UK Web Focus

Innovation and best practices for the Web

Archive for October 23rd, 2007

The UK’s Newest University

Posted by Brian Kelly on 23 October 2007

What is the UK’s newest university? I thought that it was probably Edge Hill University. But I recently discovered that the University of Central England is now BCU – Birmingham City University. I’m assuming this is the UK’s newest university.

What are the implications of changing the domain name for a well-established Web site (http://www.uce.ac.uk/) to something new (http://www.bcu.ac.uk/)? Do you lose your ‘Google juice‘ and have to start all over again in regaining your Google ranking? Or are there techniques you can use which will ensure that links to your old site will be transferred, not only to provide a seamless transition for users but also ensure that automated tools, such as indexing software, will migrate your site’s ranking data, and not treat this as an attempt to masquerade a porn site as a legitimate site.

Anthony Colebourne has described his experiences in a post to the web-support JISCMail list, which summarised what happened when the Victoria University of Manchester and UMIST changed their domain names from http://www.man.ac.uk and http://www.umist.ac.uk to http://www.manchester.ac.uk and The University of Manchester was formed. However as this JISCMail list seems to require a JISCMail username and password I will include his comments here:

We begin with a new http://www.manchester.ac.uk site running in parallel with the old sites. The old sites informed visitors of the change and provided a hyper link to the new site. Very quickly our new site rose up the results listings (without any special effort on our part) to a point where we were competing with ourselves.

Some of our sub domains had setup Aliases of old domains to new ones. However many search engines saw the 2 domains as separate sites. So again these site were competing with themselves for position in search listings and also confusing our users too.

1) The longer domain achieved higher ranking in most cases, possibly this was due to the more relevant keyword in the domain ‘manchester’ as opposed to ‘man’ plus the new site was ‘better’!

2) Our local GSA also indexed everything twice, using up paid for page limits.

3) Our marketing people preferred that the domain in the users address bar to change (i.e. Apache Redirect preferred over Alias).

We configured old addresses to issue Redirect Permanent (301). Firstly for individual sections as we were able, then for everything (/).

We formally merged in Oct 2004, we took down our old sites home pages and redirected them in June 2005. We currently still receive around 500K hits a month to the old domains that get redirected to our new site. We are monitoring usage of the old domains and are not consider removing the redirects until usage drops significantly.

Completely closing down our old domains is a huge task, when you begin to consider non web uses of DNS (email, desktop / server host names etc) and the dependences. It will be many years before our old domains are completely decommissioned. However to the outside world we are now manchester.ac.uk.

These comments are, I feel, very valuable. But what is missing is the implications of a domain name change in a Web 2.0 environment. What will it mean if third party services are used to annotated page on your Web site? What will happen if you have embedded third party content in your Web site, and authenticated based on the URI of the page embedding the content is used? Similarly what will happen to data kept by Web statistics counters?

Answers to these questions will be of interest to many readers, I think. It strikes me that the BCU change may provide a valuable opportunity for research on the implications of changes to a domain name and advice on best practices. An interesting student project, perhaps?

Posted in General | Tagged: | 11 Comments »