By now, most people in Germany will have noticed the changes to DENIC's registration terms. Since 23 October 2009, it has become possible to register all .de domains without the previous restrictions. Before that date, one- and two-character domains, as well as domains consisting only of numbers, could not be registered.
This policy change triggered intense competition for these highly sought-after domains. Many of them were considered extremely valuable, with some estimated at €10,000 or more. In practical terms, securing one of these names felt less like ordinary registration and more like a lottery. That challenge alone was enough to draw me into the process.
I focused primarily on two-character domains and the newly available registration plate domains. My first step was to generate a complete list of all possible combinations:
aa
ab
ac
ad
...
zz
00
01
...
99 After that, I added all three-letter German registration plate abbreviations to the list. Creating the list was straightforward; the real challenge was evaluating it and prioritizing the most promising domains for registration.
Some registrars introduced a reservation feature, which added another layer to the process. At the time, there were 273 accredited registrars, and each registrar could register up to four domains per minute. That means a theoretical total of 1,092 domains per minute. With approximately 1,466 relevant domains in play - calculated as 36² plus roughly 170 registration plate domains - the entire pool could, in the worst case, have been exhausted in about one minute and twenty seconds.
Fortunately, domain registrations do not happen in a perfectly uniform way. Most people tend to go after the most obvious and memorable names first. If that distribution could have been modeled accurately, it would have been possible to predict the remaining opportunities far more effectively. Still, the best practical approach was to rank the domains as carefully as possible in advance.
My first step in that ranking process was to run a Google lookup for every candidate term, which already provided a useful initial filter. However, that alone was not enough. A domain's value is also influenced by its structure. For example, "los" is clearly more attractive than "a3y". In domain terminology, sellers often distinguish between patterns such as LLL and CVC, both of which are generally preferable to something like LNL. To reflect that idea, I wrote a small PHP function to assign a simple structural score:
<?php
function rateDomainName($str) {
$score = 0;
$extra = 0;
for ($i = 0, $len = strlen($str); $i < $len; $i++) {
$extra+= ord($str[$i]);
switch ($str[$i]) {
case 'a': case 'e':
case 'i': case 'o':
case 'u':
if ($i === 1 && $len === 3) {
$score+= 1.0;
} else {
$score+= 0.8;
}
break;
case '0': case '1':
case '2': case '3':
case '4': case '5':
case '6': case '7':
case '8': case '9':
if ($i === 1 && $len === 3) {
$score+= 0.1;
} else {
$score+= 0.2;
}
break;
default:
$score+= 0.5;
}
}
if ($extra === $len * ord($str[0])) {
return $score + 2.0;
} else {
return $score;
}
}
?> Combining the result of rateDomainName() with the number of Google search results improved the ranking considerably, but it was still not enough. The final step was to maintain both a blacklist and a whitelist. Certain names were simply too problematic - for example o2.de, 4u.de, or other names that could raise trademark or ownership issues. Domains like that were likely to attract attention from many people, and in some cases could even waste a registrar's limited registration window.
Time, however, remained the decisive factor. To maximize my chances, I created accounts with as many registrars as possible and prepared my domain lists in the exact order I wanted to submit them. In the end, all that work led to the same result: I did not secure a single domain.
Even so, the experience was revealing. It showed quite clearly that success in this kind of land rush was less about ingenuity than about resources. DENIC may be a non-profit organization, but many of the 273 registrars likely earned substantial amounts from this small policy change by effectively turning queue position into a commercial advantage. The result was predictable: a wave of speculative registrations and thousands of new parked domains.