If Google doesn’t know we’re hosted in the UK, does that affect our SERPs?
-
Hi,
In November 2011 our eCommerce website dropped from between 3rd and 4th position in the UK SERPs down to 7th and 8th. A year after this happened, we still haven’t moved back up to the original ranking despite all our best efforts and we’re looking for a bit of insight into what could have happened. One of our theories is this, do you think it might be the problem?
In October 2011 we moved from a single-site custom built CMS hosted in the UK to a multi-site custom built CMS hosted on a much better server based in the UK. As part of this move we started using CloudFlare to help with security and performance (CloudFlare is a security CDN). Because CloudFlare’s servers are in the US, to the outside world it almost looks like we went from a slow hosting company in the UK to a much quicker hosting company in the US.
Could this have affected our rankings? We know that Google takes the server IP address into account as a ranking factor, but as far as we understand it’s because they (rightly) believe that a server closer to the user will perform better. So a UK server will serve up pages quicker to a visitor in the UK than a US server because the data has a shorter distance to travel.
However, we’re definitely not experiencing an issue with being recognised as a UK website. We have a .co.uk domain (which is obviously a big indicator) and if you click on “Pages from the UK” in the SERPs we jump up to 3rd place. So Google seems to know we’re a UK site.
Is the fact we’re using CloudFlare and hence hiding our real server IP address – is this penalising us in the SERPs?
Currently out of the 6 websites above us, 4 are in the US and 2 are in the UK. All of these are massive sites with lots of links, so smaller ranking factors might be more important for us. Obviously the big downside of not using CloudFlare is that our site becomes much less secure and it becomes much slower. Images and some static content is distributed via a local CloudFlare server, which means it should tick Google’s box in terms of providing a quick site for users.
CloudFlare say in a blog post that they used to have Google crawl rates and geo-tagging issues in the past when they were just starting out, but in 2010 they started working with “the big search engines” to make sure they treated CloudFlare like a CDN (so special rules that apply to Akamai also apply to CloudFlare). Since they’ve been working with Google, CloudFlare say that their customers will only see a positive SEO impact.
So at the moment we’re at a loss about what happened to our ranking. Google say they take IP’s into account for ranking, but by using CloudFlare it looks like we’re in the US. We definitely know we’re not having geo-tagging issues and CloudFlare say they’re working with Google to ensure its customers aren't seeing a negative impact by using CloudFlare, but a niggling part of us still wonders whether it could impact our SEO.
Many thanks, James
-
Hi Des, Thanks, I didn’t know that’s why we were assigned a “special crawl rate” in WMT. Could the special crawl affect our ranking? For example Google puts a lot of weight on freshness, so if Google is crawling us less (we can’t tell if it’s more or less than before), could this make our site look less fresh? We have really tried our best to rule out all other possibilities. Our content is much better and more frequent than it was before and our link building is natural and gradual. We’ve also looked at over optimisation and our competitors. Our competitors are Wikipedia, a couple of national UK newspapers, Harvard, a medical encyclopaedia and a single American competitor. We’re the first UK company to appear the in the SERPs. Whilst these are obviously very big companies, none of them (with the exception of the American company) targets the keyword as much as our website does. Incidentally we did come back up to 4th yesterday but we’ve already a dropped a place today so it doesn’t look like it’ll last. The other thing we found really strange is that the singular version of our keyword didn’t drop at all and has stayed very stable; it’s just in the plural keyword that we dropped. The vast majority of our anchor text is using the plural version (it’s in our brand name) and the domain also contains the plural version. Was there an algorithm change around that time, or maybe are we over optimising the plural keyword? (Is that even possible?) Thanks James
-
Hi SEO5, Thanks for your response. I had come across that forum post before, which incidentally led me to CloudFlare's article about how they work with Google. Maybe I wasn't being explicit enough in my question. We definitely know that Google prefers a faster website and that they have special rules for CDN's. So just to clarify, we were just wondering if there is any way that could using CloudFlare (and therefore not making it clear we’re hosted in the UK) negatively affect our rankings? We’re specifically looking at the UK SERPs ,rather than the US SERPs. Also could you clarify with what you mean with server change? The new server is faster and more reliable, but are there other factors than server speed and server location that Google take into account? We’ve also looked at over optimisation and our competitors. Our competitors are Wikipedia, a couple of national UK newspapers, Harvard and a medical encyclopaedia and a single American competitor. We’re the first UK company to appear the in the SERPs. Whilst these are obviously very big companies, none of them (with the exception of the American company) targetx the keyword as much as our website does. Thanks, James
-
1. Cloud flare is a distributed service with edge servers in many countries around the world. These servers will send your website content from the edge. Users in the uk will most likely be served from a uk server. 2. Google is well clued up on cloud flare and will as you said know that it is dealing with a cdn. Hence in webmaster tools you will not be able to adjust the google crawl rate. 3. In google webmaster tools you can set your desire location, tld that you are targeting. In your case .co.uk. This will tell google I am interested in .co.uk based queries. 4. You should not rule out other possibilities. As sherlock says "when you have eliminated the impossible, whatever remains, however improbable, must be the truth? "
-
Hi James,
The loss of ranking could be attributed to various factors:
- Over optimization penalty
- The changing of the servers
- Competitors outranking you with more quality and aggressive SEO
If you do have a co.uk domain then the changing of hosting from the US to UK could have impacted the rankings.
Here's a link to a discussion on Cloud fare's impact on SEO.
Got a burning SEO question?
Subscribe to Moz Pro to gain full access to Q&A, answer questions, and ask your own.
Browse Questions
Explore more categories
-
Moz Tools
Chat with the community about the Moz tools.
-
SEO Tactics
Discuss the SEO process with fellow marketers
-
Community
Discuss industry events, jobs, and news!
-
Digital Marketing
Chat about tactics outside of SEO
-
Research & Trends
Dive into research and trends in the search industry.
-
Support
Connect on product support and feature requests.
Related Questions
-
One keyword gone in Google SERPs - Fred?
I have an ecommerce site. One keyword, which I use to rank #1 for on Google years ago, I'm now completely gone from the SERP's as of a couple weeks ago. I'm scratching my head here, my other keywords don't seem to have changed much recently. Around mid-March of this year, which seems to line up with the Fred update, I noticed I went from page 3 to middle of page 1 for a few days with this keyword. It was a very happy few days. Then it slipped down and down and hovered around page 6. But as of a couple weeks ago, it's now gone. Before the Fred update, I changed a bunch of product pages within the keyword category that had duplicate content because they were kits of items arranged different ways. So instead of repeating the individual item descriptions over and over in the different kits, I changed the descriptions on the kits to links to the individual items within the kits. After the Fred update, at the end of March, I set all these kit item pages that I reduced to very thin content with just links to noindex. My theory is that the Fred update reset algorithmic penalties for a couple days as it was being introduced. So the penalty of duplicate content that I may have had was lifted since I took out the duplicate content, and I made it back to page one. Then as Fred saw I now had a new penalty of thin content, I got hit and slid back down the rankings. Now that I updated the pages that had very thin content to be noindex, do you think I'll see a return of the keyword to a higher position? Or any other theories or suggestions? I remember seeing keywords disappear and come back stronger years ago, but haven't seen anything like this in a long time.
Algorithm Updates | | head_dunce0 -
Google live blog schema
Live blog markup was rolled out to a selected group of publishers last year 2015. I'm trying to find out whether it has been released to other publishers yet (we are a news site). Not seeing any updates about it anywhere and wondering how I can find out if/when it will be available. This is the latest I can see which is not much help https://developers.google.com/search/pilot/open/live-blogs. Any insider info would be very much appreciated.
Algorithm Updates | | hjsand2 -
Google Update August 2016
Hi Has anyone noticed anything strange with their rankings? I've had a few drops out of the blue at the start of August. Thanks!
Algorithm Updates | | BeckyKey0 -
Is it possible that Google may have erroneous indexing dates?
I am consulting someone for a problem related to copied content. Both sites in question are WordPress (self hosted) sites. The "good" site publishes a post. The "bad" site copies the post (without even removing all internal links to the "good" site) a few days after. On both websites it is obvious the publishing date of the posts, and it is clear that the "bad" site publishes the posts days later. The content thief doesn't even bother to fake the publishing date. The owner of the "good" site wants to have all the proofs needed before acting against the content thief. So I suggested him to also check in Google the dates the various pages were indexed using Search Tools -> Custom Range in order to have the indexing date displayed next to the search results. For all of the copied pages the indexing dates also prove the "bad" site published the content days after the "good" site, but there are 2 exceptions for the very 2 first posts copied. First post:
Algorithm Updates | | SorinaDascalu
On the "good" website it was published on 30 January 2013
On the "bad" website it was published on 26 February 2013
In Google search both show up indexed on 30 January 2013! Second post:
On the "good" website it was published on 20 March 2013
On the "bad" website it was published on 10 May 2013
In Google search both show up indexed on 20 March 2013! Is it possible to be an error in the date shown in Google search results? I also asked for help on Google Webmaster forums but there the discussion shifted to "who copied the content" and "file a DMCA complain". So I want to be sure my question is better understood here.
It is not about who published the content first or how to take down the copied content, I am just asking if anybody else noticed this strange thing with Google indexing dates. How is it possible for Google search results to display an indexing date previous to the date the article copy was published and exactly the same date that the original article was published and indexed?0 -
Google "In-Depth Article" Question
Google started featuring "In-Depth Articles" a few days ago. You can read about them here and here. I have two questions about them... If you already hold a great position in the SERPs. Let's say your existing article ranks at #2 or #3. If that article becomes one of the "In-Depth Articles", will it disappear from the #2 or #3 position? I have lots of content that I could mark as an In-Depth Article, but I don't want to do that if it will pull me out of a hard-earned SERP position. Has anyone seen "In-Depth Articles" that do not have the Schema markup? Thanks!
Algorithm Updates | | EGOL1 -
Does the use of an underscore in filenames adversely affect SEO
We have had a page which until recently was ranked first or second by Google UK and also worldwide for the term "Snowbee". It is now no longer in the top 50. I ran a page optimization report on the url and had a very good score. The only criticism was that I had used an atypical character in the url. The only unusual character was an underscore "_" We use the underscore in most file names without apparent problems with search engines. In fact they are automatically created in html files by our ecommerce software, and other pages do not seem to have been so adversely affected. Should we discontinue this practice? It will be difficult but I'm sure we can overcome this if this is the reason why Google has marked us down. I attach images of the SEO Report pages 8fDPi.jpg AdLIn.jpg
Algorithm Updates | | FFTCOUK0 -
Do Google donations to Wikimedia show any bias in search results?
Over the weekend we learned Google co-founder Sergey Brin donated $500,000 to Wikimedia, the parent company of Wikipedia.com. Last year I believe Google donated $2 million to Wikimedia. I now ask, is this suspicious in anyway, seeing that Wikipedia ranks so well in Google for so many terms? There are several blog posts about it online here and there throughout the years, but what does everyone think?
Algorithm Updates | | DanDeceuster0 -
CTRs on mobile SERPs
Is there any data out there that tells the % of CTRs sites get on mobile SERPs? Thanks!
Algorithm Updates | | bonnierSEO0