Archive for the 'Web Development' Category

Is your web server’s location damaging your rankings?

I’ve been asked to do some search engine optimisation for classical guitar shop, Kent Guitar Classics. We’ve only just begun the keyword research phase, so don’t flame me for the site’s current SEO!

What I noticed whilst conducting that research, is that even for the name of the business (usually an easy number one spot unless you have a very generic business name), the site only comes second when using the “pages from the UK” option in Google. The number one result using “pages from the UK” is a page on the Venezuelan UK embassy’s website! As you would expect, Kent Guitar Classics comes first if you just search “the web” using google.co.uk. Here are a couple of screenshots for posterity:

 

Kent Guitar Classics web search

 

Kent Guitar Classics UK search

 

A bit of investigation using a whois service like Domain Tools shows that the website is hosted in Oslo!

Kent Guitar Classics whois lookup

 

Why is this important? Well, Google’s search results are biased according to the country in which the search is being performed. This is because it knows that most searchers are looking for something local to them. Google uses lots of information to decide whether a site is in the same country as the searcher: the domain extension (e.g. .co.uk), the postal address on the site (if it can find one), the geographic-targeting setting in Webmaster Tools, links from local websites and quite possibly numerous other factors.

One other factor is the physical location of the web server, i.e. if it is hosted in the same country. Clearly, in Kent Guitar Classics’ case, it isn’t – it’s hosted in Norway. As a result, one of the big pointers Google uses to determine a site’s country of origin is way off. Naturally, I have advised Miles at Kent Guitar Classics to move server.

 

An interesting aside I noticed while researching the site’s setup is that for some reason, the deafult homepage for www.kentguitarclassics.com is index.html, but the homepage appears to be index.asp. This could be another problem for Google, as it doesn’t like “bounce”-type redirects. A quick disabling of Javascript and meta refresh tags using the excellent Web Developer Toolbar for Firefox means that I can see this page:

Kent Guitar Classics redirect page - click to enlarge

Kent Guitar Classics redirect page - click to enlarge

 

Examine the code, and there is a Javascript redirect to index.asp – not something that Google will take particularly kindly to. This could be because the developer originally used index.html and when the change to index.asp was made, they didn’t want to break peoples’ bookmarks, so they used a redirect to ensure everyone still got the homepage.

This is one of the problems with Windows web servers running Internet Information Server (IIS) – there isn’t an easy way to create permanent (301) redirects, because the .htaccess files used by Apache (the usual web server on Linux machines) mean nothing to IIS. Instead, you either have to code the redirect into the page using ASP, or make changes directly in IIS (or install an ISAPI filter), which on anything but a dedicated server, the host won’t let you near.

That’s a completely separate problem to the physical location of the server, but I thought I’d mention it whilst looking at that site! :)

Google Chrome experience

Just read this short article that I think is a good overview of Chrome from a user’s perspective, following last week’s release.

Google Chrome Browser – The Third Way?

After much speculation over the years, Google have announced the release of their own web browser, Chrome, via this online comic book.

They’re claiming greater stability, more speed, better security and a new approach to browsers based on their 21st Century use (social media, interaction, video etc.) Like Firefox, it will be open source, meaning other developers can work on the browser and make their own improvements, plug-ins etc.

One of the big things is multi-threading – allowing the browser to get on with doing other things whilst waiting for another process to complete. This will make interaction much quicker and smoother, but the real need for this comes from the use of the browser as a home for applications. If you’ve used Google Docs or other online applications, you’ll be familiar with their restrictions – a multi-threaded browser takes it much closer to an operating system that can support “proper” applications. This is Google taking the big fight to Microsoft.

It looks like this will be the download URL when Chrome is released tomorrow: gears.google.com/chrome/?hl=en. Meanwhile, you can see preview screenshots and videos on TechCrunch.

So what does this mean? Well, it will be interesting to see how and if Google continues to support Firefox. That’s unlikely to change for a while, as TechCrunch reported that they have outbid everyone to be the default search engine in Firefox for the next three years. However, you can’t see that being the case forever more.

Microsoft is bound to get excited – Internet Explorer 8 is in development and will apparently have some new features, but nothing to rival the “ground-up” approach that Google has taken. If Chrome starts making our popular websites and applications run faster/better (think Facebook, YouTube etc.), Microsoft has a problem, never mind the enormous marketing clout that Google can exert.

Finally, spare a thought for the humble web developer, who will now have three PC browsers plus Safari and other minor players to develop for! Testing across several versions of IE and FF is already a depressing task, adding a completely new rendering engine to the game could have significant impacts on development time (although with Google’s test-driven development, that’s unlikely). Nevertheless, there will be a period of time where some websites run better on some browsers than other, or have more functions/work faster on Chrome while the others catch up.