Sunday, July 28th, 2013...9:54 am

The Strange Story of the UK Parliament Site

Jump to Comments

The story began with me looking at the payday loans SERPs and discovering something pretty weird in the second page:

Given the recent government attempts at regulating payday loans (which have become quite a disaster both online and offline), I would have been very surprised, if not for one little detail: it is the homepage of the Parliament site ranking, not any payday loans related page or document (of which there is certainly an abundance there). It’s not even the search results page off the site listing all the existing documents and pages covering the topic. Why the homepage?

Curious about the reasons, I started digging. What I have discovered goes way beyond this little incident and uncovers things I bet none of us has known about the Parliament site.

So, the starting point of my investigation has been MajesticSEO. I first suspected it could be an old style Google bomb by some payday loan company not quite happy about the regulations and had to look at the site’s backlinks to check my guess.

That’s where I discovered parliament.uk is not quite a usual site. Thing is, it has a redirect from its non-www version to the www one but most SEO tools either do not recognise parliament.uk (the non-www version) as a valid domain or struggle making the connection between the www and non-www version.

Usually when you enter domain.com into MajesticSEO you see all links pointing to the whole domain – www, non-www and all possible subdomains. It further suggests you to filter the links by the www version only or by the specific URL of the home page or whatever URL you enter into the tool. Not this time. Here is what I got for parliament.uk:


I was very surprised. Just over a thousand links – only? For a 1996 domain? For the PARLIAMENT site??? Have they just recently switched to that domain from, say, a .gov.uk one?

Apparently not. As confirmed by the Internet Archive, the site has existed on this same domain all the way since 1996.  Failing to believe what I see, I headed over to SearchMetrics to get a second opinion, and my surprise even got greater when that resulted in no joy either:

By this time, I already had the first screenshot of the SERPs posted in my Facebook and Twitter and a few people started discussing it with me. When I complained about the strange picture with links, Rishi suggested I add www to the domain name – and that did the trick. Compare this to the above Majestic screenshot:

SearchMetrics liked this version of the domain better as well, and I have been able to confirm the sudden recent appearance of rankings for “payday loans”:


There were links, too. Some are dead, like this one:


(this was a link off an unrelated site in a hidden div, apparently the site got hacked – whoever did that has already replaced the link to parliament.uk with all sorts of other payday loans and assorted spam related links)

Some are quite alive, and pointing exactly to the home page:

But still, it wouldn’t suit me as an explanation, and there were other strange things I’ve noticed. For example, when running a site: search for the domain, one of the first results showing is a subdomain, ramanujam1.parliament.uk:

 

The subdomain resolves to a duplicate version of the site’s home page. That’s a clear issue right there. Who in the world could have set it up like that?

Further research shows that at some point indeed, parliament.uk’s internal structure used subdomains for various purposes:

That was a few years back, however, and has since been replaced by subdirectories and corresponding redirects from the subdomains, e.g. http://images.parliament.uk now redirects to http://www.parliament.uk/about/images/.

Not quite like the same story with this ramanujam1.parliament.uk subdomain. Not only does it still resolve, it has a lot of internal links pointing to various pages located on it, and this is a much later setup than the rest of subdomains:

Furthermore, the subdomain is hosted on a different IP compared to the rest of the site: www.parliament.uk – 192.221.103.254, ramanujam1.parliament.uk – 62.60.19.79.

Who is this ramanujam1 and how did they end up with a personal subdomain on the parliament site? A Google search has confirmed that this name is probably no less common than Smith, and surely it would be too daft to believe the UK parliament has hired a random freelancer like this one to work on their site, right?

However, all of this, while uncovering a lot of curiosities, does not answer the main question:

Why the Homepage?

There is no clear clue in all the data I have looked at that one could take as the only possible explanation of why it is the home page ranking for such a seemingly unrelated query, hence here are some theories I find more or less plausible.

  1. Google bomb
    As seen from the previous data, I have been able to uncover a number of links with anchor texts related to payday loans. This confirms the theory, as well as lack of previous rankings for the keyword and sudden appearance of the site in these SERPs.
  2. “Trust list”
    With payday loans spam being the big news and Google struggling to clean things up, these SERPs have seen quite a lot of manual action recently. But Google’s rule #1 for any SERPs is they cannot be empty – whatever amount of spam is being cleaned up, Google still needs to rank something in its place, and it better be at least somehow relevant to the query. Similarly to Wikipedia having been the trusted authority source for Google for ages, I would not be very surprised if they created a whole list of sites that can be seen as trusted and clean for this specific niche. The SERPs for payday loans are full of news sites these days, why not the parliament site as well. This does not quite explain the fact that we see the home page instead of any internal page actually relevant to payday loans, or at least the search results on the parliament.uk site for “payday loans” (which, btw, is indexable) – but if we assume that this is another (lazy?) manual action as not a single one of the internal pages has been strong enough to rank by itself, then everything seems to fall into places. Moreover, the current cache of the home page in Google does not show any traces of payday loans but I suppose when the discussions have been taking place in the parliament this has been mentioned as current news and appeared on the home page so that’s an extra reason to rank the site, albeit a bit too late (as of the time of this writing, the site has dropped to the very bottom of the second page of the SERPs – something making this version even more plausible).
  3. Removed hack history
    This is something I have seen quite a lot across top competitive SERPs: a site gets hacked for the sake of parasite hosting, the spammer ranks the page he created on the hacked site, the site owner discovers the problem and cleans up his property, redirects the non-existing spam page to the home page – bingo, an unrelated site’s home page is ranking for a top competitive query. This theory was the reason why I looked so much into the ramanujam1 subdomain. However, apart from the above described weirdness of that whole setup, I have not discovered any traces of previous hacking – according to SearchMetrics, the ranking appeared just recently and just out of the blue, without any other page off the domain previously ranking, and there were no links with payday loans related anchor texts pointing at any specific page that no longer exists on the site, so this version is the least plausible one.

That’s me done digging. I would love to see the input of anybody who can shed some light on this strange story, this surely is an interesting one.