsive sites aren't going to be affected at all. But don't you think they'll be affected even a little bit? Because when the screen size resizes, the responsive layout looks different. I mean, yes, you can have everything substantially of the same. But to the bot, to the mobile bot, things are going to look just a little bit different enough so that isn't there going to be some sort of trigger where we're going, oh, where this is a new layout I'm looking at? Everything that I've known about this page before, I have to reformulate. And isn't that going to cause some waves, for lack of a better word, in the ranking system now that the mobile bot sees a slightly different look? Some buttons or moves, the navigation is different. And I know you've said before that this is all boilerplate and ignored. But I mean, it's just a computer. It's just an algorithm. It's going to sense some differences. So isn't everybody going to be disadvantaged, even the responsive sites? I'd say for the most part, responsive sites have fewer problems because all their content is still linked from the same pages. So when you look at differences such as internal navigation and you have an m.site that has a very minimal internal navigation and the responsive site that essentially has the same internal navigation across the desktop and the mobile versions, then that's kind of a difference that we would see. Or with regards to structured data, the responsive version has the same structured data as the desktop versions, like all the same stuff. And the HTML. So those are kind of the situations where we would say it's essentially the same thing. And with regards to the algorithms, seeing things a little bit differently when they render the page. Undoubtedly, we'll see some things a bit differently. But from what I've been seeing with the team, that's not something where sites would see any difference with regards to how we would rank and kind of collect those signals. So John, just to have been attaching. I have one more point. Sorry. I'm never on these videos. So let me just take a few questions. Definitely, I was only agreeing with you, but you carry on, sorry. Okay, the last question I have is, so with hidden content and content under tabs right now with the desktop site, you've said that those are discounted in the rankings, but that when we move to mobile, those won't be discounted anymore. But won't the policy reason for discounting that apply on the mobile version too? If you're sending users to mobile sites where content is hidden or they can't easily find what they were searching for, aren't they going to be just as frustrated as before? And that was the reason that you discounted those to begin with? I think we'll have to see how that kind of pans out. At least for the moment, we really want to make it easy for sites to have a mobile version. And in that sense, when we recognize that the content is hidden behind a tab or accessible in some other way, then that's something where we'd say, well, on mobile, there are very few ways to actually do this in a way to always present the full content. So that's kind of why we're taking this more pragmatic approach here. I think in the long run, it's certainly possible that at some point we'll say, well, make sure that your mobile site is good in this specific regard because we've seen lots of complaints from people who go to mobile sites and they feel this is really terrible. So that's something where maybe it'll change, I don't know, in the next couple of years, but probably not in the near future. Okay, thank you. John, just very quickly jumping on to Danny's point as well, which is that, I mean, again, and I think we've talked about this before, but the idea with mobile first also you've said is that things that are kind of hidden through CSS will no longer be completely ignored and will start being looked at. And therefore, that's another reason why responsive site can look completely different when hidden stuff is ignored versus hidden stuff isn't ignored. But you seem to be kind of insinuating there when you're answering the question that mobile first is gonna be looking very largely internal linking, unless so at kind of responsive pages and if the kind of the hiding type stuff. Is that kind of also where mobile first is going? Because people have been looking at mobile first very largely from an individual page point, excuse me, point of view from content. You kind of mentioned the whole kind of concept of the internal linking because again, a lot of mobile sites, pure mobile sites are page light or link light and they kind of look a bit different. Is that kind of where you're coming from a bit? I just think that it's important to keep in mind that a website is more than just the visible content. So in particular, things like the internal links, the structured data, things in the head of a page, all of that is a part of a page too. And just because it's not visible to the user by default doesn't mean it has no importance for search engines. And these are things that you need to make sure are applied to your mobile versions as well. For example, if you have some pages set to no index on your desktop, then obviously the mobile version needs to be no index as well. And that's not something that the user would see but something that search engines would have to pull out. Okay, so again, the takeaway is don't just, because a lot of people is looking at individual pages and nothing else and kind of saying like do I add a hide here or take a hide here or remove a bit of content or add a bit of content but it's holistically looking at the site not just on an individual page by page basis. Exactly, yeah. Let's see, Graham has a question about AdWords. I really have no idea about AdWords. So that's something where I'd recommend going maybe to the AdWords Help forum and double checking with the guys there. All right, it looks like we ran over. We still have a bunch of questions left. So if there's anything with those remaining questions that you really, really want to have answered, make sure to drop that into the next hangout. I think that's on set for Friday. And maybe we'll see some of you then again and we can continue chatting about mobile first or any other topic around search. Thanks again for coming over for asking lots of questions, leaving lots of insights as well. And hopefully we'll see each other again in one of the future versions. Bye. Thank you.