4,643 reputation
1820
bio website johnmu.com
location Switzerland
age
visits member for 4 years
seen 6 hours ago

Webmaster Trends Analyst at Google Switzerland. I work together with the web-search engineers and the webmaster-support team to help webmasters to make awesome content that search engines can understand and recommend.

I primarily post in the Google Webmaster Help forums (in English & German). You can also catch me on Google+, occasionally even in live hangouts :-).


Jul
18
comment Google sitemap priorities, am I doing the right thing?
FWIW Google pretty much doesn't use the priority value at all for web-search. changefreq only makes sense if you're not specifying a date (which you're doing here). Hope that saves you some unnecessary work :)
Jul
16
comment How to disallow robots from the first 185 pages?
Google recommends not using the robots.txt file as a way of handling duplicate content, so using the robots meta tag (or x-robots-tag HTTP header) is preferred.
Jul
16
comment Can Google index a non-80/non-443 URL?
You can definitely do that, but I'd tend towards using subdomains / hostnames: many users don't understand port numbers and might be confused with them (eg how do they recommend the website to others?).
Jul
1
comment Why has Google removed authorship images from its SERPs?
Just to add, authorship (as a concept & the semantics) itself isn't gone, it's really just the profile photos & the circle count. Google is still processing & using authorship markup, we're still showing the name of the author as well.
Jul
1
comment How do you tell search engines not to index this page just yet, but maybe in the future?
FWIW "follow" is the default behavior, you can drop it if you want (it doesn't cause any problems, and maybe it's meant more for humans reading it, so you can leave it too).
Jul
1
comment Should a website be directly accessible by its IP address?
This answer is definitely not true. Many websites are accessible on their IP address, it does not cause search engines headaches, it won't kill your "SEO." Duplicate content is a technical problem that has been around since the beginning of the web, search engines have various ways of dealing with it. It's not the end of the world. That said, if you have a preference regarding indexing, adding a rel=canonical is a great way to guide search engines towards that.
Jul
1
comment Search results for my website show content that isn't mine including pharmacy (Viagra and Cialis)
FWIW sometimes these hacks are only visible to search engines (IP addresses or user-agents), check with a tool like Fetch as Google if you're unsure: googlewebmastercentral.blogspot.ch/2009/11/…
Jun
10
comment Canonical tags for separate mobile URLs
If the primary content isn't equivalent, don't use the rel=canonical.
Jun
10
comment 404 - Page not found error while using Google's Instant Preview in Webmaster Tools
404 in the browser when accessing the tool, or the tool showed a 404 from your website?
Jun
10
comment Can we 301 redirect to a new page, but still publish the old content somewhere else?
Another idea is to have a more-generic page be the main landing page (eg /blue-widgets) which shows the current information, and have the archived pages (/2013/blue-widgets, etc) linked from there. That way, it's easy to focus on the current information (also for search), but still possible to find the rest.
May
28
comment Massive 404 attack with non existent URLs. How to prevent this?
A "404 attack" will definitely NOT affect your site's pagerank, nor its ranking. (If your competitors are spending time linking to pages that 404, that's less time they're spending on doing something useful, so be happy :).) Sites are supposed to have 404s, it's a sign that you have set up the server properly, so if anything, it would be a good signal for us.
May
28
comment Need to add 30K new pages to a 10K page website - troubles ahead? (SEO)
30k pages isn't that much, I'd just put them up at the same time (with the caveat that they should be great pages, otherwise I wouldn't put them up at all).
May
21
comment Are keywords in URLs good SEO or needlessly redundant?
For search, you could use breadcrumb markup to have a readable URL line even if you use parameters / IDs. FWIW googlewebmastercentral.blogspot.ch/2008/09/… is also still very relevant.
May
21
comment Does Google crawl and index sites hosted on an IP address only (with no domain name) and non-standard port?
I'm not sure how these would be shown in Webmaster Tools, so maybe that's a bit confusing (send me an example via Google+ if you want me to check with the team).
May
21
comment Does Google crawl and index sites hosted on an IP address only (with no domain name) and non-standard port?
Yep, we can index sites on IP addresses and/or on non-standard ports.
May
21
comment Why is a query string appearing on my URLs in the Google search results?
Also, don't combine a rel=canonical with the robots.txt - Google won't see the rel=canonical in those cases.
May
21
comment lazy loading images and effects on SEO
I wouldn't rely on "noscript" -- we see a lot of spam trying to use it, so we're not always confident enough to trust content placed there. I like the idea of using links to the images. In general, if an image is one of the primary elements on a page, I'd still tend towards just embedding it naturally (at least one version of it, if you're using different sizes) so that you're sure it's indexed normally. We're working on a better story for this, but I don't see it changing today or tomorrow.
May
14
comment Does using OpenGraph meta tags help webpages to get indexed better?
Google doesn't use Open-Graph meta tags for indexing in web-search. AFAIK we use them for Google+ snippets though ( developers.google.com/+/web/snippet ).
May
14
comment Implications of using subdomain or traditional folders SEO
Subdomains let you split things off to separate servers (maybe in different locations), which can make sense sometimes. However, it also means that search engines will have to figure out crawlrates for each server separately (which might cause inefficient crawling). Using wildcard subdomains makes crawling extremely inefficient, so try to avoid that if you can.
May
14
comment Is it better to have one website with two languages or two websites with different languages in terms of SEO?
Another consideration is that sometimes there are legal aspects that make it important to either get a local ccTLD or not get one.