The domain clear-out

15 March 2019
For the last 18 or so months I have been disposing of many of my domain name registrations, something that was strongly implied in a previous article regarding the price of UK domain registrations. I had come to the conclusion that most of the domains I had registered were really just baggage for which the notional use-cases were unlikely to be put into practice — many were registered between 2006 & 2008 and had more often than not never been used. Back in the summer of 2017 when I did the first batch of registratrion cancellations I had around 60 domains, whereas today this is down to 39 domains, and I expect to only have 20-25 left by the end of the year — I estimate that this reduction will save me at least half a grand a year, which at the very least is more than I could realistically expect to profit from holding onto the domains.

Why so many domains?

Back in the summer of 2017 when I did the first batch of registratrion cancellations I had 61 domains, whereas today this is down to 39 domains, and I expect to only have 20-25 left by the end of the year. About half of the domain registrations consisted of sets of domains, typically including most of the major suffixes: .com, .org, .net, .co.uk .org.uk, and & .eu. This was a common thing to do back in the mid-2000s and was a combination of anti-cybersquatting, a desire for subtly-different domains for different purposes, convenience as some people remember the domain but forget which prefix, and outright vanity. Auto-renew played a major part in me still having all the domains years later. Most were on two-year renewal cycles, so even when there were domains that I considered surplus to requirement, the long time before I had the opportunity to kill them off made me forget about them. The earlier half of 2017 ended up being unusually busy for renewals, so when I finally decided to actually start cancelling registrations, many had recently been renewed for 2018 & 2019.

Declining value of domains

However it is looked at domain names simply do not have the value they had 10-20 years ago, and these days money that would in the past be spent on a securing a good domain name is in most cases simply better spent on online advertising. The location bar in most modern web browsers is really a search field so typing in the wrong suffix is less of a concern, and in the last few years there has been such proliferation of novelty top-level domains (.info, .media, .london, etc) that people no longer bother even keeping track of them. With the introduction of the top-level .uk domain it looks like interest in .co.uk & .org.uk is going through the floor, which would explain why they are practically being given away for new registrations. When the top-level .uk became available I got nine of them for free on special offer from having the lower-level domains, and these account for almost all the domains I obtained 2013 onwards.

For me personally geographic domains have far less appeal than they used to do, and I have particular ire towards UK domains due to inflation-busting price rises. The latter combined with it now being the best part of seven years since I stopped working in the UK made me decide to simply dump the vast majority of my circa 30 UK domain names in favour of the non-geographic ones. I also had a few semi-geographical EU domains but I dropped almost all of them as I could not see any convincing use-case for them, the one exception being dhondt.eu which is in active use for a site I plan to develop further. Even though they are geographic I am holding onto the likes of pgu.org.uk and remy.net.nz because domains that short are hard to come by.

Manageability issues

Just keeping track of 50+ domains is problematic, and an audit I did in 2008 or 2009 revealed that almost all the registrations had at least one error, most commonly the telephone number that was probably never correct in the first place. Spin forward a decade and a surprisingly large number still had the address of my Ph.D research lab, and yet somehow they still passed the cross-validation with “third party sources” Nominet did in 2012. In the case of bristolshooting not all of the domains were correctly setup — in other words correct DNS entries and correct web-server configuration so putting the domain into a web-browser did not go to the right place — which defeated the purpose of having the full set. I even had one instance with a different domain where a stale DNS entry ended up pointing at someone else's website. I concluded that I should at least reduce my holding of domains to something I could at least give due attention to.

A major motive for maintaining multiple domain names was as a way to filter emails — for instance a .net domain for hosting-related stuff, a .com domain for company-related signups, and others for things like university societies or use on job-site CVs. This both kept tabs on where my email address was being obtained from, as well as being able to keep apart emails based on relative importance that would not be affected by undisclosed-recipient monkeying. In reality two domains became a sink for most stuff, and in any case a combination of the GDPR and advances in email filtering mean that junk rendering any given address unuseable is no longer a problem. Half the domains I have setup for email I only keep that way because they are linked to things like alumni mailing lists that are a pain to have changed, and I certainly do not intend to expand the set of domains that are actively wired up for email. For me email is a communication medium of decreasing importance.

In hindsight

A pruning of domain registrations was well overdue, especially with the spiralling cost of renewals, but this wholesale clearout is the result of a broader divestment from running my own internet-related services. I have long discounted web development and system adminstration as a professional career path, and my personal technical interests (read: hobbies) are deviating from these areas as well, so I felt it time to close things off. When I took the decision to cull a lot of domains had recently-ish auto-renewed, and they were opn a mix of one- and two-year renewal cycles, so it has taken a long time for the cancellations to filter through.