Month: December 2021

Server Location

Q. Server location doesn’t affect geotargeting, but it can affect the website speed, and, thus, ranking.

• (00:37) There’s nothing around geotargeting with the server location. From that point of view, server location doesn’t matter so much. Sometimes there’s a bit of a speed difference, so it’s better to look for an opportunity to host the server closer to where most of the users are, since then it tends to be a bit of a faster connection. Sometimes that plays a bigger role, sometimes that plays a smaller role. Sometimes it’s worth trying out. Also, if the content delivery network is used, then oftentimes the content delivery network will have nodes in different countries anyway, and it’ll essentially be the same as if the server was in multiple locations. So from an SEO point of view, from a geotargeting point of view, it’s a non-issue. From a speed point of view, maybe. That can be tested as well. If it’s a critical speed issue that affects the Core Web Vitals and the page experience ranking factor on the website’s side, then that could have a ranking effect. But it’s not so much that it’s because the server is in a different location, rather because the website is perceived as being slow by the users.

Not Ranking For Brand Name

Q. When Google tries to adjust to the search intent of the users, some ambiguous brand names might lose their rankings.

• (09:11) The person asking the question is concerned with the fact that the sexual wellness website he works on, Adam and Eve, stopped ranking for the branded “Adam and Eve” keyword in France altogether. John says that if Google’s systems recognise somehow that people are looking for something that is very different kind of from the style, specifically with regards to kind of this adult content, then that’s something where they might see that even though the website is called exactly like this particular term, there are a lot of people who don’t expect this website will be shown in search, and they might be confused. Or they might find something is wrong. For example, if someone is searching for some Disney character and suddenly there is a sex toy store, then that would be kind of unexpected, and the average person who is looking for that character might be confused about why Google would show that.

503 code

Q. If 503 status is removed in less than a day, then Google automatically picks that up, otherwise the page needs to be recrawled again.

• (14:12) If the 503 status code stays on a page for less than a day, then Google will probably automatically pick that up and retry again. If it’s more than a couple of days, then it can happen that the URL drops out of the index, because Google thinks it’s more of a persistent server error. In a case like that, it essentially needs to wait until Google recrawls that URL again, which can happen, depending on the URL, after a couple of days. Maybe it takes a week or so to kind of be recrawled, but essentially that is something where it goes through the normal recrawling process, and then picks up again.

Automated Translation

Q. Google still has the same stance on automatically translated content in the sense that it’s not optimal to have.

• (25:12) Running the content through Google Translate is viewed as automatically translated content, if it’s an automatic translation without any kind of review process around that. Google’s stance on automatically translated content is still the same, and at least from the guidelines, that would remain like that for the moment. One way that could work together with these guidelines is to, for example, for the most important content, do automatic translation and have it reviewed by someone local, and if it’s alright, then make it indexable. That’s still something different than just taking a million pages and running them through Google Translate for 50 languages and then publishing them.

Hreflang

Q. When a website has different country versions, and there is a need for the users to land on the correct version, it’s always good to use hreflang, as well as make sure there is a way to guide the user to the correct version of the website, in case hreflang doesn’t work properly in some cases.

• (29:06) If there are different versions of content and sometimes wrong versions pop up in search, that’s essentially the situation that hreflang tries to solve, in that there is the content available for different locations or in different languages, and sometimes the wrong one is shown. With hreflang Google can guide that to be more the correct version. The other thing to keep in mind is that geotargeting and even hreflang is never perfect. So if the website is very reliant on the right version being shown to the right users, then there always needs to be sort of a backup plan. John’s recommendation for backup is usually having some kind of a JavaScript-powered banner on top or the bottom or somewhere that essentially states that there is a better version of this content for the user, specifically for his location, for his language and linking it from there. That way, Google can still crawl and index all of the different versions, but users, when they end up on the wrong version, can quickly find their way to the correct version.

Language Versions Of A Website

Q. It doesn’t matter for Google whether only a part of the website is translated for another language version of the website, or the whole website is mirrored in another language version.

• (31:27) First of all, when Google looks at language, it looks at that on a per-page basis. So it’s not so much that Google tries to understand that there is a part of the website in a certain language and the other part is in a different language. Google essentially looks at an individual page, and says, for example,  “Well, it looks like this page is in Spanish”, and when someone searches for something Spanish, Google will be able to show that to them. From that point of view, it doesn’t matter if only a part of the website is translated into a different language. Usually, website owners start somewhere and kind of expand from there. John says that the aspect of internal linking could be a bit tricky in that it could provide a bad user experience if the intern linking is all focused on the first language version, but if there are individual pages that are in that language only, and they’re linked to from another language version, that’s fine. That is something that’s pretty common across a lot of different websites. It’s just, for the most part, it’s good to make sure that the other language version is also properly linked.

Similar Keywords With Different Search Volumes

Q. When there is a situation, where there are slightly different versions of the same keyword (for example, brand name) with different search volumes, it is worth trying to use different variations of the keyword on a page and see what it does.

• (33:31) It’s not so much that Google looks at the search volume of similar keywords and treats them differently, but rather what happens in cases like this is, on the one hand, Google tries to figure out what the tokens that are involved in the query and on the page. Google looks at things like a word level and tries to match those. It also takes into account things like synonyms, where if it can tell that this is a common synonym that people use for different versions of a brand name, then it will take that into account as well, and oftentimes, words that always come together. So if there is a brand that is always one word and the next word, then Google will try to treat that as an entity, and then it will be similar to someone searching for things with one word, for example. So all of these are different ways for essentially Google systems to try to look at that situation and recognise the similarity between the keyword versions and rank them in a similar way. Depending on the actual situation, that kind of very similar keywords might still be different enough to tell them apart. John’s recommendation here is to really look at the search results, and based on that, decide whether it makes sense maybe to mention slightly different versions of the brand name. Because a lot of people look for it, for example, without a space in maybe a specific language. Or is Google already figuring out that these things are the same thing and the search results are similar enough that there’s no reason to do that manually? So that’s something where it kind of depends. These kinds of things need to be tested case by case to figure out what applies and works best. The good thing is that Google doesn’t penalise anyone for using slightly different versions of the brand name on the website pages.

Duplicate And Canonical URLs

Q. John is not sure whether the same URL with a question mark in the end and without one would be treated as the same or separate URLs, but there is a quick way to find out.

• (37:06) For the most part, if there are parameters at the end of a URL or no parameters at the end, Google treats them as clearly separate URLs. However, it does have some systems in place that try to do some almost lightweight canonicalisation for a website owner, in that they try to figure out what simpler version of the URL Google could actually be showing even if the website itself doesn’t provide a rel=”canonical”, or it doesn’t redirect to a simpler version of a URL. So a really common one is if there is a page called index.html and it is linked to, then that’s often the same as just linking to a page that’s called “slash”. So if that’s on the home page, if it’s website.com/index.html, if Google is to see a link like that, it could say that index.html essentially is irrelevant here. It can just drop that automatically, and that kind of canonicalisation happens essentially very early in the Google systems. It takes place without things like rel=”canonical” or sitemaps or redirects, all of those other things. John doesn’t know offhand if just a plain question mark at the end would also fall into this category. However, if there is already a setup on the website, then it can be told fairly quickly if that extra question mark at the end is actually being used by Google because when looking at server logs, one can see if that question mark is there or not. When looking at things like Search Console at the URLs that are shown in the performance report, it is possible to see if that question mark is there or not. If Google doesn’t drop it automatically, and it’s important to have the URL cleaner, it’s good to make sure that at least there is a rel=”canonical” setup to remove that for the website owner.

Adult Content

Q. Google doesn’t really penalise adult content, it’s just Google tries to recognise better whether the search intent is actually adult content or something else with similar terms and keywords.

• (39:45) The person asking the question is concerned with the fact that his adult content website alongside with some other websites similar content-wise have dropped from ranking for some major keywords and is wondering whether that has to do with Google penalising certain types of content. John argues that Google doesn’t really penalise adult websites in that regard, but it does have systems in place that try to figure out whether the intent of the query is actually to find something that would fall into the category of maybe adult content. If the intent is clearly not for someone to find adult content with that kind of query or for the most part not to find adult content, then it is something where Google would try to filter those things out. That’s something that usually makes a lot of sense because sometimes there are adult websites that are named very similar or different types of adult content that are named very similarly to things that are maybe children’s toys or things like that. Google wouldn’t want someone who’s looking for a child’s toy to actually run into an adult toy website just because it’s ranking for the same term. That’s the kind of thing where Google systems try to almost silently figure out what the intent is behind certain queries and then to adjust that so that they show something that matches a little bit more with what the perceived intent is. 

• Understanding what the intent behind a query is really hard sometimes, and sometimes Google gets it wrong. So if there are certain queries where Google has totally messed up, because the intent was to find adult content and Google is not showing any of that content at all and it looks really weird, then those are the kind of things that Google’s team would love to have examples of. So it’s not something that Google has something against adult websites.

Sponsored Links

Q. rel=”sponsored” links don’t help with SEO, they’re basically just advertisements.

• (42:26) rel=”sponsored” attribute attached to a link doesn’t help with SEO rankings. The idea here is essential that the website owner pays someone for that specific link, and it’s a kind of advertising where people can click on that link to go to the website if they like it. If they really like it, they can also recommend it to other people. But the reason for that link being on that other website is because there’s some kind of financial exchange that took place or some kind of other exchange that took place, and it’s not a natural link that Google would take into account for things like search. So from that point of view, it’s fine to have sponsored posts and to have links in sponsored posts. If they’re flagged with rel=”sponsored”, that’s essentially the right way to do it. It’s just that these don’t have any effect on SEO initially. And again, if people go to the website, because they found this link and then they recommend it themselves, then that indirect effect is something that can still be valuable. Oftentimes, especially new businesses  will take kind of an approach to using advertising to initially drive traffic to their website. If they have something really good on their website, they kind of hope that by driving all of this traffic to their website they will get some awareness for the cool things that they have. Then those people share that further, and the website gets some value from that.

Brand Mentions

Q. Google doesn’t really take the brand mentions as a clearly positive or negative signal because it’s hard to tell what the subjective context of the mention is.

• (58:12) It’s hard to use brand mentions or anything like that with regards to rankings. Understanding the subjective context of the mention is really hard. Is it a positive mention or a negative mention? Is it a sarcastic positive mention or a sarcastic negative mention? How can one tell? All of that together with the fact that there are lots of spammy sites out there and sometimes they just spin content, sometimes they’re malicious with regards to the content that they create, all of that, makes it really hard to say if Google can use that as the same as a link. From that point of view, it’s something where for the most part, Google doesn’t mention it as something that positively affects the website or negatively affects the website. It’s just too confusing to use as a clear signal.

Sign up for our Webmaster Hangouts today!

Website Is Down Temporarily

Q.  There isn’t really a way to tell Google that the website will be temporarily down without getting dropped out of the index.

• (04:17)Regardless of what is set up on the website, there isn’t really a way to tell Google the website is down only temporarily and will be back live again after some time. For an outage of maybe a day or so, using a 503 results code is a great way to tell Google that it should check back. But after a couple of days, Google thinks it’s a permanent results code, and that the pages are just gone, and they will be dropped from the index. When the pages come back, Google will crawl them again and will try to index them again. But essentially, during that time, Google will probably drop a lot of the pages from the website from the index. There’s a pretty good chance, that it’’ come back in a similar way, but it’s not always guaranteed. So any time there is a longer outage, more than c couple of days, the assumption is that, at least temporarily, there will be really strong fluctuations, and it’s going to take a little bit of time to get back in. it’s not impossible, because these things happen sometimes, but if there’s anything that can be done to avoid this kind of outage, it’s better to try to do that. That could be something like setting up a static version of the website somewhere and just showing that to users for the time being. Especially if it’s being done in a planned way, it’s advised to try to find ways to reduce the outage to less than a day if at all possible.

304 Response Code

Q. Google doesn’t change its crawl rate on the website based on existing 304 pages.

• (11:48) 304 is a response to the “if modified since” requests, where Googlebot tries to see if this page has changed. 304 response code would not apply to the crawl budget side of things. That basically means for Google that it can reuse that request and crawl something else on a website. The other aspect with regard to crawling that specific URL less – that shouldn’t be the case. Google does try to figure out how often pages change and try to recrawl pages based on the assumed page frequency or update frequency that it has. So it’s not so much that particular URL would get crawled less frequently, it’s more that Google understands it’s better how often these pages change. Then based on that. Google can update its refresh crawling a little bit. If most of the pages on the website return 304, Google wouldn’t reduce the crawling rate. It would try to just focus more on the parts where it does see updates happening. So there’s no need to artificially hide the 304s in the hope that it improves the crawling .

301 Redirects

Q. It’s okay to have a lot of 301 redirects on a website.

• (14:57) A huge amount of 301 redirects on a website is perfectly fine and doesn’t do any harm. If there are changes and redirects made on the website it’s fine.

Launching Big Websites

Q. If a huge website for different country version is to be launched, it’s better to start with fewer country and language versions, and expand incrementally from there in case they prove to be working well.

• (15:52) If there’s a huge new website with lots of country and language versions, it’s hard for Google to get through everything. John recommends starting off with a very small number of different countries and language versions and making sure that they’re working well, and then expanding incrementally from there. With regard to international versions, it’s very easy to take a website and, say, just make all English language versions of the website, but it causes so many problems and makes everything in the whole crawling and indexing and ranking cycle so much harder. .

M-dot Website

Q. Even though Google is not supposed to have any problems with M-dot setup, it’s better to go with the responsive setup for mobile versions of websites.

• (30:20) From Google’s point of view, it doesn’t have any problems with M-dot domains in general, in the sense that this is one of the supported formats that it has for mobile websites. John doesn’t recommend the M-dot setup – if there’s a new website being set up, it’s best to avoid that as much as possible and instead use a responsive setup, but it’s something that can work. So if it’s a regular thing on the website that Google is not able to index the mobile content properly, then that would point more at an issue on the website where when mobile Googlebot tries to crawl, it’s not able to access everything as expected. The one thing that throws people off sometimes with M-dot domains is with mobile first indexing, Google switches to the M-dot version as the canonical URL, and it can happen that it shows the M-dot version in the desktop search as well. So there’s also need to watch out for not only redirecting mobile users from the desktop to the mobile version, but also redirecting desktop users from the mobile to the desktop version. That is something that doesn’t happen when having a responsive setup – it’s another reason to go responsive if possible .

rel=”sponsored”, rel=”nofollow

Q. rel=”sponsored” is a recommended setup for affiliate links, but it’s not penalised in case that is not in place

• (31:59) From Google’s point of view, affiliate links fall into that category of something financial attached to the links, so John really strongly recommend to use this setup, But for the most part if it doesn’t come across as selling links, then it’s not going to be the case that Google would manually penalise a website for having affiliate links and not marking them up, he says

CMS Migration

Q. Fluctuations after CMS migration depend on what the migration covered

• (32:45) When moving from one domain to another, it’s very easy for Google to just transfer everything from that domain to the new one. That’s something that can be processed within almost like a week or so in many cases. However, if the internal URLs within the website are changed, then that’s something that does take quite a bit of time, because essentially, Google.can’t just transfer the whole website in one direction. It has to almost reprocess the whole website and understand the context of all of the pages on the website first, and that can take a significant amount of time. During that time, there will almost certainly be fluctuations. The offhand guess is thet there will be at least a month of fluctuations there, perhaps even longer, especially if it’s a bigger change within the website itself. The other thing is that when the CMS is changed, oftentimes things that are associated with the CMS also change, and that includes a lot of things around internal linking, for example, and also the way that the pages are structured in many cases. Changing those things can also result in fluctuations. It can be that the final state is higher or better than it was before, it can also be that the final state is less strong than it was before. So that’s something where changing a CMS and changing all of the internal linking on a website, changing the internal URLs on a website, changing the design of these pages – these are all individual things which can cause fluctuations and can cause drops and perhaps even rises over time. But doing that all together means that it’s going to be messy for a while. Another this to watch out for with this kind of migration is oftentimes there’s embedded content that is not thought about directly because it’s not an HTML page. A really common one is images. If old image URLs are not redirected, then Google has to reprocess them and find them again because it doesn’t have that connection between the old images and the new ones. So if the site was getting a lot of image search traffic, that can also have a significant effect. Setting up those redirects probably still makes sense, even if the website was moved a month ago or so.

Mobile and Desktop Versions

Q. Having the main navigational links a little different on the mobile and desktop versions of the website is a common thing for responsive design and not that big of an issue.

• (39:24) Having the main navigational links on the desktop version of the website different from the mobile version is a fairly common setup for responsive design, in that both variations are visible in the HTML. From that point of view, that is not super problematic, but it’s probably also not very clean because both variations of the UI have to be maintained within the same HTML, rather than there being just one variation in the HTML that is adjusted depending on the viewport size of the device. From that point of view, it should not be penalised, because it’s a very common setup. It’s probably not optimal, but it’s also not something that needs to be fixed right away.

Old Pages

Q. Whether or not some old pages on a website should be deleted is not only a matter of traffic on these pages.

• (41:00) Deleting old blog posts that don’t get any traffic anymore is up to the website’s owner – it’s not something where from an SEO point of view, there would be a significant change unless these are really terrible blog posts. The main thing to watch out for is that just because something doesn’t have a lot of traffic, doesn’t mean that it’s a bad piece of content. It can mean that it’s something that just gets traffic very rarely, maybe once a year. Perhaps it’s something that is very seasonal that overall when looked at from a website point of view, is not very relevant, but it’s relevant, for example, right before Christmas. From that point of view, it’s fine to go through a website and figure out which parts need to be kept and which need to be cleaned out. But just purely looking at traffic for figuring out which parts to clean out is too simplified.

Image URLs

Q. Changing image URLs is fine, as long as it doesn’t interfere with the typically quite long process of Google crawling and indexing it.

• (43:52) Changing image URLs, for example, adding a query string at the end of an image source URL, wouldn’t cause issues with regard to SEO. But with images in general, Google tends to recrawl and reprocess them much less frequently. So that means that if the image URLs linked on the website are changed regularly, Google would have to refind those again and put them in its image index again, and that takes to take a lot longer than with normal HTML pages. From that point of view, if doing that too often can be avoided, it’s recommended to do so. If it’s something that happens very rarely, and where it doesn’t really matter too much how things are processed in image search, because the website doesn’t rely on image search for traffic to the website, in that case, that’s totally fine. The thing to avoid, especially here with the image URLs is embedding something that changes very quickly. So something like a session ID or just always today’s date, because that would probaby change more often than Google would reprocess the image URLs, and then it would never be able to index any of the images for image search.

Google Discover

Q. There are several reasons why a website can suddenly lose traffic coming from Google Discover.

• (45:38) It’s always tricky with Discover because it’s very binary in that either a website gets a lot of traffic or doesn’t get a lot of traffic from Discover. That also means that any changes there tend to be very visible. The main recommendation is not to rely on Google Discover for traffic, but rather to see it as an additional traffic source and not as the main one. When it comes to Discover, there are a few things that play in there. One of them is the core updates. Google recently had a core update, so it’s good to check the blog post that there is about core updates with lots of tips and ideas. The other thing is with Discover in particular – Google has a set of content guidelines that it tries to stick to in an algorithmic way. Depending on the website itself, it might be something where some of these content guidelines are not followed through, and the website is kind of borderline. For example, there is a guideline around clickbait-y titles or clickbait-y content in general, or adult-oriented content. It might be that a website is a kind of borderline with regards to how Google evaluates it. Then it can also happen that the algorithms see that a large part of the website is just clickbait or one or the other categories that Google lists in the content guidelines. Google then will be a lot more conservative with regard to how it shows the website in Discover. .

One-page website

Q. It’s not always important to be authoritative to provide value with a website. .

• (49:40) The question goes back to one of John’s Reddit posts where he allegedly says that 30-page website can’t be authoritative, and the person asking the question wonders what he would approach one-page websites. John says that it’s possible to make good one-page websites and clarifies that by his post he meant that he actually talked about the reasoning that goes “I created 30 blog posts, and they’re really good, and therefore my website should be authoritative”. From his point of view, going off and creating 30 blog posts doesn’t automatically make the website authoritative. Especially for the higher or the more critical topics, it’s something where it’s not right to just create 30 blog posts on a medical topic and claim to be a doctor. That was the direction he headed there. For a lot of websites, it’s not that the author needs to be seen as an authority. He just puts the content out there. If it’s a small business that sells something, there’s no need to be an authority. Especially things like one-page websites are focused on this one thing, and there’s no need to be an authority to do that one thing, for example, to sell an e-book or to give information about opening hours for a business. From that point of view, having a one-page website is perfectly fine. It’s just useful to think where to go from there at some point – maybe creating more pages and trying to find a way to not paint oneself into a corner by having to put everything on one page all the time, but rather expanding whenever that fits.

Sign up for our Webmaster Hangouts today!

Same Language Content for Different Countries

Q. It’s more reasonable to have one version of the website targeting different countries with the same language, rather than different versions of the same language website.

• (00:47) The person asking the question has a website that has almost the same but separate versions for UK and USA, and he is not sure what the best strategy for managing them is. John says, that having English US and English UK versions means that Google would swap out the URL for the appropriate version depending on the country, So if there is a different content for the two versions, even if that’s something like a contact address or currencies or things like that, then that makes sense to have it on separate URLs. If it’s all the same content, if it’s really just like a text article, then it’s more reasonable to make it one English version. The content can’t be limited to those two countries anyway, so having one version is an easy solution. Another advantage of having one version, apart from having to do less maintenance work, is the fact that for Google it’s a lot easier to rank that one page versus multiple pages with the same content.

Image Formats and Image Search

Q. The way the format of images on the website is changed defines whether or not there might be changes in rankings.

• (06:45) The person asking the question is redesigning her website in AMP pages and is converting all her images to WebP format, and she’s trying to use the same URLs. she is concerned with the fact that converting JPEG images to WebP might affect her rankings. John agrees, that it could potentially affect the rankings. He mentions that he has seen some people use the same image extensions and make them WebP files. If that works, that would help lower the amount of work that needs to be done. Because then the content would just be swapped out, but URLs will be the same and all of that will continue to work. Whereas, if the image URLs are changed, or if the URLs of the landing pages for the images are changed, the image search takes a little bit longer to pick up. The thing to keep in mind here is that not all sites get significant traffic from image search, so sometimes it’s something where theoretically it’s a problem to make these changes that take time.

Deleting JavaScript from the Website

Q. When deleting JavaScript from a page to simplify it for Googlebot, it’s important to not end up giving Googlebot and users a very different page experiences.

• (11:30) If JavaScript is not required for the website pages, for the content and for the internal linking, then probably deleting it doesn’t change anything. John says to be cautious about going down the route of making more simplified pages for Googlebot, because it’s easy to fall into a situation where Googlebot sees something very different than what the users usually see, and that can make it very hard to diagnose issues. So if there’s some bug that’s only affecting the Googlebot version of the site, the website owner wouldn’t see it, if users always see a working website.
  Another thing to watch out for is the fact that in product search, Google sometimes checks to see what happens when users add something to their carts, just to kind of double-check that the pricing and things like that are the same. And if, for example, the Add to Cart functionality is removed completely for crawlers, then maybe that affects kind of those checks that product search does. However, John says that he doesn’t know the details of what exactly product search is looking for.
  In general, Google tries to render the page to see if there’s something missing, but it doesn’t interact with the page, as it would take too much time to crawl the web if it had to click everywhere to see what actually happened. So Googlebot’s experience is different from what users see, and removing JavaScript might affect that.

Featured Snippet

Q. Sometimes it’s hard for Google to determine from the search query whether the user needs local results or the ones in a more global context. The same goes for both the regular search and the Featured Snippets.

• (16:44) The Featured Snippet from Google’s point of view is essentially a normal search result, that has a little bit of a bigger snippet and a little bit more information there. But otherwise it’ a normal search result. And from Google’s point of view, it tries to do two things when it comes to searches. On the one hand, it tries to recognise when a user wants to find something local. And when it recognises that, it uses the geotargeting information that it has from the websites to figure out which are likely the more local results that would be relevant for the user. The local aspect is something that helps to promote local websites, but it doesn’t mean that they will always replace anything that is global. Global in this context might mean bigger websites. So Google sees these global websites, and on the other hand, kind of local results from the same country. And depending on how it understands the query, it might show more local results or more from the global search results. For example, when someone is searching for Switzerland, then, of course, Google recognises that the user wants something from Switzerland, and it can strongly promote local results. But without that addition, sometimes it’s hard for Google to determine whether the local context is critical or not for this particular query. And sometimes it will just take global results in a case like that. And that’s not really something that a website owner can influence.

Website Authority

Q. With the fast-changing dynamics of the Internet, Google doesn’t have a long-term memory of the things that were wrong about the website.

• (22:38) Google pretty much has no memory for technical issues on websites, in the sense that if it can’t crawl a website for a while, or if something goes missing for a while and it comes back, then there is that content again, Google can have that information again, it can show it. And that gets picked up pretty quickly. That is something Google has to have because the Internet is sometimes very flaky, and sometimes sites go offline for a week or even longer, and they come back, and it’s like nothing has changed, but the website owners fixed the servers. And Google has to deal with that since the users are still looking for those websites. 
  It’s a lot trickier when it comes to things around quality in general, where assessing the overall quality and relevance of a website is not very easy, and it takes a lot of time for Google to understand how a website fits in with regards to the rest of the Internet. And that means on the one hand, that it takes a lot of time for Google to recognise that maybe something is not as good as it thought it was. And, similarly, it takes a lot of time for Google to learn the opposite way. And that’s something that can easily take a couple of months, a half a year, sometimes even longer than a half a year. So that’s something where compared to technical issues, it takes a lot longer for things to be refreshed in that regard.
  John also points out that there are these very rare situations, when a website gets stuck in some kind of a weird in-between stage in Google’s systems, in that at some point the algorithms reviewed the website and found it to be absolutely terrible. And for whatever reason, those parts of the algorithms just took a very long time to be updated again, and sometimes that can be several years.
  It happens extremely rarely, especially now, says John. But he suggests that if someone struggles and really sees that he’s doing a lot of things right and nothing seems to be working, it is worthwhile to reach out to Google stuff and see if there is something on the website that might be stuck.

Alt Text and Lazy Load Images

Q. It is not problematic to add alt text to the image that is lazy loaded, even if it’s a placeholder image.

• (26:31) When Google renders the page, it has to or tries to lazy load the images as well, because it tries to load the page in very high viewport, and that triggers lazy loading. And usually, that means Google can pick up the alt text and associate with the right images. If the alt text is already in place, and the placeholder image is currently there, and Google just sees that, then that shouldn’t be a problem per se. It’s kind of like giving information about an image that is unimportant. But it’s not that the rest of the website has kind of like a bad or a worse standing from that point of view. The thing to watch out for here more is that Google can actually load the images that are supposed to be lazy-loaded here. So in particular, Google doesn’t watch out for things like the data source attribute. It essentially needs to see the image URL in the appropriate source attribute for the image tag, so that it can pick it up as an image.

Google Analytics Traffic

Q. Sometimes traffic from Google Discover might create a situation where, when checking the website analytics, lots of traffic comes from direct traffic at random times.

• (32:02) If there is a situation where a huge amount of traffic starts dropping into the direct channel. One of the things that could be playing a role in that case is Google Discover. In particular, Google Discover is mostly seen as direct traffic in Google Analytics. And Google Discover is sometimes very binary, in the sense that it’s either the website gets a lot of traffic or it doesn’t get a lot of traffic from Google Discover. So that could be something where if the website owner is to just look at analytics, there might be these spikes of direct traffic happening there. In Search Console, there’s a separate report for Google Discover, so this kind of thing can be double-checked there.

Shop Ratings

Q. For an e-commerce shop it’ more advised to use product ratings, while for a directory of other e-commerce sites it’s fine to have ratings on those individual sites.

• (34:33) When it comes to shop ratings, Google wouldn’t try to show them if they’re on the shop themselves. So essentially, for local businesses, the idea is that if there is a kind of like a local directory of local businesses, then putting ratings on those local businesses will be fine, and that’s something Google might show and search. But for the local businesses themselves, putting a rating on their own websites is not an objective rating. It can be manipulated to look a little bit more legitimate. And that’s not really something that Google can trust and show in the search results. From that point of view, for an e-commerce shop, it’s better to use products ratings, because the individual products can be reviewed either by the website itself – the shop can clearly specify that it was reviewed by the shop itself. Or aggregate ratings can be used, which would be from users. On the other hand, for a directory of other e-commerce sites, of course, having ratings on those individual sites would be an option.

Search Console

Q. If the number of good pages goes down while the number of bad pages goes up in Search Console, that’s the sign of some kind of problem on the website, if both numbers just go down, it means there’s not enough data for Google to make any conclusion, and that’s perfectly fine.

• (37:35) Essentially, what Google does with regards to the Core Web Vitals and the speed in general – it tracks the data based on what it sees from users, a specific kind of user. That’s documented on the Chrome side. And only when Google has sufficient data from users, will it be able to use that in Search Console, and additionally, in Search Console, it creates groups of pages where it thinks that this set of pages is the same. And if Google has enough data for this set of pages, then it will use that. That also means that if there’s just barely enough data for that set of pages, then there can be a situation where sometimes Google has enough data and it would show it. And sometimes it doesn’t have enough data, and it might show as a drop in the number of good pages. That essentially doesn’t mean that the website is bad, it just means there’s not enough data to tell. So if just the overall number of the pages goes down in Search Console, and over time the overall number goes back up again, then it means there is almost enough data for Google to use. If the number of good pages goes down and the number of the bad ones goes up, that’s a sign that there’s a problem that Google sees there. So if just the overall number goes down, it could be ignored, it’s perfectly fine.

Out Of Stock Pages

Q. How Google will treat out of stock product pages depends on whether it will see them as soft 404 pages or not.

• (44:29) Google tries to understand when a page is no longer relevant based on the content of the page. So, in particular, the common example is a soft 404 page, where there is a page that looks like it could be a normal page, but it’s essentially an error page that says “This page no longer exists”. And Google tries to pick up things like that for e-commerce as well. When out of stock products are seen as soft 404 pages, Google drops them completely from search, if it keeps the page indexed despite being out of stock, the ranking of the page will not change. It essentially will still be ranked normally. It’s also still ranked normally if the structured data is changed to say that something is out of stock. So from that point of view, it’s not that the page would drop in ranking, it’s more that either it’s seen as a soft 404 page or it’s not. If it’s not seen as a soft 404 page, it’s still a normal page.

User Reviews

Q. There are different ways on how to go around showing customers’ reviews about the particular business, but it’s still not possible for the business to show those in the search results.

• (54:03) The customer rating reviews can be put on the homepage. Google would see these more as testimonials because the website owner is kind of picking and choosing what he wants to show. Using structured data for that on the website is something that Google wouldn’t like to see – it would probably ignore that. But sending users to a third-party review site where they can review it is kind of the best approach here. Because what would happen then is if Google shows that listing from that third-party review site, it can show the structured data for the business there. For example, the business is listed on Yelp, and they leave the reviews there. When that Yelp listing is shown in search, Google will know that this is a list of businesses essentially, so it can show the structured data about the business. And then it can use that review and markup and show those stars in the search results. So if Yelp has structured data, then that’s something that Google could pick up there. Showing it on the website as well in a textual form is perfectly fine.
  And testimonials are very common and popular, but it’s just Google wouldn’t show the stars in the search result for that. 

Sign up for our Webmaster Hangouts today!

Mobile Friendliness

Q. Pages that are not mobile-friendly don’t fall out of being indexed

• (00:40) Even if website pages are not completely mobile-friendly, they should still be indexed. Mobile-friendliness criteria is something Google uses as a small factor within the mobile search results, but it definitely still indexes those pages. Sometimes this kind of issue can come up temporarily where Google can’t crawl one of the CSS files for a brief time, as then it doesn’t see the full layout. But if these pages look okay when being tested manually in the testing tool, then there isn’t really a problem and things will go back to normal eventually.

Title Length

Q. On Google’s side, there are no guidelines on how long a page title should be

• (03:02) Google doesn’t have any recommendations for the length of a title. John says that it’s fine to pick a number as an editorial guideline based on how much space there is available, but from Google’s search quality and ranking side, there aren’t any guidelines that state some kind of required length. Ranking doesn’t depend on whether the title is shown shorter or slightly different – the length doesn’t matter.

URL Requirements

Q. Length of URL and words contained in URL matter mostly for users, not for SEO

• (04:53) URL length doesn’t make any difference. John says that it’s good practice to have some of the words in the URL so that it’s a more readable URL, but it’s not a requirement from an SEO point of view. Even if it’s the ID to a page, it’s okay for Google too. It’s good to have words, but it’s essentially something that just users see. For example, when they copy and paste the URL, they might understand what the page is about based on what they see in the URL, whereas if they just see the number, it might be confusing for them.

Doorway Pages

Q. If a website has a very little number of similar landing pages, they are not considered doorway pages

• (14:41) The person asking the question is worried about the fact that his seven landing pages that target similar keywords and have almost duplicate content would be flagged as doorway pages and would be de-listed. John explains that with just seven pages, he probably wouldn’t have any problems, even if someone from the Web Spam Team was to manually look at that. They would see that it’s seven pages, not thousands of them. It would be different, if someone, for example, a nationally active company, had a separate page for every city in the country. Then the Web Spam Team would consider that as beyond acceptable and problematic, where they would need to take action to preserve the quality of the search results.

Reviews Not Showing Up in SERPs

Q.If reviews on a page are not left on the page itself, but are outsourced from some other website, they’re not going to show up in SERPs

• (21:35) For a review to show up in the search results, it needs to be based on a specific product on that page and it needs to be the thing that a user left directly on that page. So if a website owner was to archive reviews from other sources and post them, then Google wouldn’t pick those up as reviews for the structure data side. These can be kept on the page, it’s just Google wouldn’t use the review markup for that.
  It’s a tricky process because Google tries to recognise this situation automatically and sometimes it doesn’t recognise it and shows the review. There are some sites that have these reviews shown because Google didn’t recognise that it was not left on the site. But from a policy point of view, Google tries not to show reviews that are left somewhere else and are copied over to a website.

Search Console Verification

Q. It’s possible to have a site verified multiple times in Search Console

• (24:47) In Search Console, it’s possible to have a site verified multiple times, as well as to have different parts of the site verified individually. It doesn’t lose any of the data when the website is verified separately. That’s something, where it’s okay to have both the host level as well as domain level verification running in Search Console.

Crawling AMP and non-AMP Pages

Q. Google tries to keep a balance between crawling AMP and non-AMP pages of a website

• (26:42) Google takes into account all crawling that happens through the normal Google Bot infrastructure and that also includes things like updating the AMP cache on a website. So if there are normal pages as well as AMP versions and they’re hosted on the same server, then the overall crawling that Google does on that website is balanced out and that includes AMP and non-AMP pages. So, if the server is already running at its limit with regards to normal crawling and AMP pages are added on top of that, then Google has to balance and figure out what it can do there – which part it can crawl at which time. For most websites, that’s not an issue. It’s usually more of an issue for websites that have tens of millions of pages, Google barely gets through crawling all of them and when another kind of duplicate of everything is added, then it makes it a lot harder. But for a website with thousands of pages, adding another thousand pages from the AMP versions is not going to throw things off.

Indexing Process

Q. The way Google indexes pages and the way request indexing tool work have changed over the past few years

• (32:27) In general, the ‘request indexing tool’ in the Search Console is something that passes it on to the right systems, but it doesn’t guarantee that things will automatically be indexed. In the early days, it was something that was a lot stronger in terms of the signalling for indexing, but one of the problems that happens with this kind of thing is that people take advantage of that and use that tool to submit all kinds of random stuff as well. So over time Google systems have grown a little bit safer in that they’re trying to handle the abuse that they get, and that leads to things sometimes being a bit slower, where it’s not so much slower because it’s doing more, but it’s slower because Google tries to be on the cautious side. This can mean things like Search Console submissions take a little bit longer to be processed, it can mean that Google sometimes needs to have a confirmation from crawling and kind of a natural understanding of a website before it starts indexing things there.
  One of the other things that have also changed quite a bit across the web over the last couple of year, is that more and more websites tend to be technically okay in the sense that Google can easily crawl them. So on the one hand, Google can shift to more natural crawling and on the other hand, that means a lot of stuff it gets, it can crawl and index, which means because there’s still a limited capacity for crawling and also for indexing, Google needs to be a little bit more selective there, and it might not be picking things up fast.

Pages Getting Deindexed

Q. Some pages are being deindexed as new pages are added to the website is a natural process

• (37:02) For the most part, Google doesn’t just remove things from its index, it kind of picks up new things as well. So, if there’s new content added at the same time and some things get dropped on along the way from the index, usually that is normal and expected. Essentially, there are pretty much no websites that Google indexes everything on the website. It’s something where, on average, between 30 and 60 percent of a website tends to get indexed. So, if there are hundreds of pages added per month and some of those pages get dropped or some of the older or less relevant pages get dropped over time, that is kind of expected.
  To minimise that, the value of the website overall needs to be shown to Google or the users, so that Google will decide to try and keep as much as possible from the website in the index. 

Website Migration

Q. After a few months post website migration, it’s better to remove the old sitemap from the old website

• (41:58) Usually, when someone migrates a website, they end up redirecting everything to the new website and sometimes they keep a sitemap file of the old URLs in Search Console with the goal that Google goes off and crawls those old URLs a little bit faster and finds the redirect. That’s perfectly fine to do in a temporary way, but after a month or two, it’s probably worthwhile to take that sitemap out because what also happens with the sitemap file is it tells Google which URLs are important. Pointing at the old URLs is almost the same as indicating that the old URLs need to be findable in search and that can lead to a little bit of conflict in Google systems because the website owner is pointing at the old URLs but at the same time, they’re redirecting to the new ones. Google can’t understand which ones are more important to index. It’s better to remove that conflict as much as possible, and that can be done by just dropping that sitemap file.

Spider Trap

Q. Whenever there are spider trap URLs on a website, Google usually ends up figuring them out

• (46:06) If there is something on a website, like, for example, an infinite calendar where it’s possible to scroll into March 3000 or something like that and essentially one can just keep on clicking to the next day and the next day, and it’ll always have a calendar page for that, that’s an infinite space kind of thing. For the most part, because Google crawls incrementally, it’ll start off and go off and find maybe 10 or 20 of these pages and recognise that there’s not much content there, but think that it will find more if it goes deeper. Then Google goes off and crawls maybe 100 of those pages until it starts seeing that all of this content looks the same, and they’re all kind of linked from a long chain where someone has to click “next”, “next”, “next” to actually get to that page. At some point, Google systems see that there’s not much value in crawling even deeper here because they found a lot of the rest of the website that has really strong signals telling them that those pages are actually important compared to the really weird long chain on the other side. Then Google tries to focus on the important pages.

Multilingual Content

Q. When there is multilingual content, it’s advised to use hreflang to handle that correctly

• (53:13) In general, if there is multilingual content on a website, then using something like hreflang annotations is really useful because it helps Google to figure out which version of the content should be shown to which user. That’s usually the approach to take. 
  While with a canonical tag, Google knows which URL to focus on. So the canonical should be the individual language versions – it shouldn’t be one language as a canonical for all languages. Each language has its own canonical version – like there is the French version and the French canonical, the Hindi version and the Hindi canonical. So it shouldn’t be linking across languages.

Sign up for our Webmaster Hangouts today!

No-Index and Crawling

Q. If a website had no index pages at some point, and Google hasn’t picked up the pages after they became indexable, it can be fixed by pushing the pages to get noticed by the system

• (08:18) The person asking the question is concerned with the fact that there are a handful of URLs on his website that at some point had no index tag. A couple of months have passed since the removal of no-index, but Search Console still shows that those pages have a no-index tag from months ago. He resubmitted the sitemap, requested indexing via Search Console, but the pages are still not indexed. John says that sometimes Google is a little bit conservative with regards to submitting indexing requests. If Google sees that a page has a no-index tag for a long period of time, then it usually slows down with crawling of that. That also means that when the page becomes indexable again, Google will pick up crawling again, so it’s essentially that one kind of push that’s needed.
  Another thing is that, since Search Console reports on essentially the URLs that Google knows for the website, it might be that the picture looks worse than it actually is. That might be something that could be seen by, for example, looking in the performance report and filtering for that section of the website or the URL patterns to see if that number of high no-index pages in Search Console is basically reporting on pages that weren’t really important and the important pages from those sections are actually indexed.
  Sitemap is a good start, but there is another thing that could make everything clearer for Google – internal linking. It is a good idea to make it clear with internal linking that these pages are very important for the website so that Google crawls them a little bit faster. And that can be a temporary internal linking, where, for example, for a couple of weeks, individual products are linked from the homepage. When Google finds that the internal linking has significantly changed, Google will go off and double-check those pages. That could be a temporary approach to pushing things into the index again. It’s not saying that those pages are important across the web, but rather that they’re important pages relative to the website. So if the internal linking is changed significantly, it can happen that other parts of the website that were just barely indexed, drop out at some point, so that’s why changes in the internal linking need to be done on a temporary level and changed back afterwards.

Canonical and Alternate

Q. Rel=“canonical” indicates that the link mentioned is the preferred URL, rel=“alternate” means there are alternate versions of the page as well.

• (14:25) If there’s a page that has rel=“canonical” on it, it essentially means that with the link that is mentioned there is the preferred URL and the rel=“alternate” means that there are alternate versions of the page as well. For example, if there are different language versions of a page, and there is a page in English and a page in French there would be the rel=“alternate” link between those two language versions. It’s not saying that the page where that link is on is the alternate but rather that these are two different versions and one of them is in English, one of them is in French, and for example, they can both be canonical – having that combination is usually fine. The one place to watch out a little bit is that the canonical should not be across languages – so it shouldn’t be that on the French page there is a canonical set to the English version because they’re different pages essentially.

Rel=“canonical” or no-index?

Q. When there are URLs that don’t need to be indexed, the question whether to use rel=“canonical” or no-index depends on whether these pages need to be not shown in search at all or if they need to be most likely not shown in search.

• (16:49) John says that both options, rel=“canonical” and no-index are okay to use for the pages that are not supposed to be indexed. Usually, what he would look at there, is what the strong preference is. If the strong preference is not wanting the content to be shown at all in search, then a no-index tag is the better option. If the preference lies more with everything being combined into one page and if some individual ones show up, it’s not important, but most of them should be combined, then a rel=“canonical” is a better fit. Ultimately, the effect is similar in that it’s likely that the page won’t be shown in search, but with a no-index it’s definitely not shown, then with a rel=“canonical” it’s more likely not to be shown.

Response Time and Crawling Rate

Q. If crawling rate decreases due to some issues, like high response time, it takes a little bit of time for the crawling rate to come back to normal, once the issue is fixed

• (20:25) John says that the system Google has is very responsive in slowing down to make sure it’s not causing any problems, but it’s a little bit slower in ramping back up again. It usually takes more than a few days, maybe a week or longer. There is a way to try and help that: in the Google Search Console Help Center, there’s a link to a form where one can request that someone from the Google team takes a look at the crawling of the website and gives them all the related information, especially if it’s a larger website with lots of URLs to crawl. The Googlebot team sometimes has the time to take action on these kinds of situations and would adjust the crawl rate up manually, if they see that there’s actually the demand on the Google side and that the website has changed. Sometimes it’s a bit faster than the automatic systems, but it’s not guaranteed.

Indexed Pages Drop

Q. Indexed pages drop are usually have to do with Google recognising the website content as irrelevant

• (26:02) The person asking the question has seen that the number of indexed pages has dropped on her website, as well as a drop in the crawling rate. She asks John if the drop in crawling rate could be the cause of indexed pages drop. John says that Google crawling pages less frequently is not related to a drop in indexed pages, indexed pages are still kept in the index – it’s not that the pages expire after a certain time. That wouldn’t be related to the crawl rate unless there are issues where Google receives 404 instead of content. There could be a lot of reasons why indexed pages drop, the main thing that John sees a lot being the quality of these pages. Google’s systems kind of understands that the relevance or quality of the website has gone down and because of that, it decides to index less.

Improving Website Quality

Q. Website’s quality is not some kind of quantifiable indicator – it’s a combination of different factors

• (34:35) Website quality is not really quantifiable in the sense that Google doesn’t have Quality Score for Web Search like it might have for ads. When it comes to Web Search, Google has lots of different algorithms that try to understand the quality of a website, so it’s not just one number. John says, that sometimes he talks with the Search Quality Team to see if there’s some quality metric that they could show, for example, in Search Console. But it’s tricky, because they could create separate quality metrics to show in Search Console, but then that’s not the quality metrics that they could actually use for search, so it’s almost misleading. Also, if they were to show exactly what the quality metric that they use, then on the one hand that opens things up a little bit for abuse, on the other hand, it makes it a lot harder for the teams to work internally on improving these metrics.

Website Framework and Rankings

Q. The way the website is made doesn’t really affect its rankings, as Google processes everything as HTML page

• (36:00) A website can be made with lots of different frameworks and formats and for the most part, Google sees it as normal HTML pages. So if it’s a JavaScript based website, Google will render it and then process it like a normal HTML page. Same thing for when it’s HTML already in the beginning. The different frameworks and CMS’s behind it are usually ignored by Google.
  So, for example, if someone changes their framework, it isn’t necessarily reflected in their rankings. If a website starts ranking better after changing its framework, it’s more likely due to the fact that the newer website has different internal linking, different content, or because the website has become significantly faster or slower, or because of some other factors that are not limited to the framework used.

PageSpeed and Lighthouse

Q. PageSpeed Insights and Lighthouse have completely different approaches to a website assessment and pull their data from different sources

• (37:39) PageSpeed and Lighthouse are done completely differently in the sense that PageSpeed Insights is run on a data center somewhere with essentially emulated devices where it tries to act like a normal computer. It has restrictions in place that, for example, make it a little bit slower in terms of internet connection. Lighthouse basically runs on the computer of the person using it, with their internet connection. John thinks that within Chrome, Lighthouse also has some restrictions that it applies to make everything a little bit slower than the computer might be able to do, just to make sure that it’s comparable. Essentially, these two tools run in completely different environments and that’s why often they might have different numbers there.

Bold Text and SEO

Q. Bolding important parts in a paragraph might actually have some effect on the sEO performance of the page

• (40:22) Usually, Google tries to understand what the content is about on a web page and it looks at different things to try to figure out what is actually being emphasised there. That includes things like headings on a page, but it also includes things like what is actually bolded or emphasised within the text on the page. So to some extent that does have a little bit of extra value there in that it’s a clear sign that this page or paragraph is considered to be about a particular topic that is being emphasised in the content. Usually that aligns with what Google thinks the page is about anyway, so it doesn’t change that much.
  The other thing is that this is to a large extent relative within the web page. So if someone goes off to make the whole page bold and thinks that Google will view it as the whole page being the most important one, it won’t work. When the whole page is bold, everything has the same level of importance. But if someone takes a handful of sentences or words within the full page and says that these words or sentences are really important and bolds them, then it’s a lot easier for Google to recognise these parts as important and give them a little bit more value. 

Google Discover Traffic Drop

Q. There can be different factors affecting traffic drop in Google discover: from technical issues to the content itself

• (47:09) John shares that he gets reports from a lot of people that their Discover traffic is either on or off in a sense that the moment Google algorithms determine it’s not going to show much content from a certain website, basically all of the Discover traffic for that website disappears. Also in the other way, if Google decides to show something from the website in Discover, then suddenly there is a big rush of traffic again.
  The kind of issue people usually talk about is on the one hand quality issues, where the quality of the website is not so good. With regards to the individual policies that Google has for Discover – these policies are different from web search ones and the recommendations are different too. John thinks that it applies to things like adult content, clickable content etc, all of which is mentioned in the Health Centre Page that Google has for Discover. Sometimes a lot of websites have a little bit of a mix of all of these kind of things, and as John suspects, sometimes Google algorithms just find a little bit too much and then it decides to be careful with this website. 

Response Time

Q. The standard for response time for a website doesn’t really depend on the type of website, but rather on how many URLs need to be crawled

• (50:40) The response time is something that plays into Google’s ability to figure out how much crawling a server can take. Usually, the response time from a practical point of view limits or plays into how many parallel connections would be required to crawl. So if Google wants to crawl 1000 URLs from a website, then the response time to spread that out over the course of a day can be pretty large, whereas if Google wants to crawl a million URLs from a website and a high response time is there, then that means it will end up with a lot of parallel connections to the server. There are some limits with regards to the fact that Google doesn’t want to cause issues on the server, so that’s why response time is very directly connected with the crawl rate.

Sign up for our Webmaster Hangouts today!