Google Webmaster Hangout Notes - February 21st, 2020.
Google’s John Mueller hosted a Google Webmaster Central office-hours hangout last Friday, where members of the public had the opportunity to ask questions about search-related topics.
You can watch the full video below, or skip to read my notes.
Mueller is unsure if the legacy Webmaster Tools Crawl Errors report showing 404 sources is coming back 00:45
Mueller was asked whether Google has any plans to bring back the detail in Webmaster Tools showing where 404s originate.
You used to be able to drill down in the report and find out where the link is coming from, but that is no longer an option.
Mueller confirmed that there were some discussions internally about this, but that he does not “know what the current status is.” He said he would “poke at it again.”
Errors relating to Hreflang only affect the pages with errors, not the whole site 01:32
The question was concerning how redirecting or no-indexing individual hreflang pages for deduplication purposes would affect the site.
However, I think the biggest takeaway is how errors only affect those specific pages, not the entire sites’ hreflang implementation.
With hreflang, it’s always a connection between multiple pages, and if any of those pages don’t work, we just ignore those specific pages.
So if they’re like individual language pages that are no index or if they’re redirecting somewhere else or if they don’t have a return tag or anything like that, then we would just ignore that one page.
So it’s not that we would kind of say well something is broken we will ignore everything, or we’ll ignore it for the whole website it’s really just that individual page that we would ignore.
Google generally only picks one type of rich result per page 04:21
A question was asked whether it is possible to include article schema on product pages to add informative text below the products.
Mueller confirmed that it is possible, but he could not see any advantage of doing so. He said that Google “generally picks one type of rich result,” so you would “probably” not see any “big advantage of adding article markup to a product page.”
Adding lots of text to the bottom of product or category pages may trigger a Google’s Keyword Stuffing Penalty 05:08
Continuing from the last question, Mueller added another important point.
Some people think that by adding lots of text on product category pages, they can rank better for those terms. Mueller sometimes see webmasters add five Wikipedia articles of text to ensure all keywords are covered, but that doesn’t help Google at all:
On the contrary, our kind of keyword stuffing algorithms might pick that up and say well there’s just so much random stuff here on this page I don’t know what we should rank it for at all.
I would really focus on informative text that people actually read and see if you can provide value like that rather than just trying to dump as many keywords into a page as possible.
Discover is an algorithmic organic feature 07:02
Google’s Discover is based on an algorithm with Mueller not aware of any ability for Google to block sites manually.
Mueller confirmed that not every website in every country would show up as it would be “based on what users are interested in.”
Alt text for images should be about what is in the image, not for adding page keywords 09:22
Where you have a generic image, of say, an office, on a website offering a specific service, the alt text on the generic image should say what is in the image. It should not be used to add keywords for your service.
Here is what Mueller had to say:
An alternative text for an image it’s not kind of that catch-all place where you dump all of your site’s keywords in.
I would really focus on kind of what you’re showing in those images or what’s relevant with regards through those images.
So don’t […] blindly include all of your sites keywords there because that’s not really going to be useful neither for users nor for our algorithms.
Purchases by consumers on e-commerce sites are not a Google Ranking Factor. 12:03
Google does not know what users purchase on e-commerce sites, so it is hard to use that as a ranking factor.
Breadcrumb Schema on your Homepage provides no benefits 14:35
The purpose of a breadcrumb is to leave a trail back to the homepage. If you put breadcrumb schema on your homepage, there would be no trail, so there would be no advantage in doing so.
Mueller did indicate that it would not do any harm if you had to include breadcrumb schema on your homepage for technical reasons.
A gradual drop in rankings over a more extended period of time could indicate natural ranking changes 17:36
If your website is showing a slow drop in rankings over a long period, you probably are not doing anything wrong that “made everything blow up.”
IF you find in this situation, Mueller recommends that you take a step back and look at the website in general. Mueller says that you should “find areas where you can make significant improvements” to make sure your site becomes “more relevant” for the “kind of users that you are trying to target”.
The Search Console is more accurate than using the site: operator 18:44
In answer to a question about which method is more accurate for seeing the number of indexed pages in the search results, Mueller confirmed that the Search console is “extremely accurate”.
According to Mueller:
- The Search Console is based on all of the indexing they have of the site.
- The site: operator is a restricted query that will only show indexing of that site, but that it is not an inclusive query. It won’t include everything in the index.
Mueller confirms that the site: operator method can be significantly off by a factor of 10, 100, or even more. As such, Mueller recommends using the Search Console Coverage report.
URL Structure to Title Tag is a 7:1 Banana Ratio in importance according to John Mueller 23:39
Google’s John Mueller was asked how important, relatively speaking, URL structure is to Title Tag, in terms of bananas.
Here is what he had to say:
I would give it seven bananas. I think that’s a good number of bananas I should I said a lot of bananas now I don’t know. I think you’re generally asking like is the URL structure more important than a title tag or not and these are vastly different things
So that’s not something where I’d say it’s like you should focus on either one or the other, but rather if you want to work on your site in a holistic way, then try to make sure that everything aligns.
Google may ignore your Search Console geo-targeting for your domain 26:44
Mueller was asked why a webmasters domain was generating a lot of traffic from the US when it has Canada geo-targeting.
Mueller confirmed that Google does not have a mechanism where you can tell Google not to check on a site from a specific country. Their algorithms may assume that your website is globally relevant even if it is geo-targeted or on a country code top-level domain.
What’s happening here in that we think your website is pretty good and we would like to show it to as many people as possible and we don’t realize that actually you’re not able to serve them you don’t want to serve people in those specific countries.
Search Console Report only uses a Sample of the pages on your site 31:45
Mueller states that they only use a sample of pages for their Enhancement Reports in the Search Console. Furthermore, the pages used in the sample can change over time.
Mueller recommends that:
Instead of the absolute number of pages that we report there, I would focus on maybe the percentage or the part of those pages that have issues.
Changing redirects and canonicals can take time for Google to process 37:59
A webmaster asked whether it was ok to change the canonicalization of a URL should the need arise due to updated content on the canonicalized version.
Mueller confirmed that it is “perfectly fine”, but like a redirect, changing the canonicalization can take time for Google to process.
Mueller also recommended that you ensure the following are updated:
- Sitemap files
- Any links within the site are updated
- Structured data, such as Hreflang
Mueller also commented on how that it really should not matter which version of a page is canonicalized as they should all be equivalent. If they are not, Mueller recommends that you might wish to consider your canonicalization strategy.
The last 15 minutes of the hangout dealt primarily with individual site issues, including: