ArticlesI am very excited. My experiment to see whether it would be possible to strike back against domain squatters has proven successful – at least in the lab.
Re-Cap
In December last year, I modified the OddThinking theme to include a URL to a non-existent domain. I made the URL invisible to the naked eye, but clearly visible to any bot trawling the source-code. The URL is a commercially-useless name, unlikely to be used for real.
Having set the bait, I waited for a domain squatter to notice the address, realise there was no corresponding domain, and then to register it.
What I wanted to see was whether it would be possible to waste the money of a domain squatter, with minimal cost to me.
Result!
I have just noticed that, in late June, a Samoan domain squatting company registered the artificial domain name, and put a standard search engine/advertising page there.
So, it took six months for someone to take the bait, but they eventually did.
Practical Outcome?
As I described in a comment previously, there is a formula to work out the ratio of the costs to the domain-squatter compared to the costs to the vigilante-blogger.
All calculations are “back-of-the-envelope” and in US dollars.
Cost to Vigilante
The vigilante has bandwidth costs and set-up costs.
Bandwidth
The link added 125 characters to the length of every page. That could probably be halved, or smaller with some simple optimisations. There was approximately 177,000 page views. That translates to about 21 MB of bandwidth consumed by experiment.
This consumed bandwidth has two costs.
The first cost is the incremental cost in hosting fees. With my provider, that is, for all intents and purposes, zero. Other people may be on different plans.
The second cost is that to legitimate users who have to wait a tiny bit longer. Remember that 73% of those page views are from bots – and I can’t see bots getting that bored by 125 extra bytes. The remaining hits are presumably readers, and that means that I have made my poor suffering readers collectively wait for 5.7 MB of download.
That much download must add up to somewhere between 1-6 minutes of download time. How can I rate the dollar value of my reader’s time to me? I am still mulling over the economically correct way to figure that. In the meantime, I am going to be very generous, and call it US federal minimum wage: $US5.15/hour. So I value the lost time suffered by my readers as about 8 cents to 51 cents. Let’s go with 51 cents.
Now what about the set-up costs. I am ignoring the cost of this experiment, and considering the cost of installing and configuring a yet-to-be-written plug-in.
Installation would probably take about 15 minutes. It depends whether a plug-in could work out how to make a link invisible without knowing the theme’s background colour. If it could, installation time would be about 5 minutes.
Rating the average blog administrator’s time at $US25/hour that’s about $US3.13 in configuration costs, worst case.
However, that needs to be amortized over the life of the plugin. I figure the average blog would run for 2 years after the installation (remembering it would still run if the blogger fades, but the site remains). I figure the plugin would run four fake domain baits at once. That’s about 16 baits-taken in the lifetime of one plugin, or a configuration cost of about $US0.20 per domain.
So the total cost to the vigilante is about $US0.71 per domain.
Cost to Domain Squatter
I estimate it costs the domain squatter $3-5 dollars for the basic resource (i.e. domain registration), as much again for 12 months hosting and as much again for the overall business and administrative costs per domain. I have no references here – it is all finger-in-the-air estimates.
So that’s about $9-15 per domain per annum. You don’t need to sell many domains, or that many adverts, to pay back that much.
The Big Question
Would blog-owning vigilantes be willing to sacrifice valuable resources (their time, and their readers’ time) to an upper value of $0.71 to make a bad guy waste valuable resources to a lower value of $9?
Comment by Ben Poweski on September 16, 2006
Nice!
Buying domains from these scum only encourage the practice, but unfortunately the cost of many domains typically exceed the cost to forcefully take back domains based upon trademarks, etc.
I think the only answer to the problem is to increase the cost to register domains.
Comment by Julian on September 16, 2006
Ben,
I discussed this very point recently with a friend registeringa .com.au domain. The annual cost and bureaucracy involved was much, much higher than getting a .com domain.
I asked whether the extra annual cost outweighed the benefits, i.e. the registered company name was still available and the customers of the site could feel more confident that they were dealing with the authentic owner of the the business name.
He wasn’t sure; measuring the value of that benefit is difficult.
As a geek who has registered a .com domain name for a once-off party, I like the fact that domain names can be obtained so cheaply. If anything, I wish the price was lower and DNS refreshes were faster!
Comment by Aristotle Pagaltzis on September 21, 2006
How do you make the causal connection? Couldn’t the squatter have come up with the domain name entirely independently?
Comment by Julian on September 22, 2006
Yes, I’ve given that some thought already.
I tried to come up with a name that wasn’t commercially viable and was unique. I searched for the term in Google beforehand – no hits. I just repeated the search and it only has one hit – the squatter’s site.
I couldn’t publish the name before because then someone could have registered it just to spite my experiment.
I am reticent to publish it now, because if enough people were to check it out, then it would get eyeballs for the adverts and it would be a success!
I figure that, if I was to write some system to automate this, I would have it generate a list of unused names, put half of them into a random control group and never publish them. Then, I could compare the number of control sites that get registered against the published sites that get registered. The difference (if significant) could be attributed to the work of the system.
[By the way, I am repeating the experiment now. I have an anchor tag with only a space in the name, on this very page (near the “login” link. This has been implemented as a hardcoded plugin rather than as a modification to the template.]