I just read the blog post respect for DMOZ at the DMOZ blog. The anonomus editor glippitt compares DMOZ to Google and Wikipedia. This editor compares DMOZ to Google and Wikipedia and provides an argument for why DMOZ is a better resource for information about Somalia. Glippitt final claim is that DMOZ is useful for broad topics.

Nobody is searching in DMOZ

I understand Glippitt. I can imagine that it might be frustrating for an editor to spend hours on an directory that is hardly used.

I honestly can not remember the last time I did I search in DMOZ, if I were to take a wild guess it must have been more than a year ago, and I am a person with a special interest in web searching. And as a webmaster and former consultant I have and have had access to a lot of web referer statistics, across the board traffic from DMOZ, including the thousands of sites that mirror whole and parts of the directory normally would account for just a fraction of a percent of the total traffic. I guess there must be some exceptions to this rule. I would assume that sites listed in DMOZ categories that are ranked well in search engines like Google might get some traffic from DMOZ, but this would just be indirect search engine traffic.

The DMOZ search results are both good and bad

DMOZ does list some nice resources for single word searches like Cake, but is the results better than the results for the same search in Google or Wikipedia, perhaps they are? However normal searchers do not use single word search phrases all that often. Back in 2004 most searchers used two and three word search phrases, and there is no reason to believe that that has changed. So let us compare the three for the search Birthday cake: DMOZ, Google, Wikipedia and Birthday cake recepies: DMOZ, , Wikipedia. Are the DMOZ results better? In my opinion they are not bad, atleast better than the Wikipedia results, but not nearly as useful as Google.

So where is it DMOZ fails? They do not correct misspellings, take the search for birtday cake and compare it to the same Google search. And if the unlucky searcher where to search for how to make a birthday cake, he or she would get no results compared to the 4.1 million results in Google. And for common local searches like carpenter in Miami DMOZ only return a single irrelevent result, again compare this to Google.

DMOZ still affects Google and other places people actually search

The huge benefit for a webmaster to have his or hers site listed in DMOZ is the massive amount of links that are generated from a single DMOZ listing. Even though few searches are using DMOZ and the huge number of DMOZ mirrors these sites are still crawled by Google and other search engines, the search engines find the links and are using them to boost the search rankings of the sites listed in the directory.

Power in the hands of anonymous editors

It would not be a bad thing that DMOZ affects the results in Google if DMOZ was a superb quality resource and if few or none of the editors abused the power they have to list and deny listing of links on the thousands of sites that are using DMOZ data. Even though the value of a link from DMOZ in terms of Google rankings has decreased over the years due to changes in Google’s ranking algorithm, it is still reasonable to question the power of these editors.

Anyone with a valid e-mail address can apply to become a DMOZ editor, there are no background checks, and nobody can know the true motive of a new editor. Once accepted the editor can pretty much do as she or he likes in the part of the directory that they edit. They have the power to list sites that they are affiliated with and avoid listing sites that are competing with the sites they are affiliated with. And since there are no transparency of the selection process and since nobody knows who hides behind an editor nickname it is very hard to uncover and to prove abuse of editor power. Historically bribing DMOZ editors eighter directly or indirectly was often the only way to get listed in certain parts of the directory, I am not sure if that has changed in the last 18 months, but I some how doubt it.

A future for DMOZ?

The editor community are working on improving the directory. Two of these projects are “Project 27″ which aims to change the directory model and introduce tagging and DMOZ v.2 which is currently in beta.

It is good that they are prepared to change this dinosaur of a directory, but I fear that it will not be enough. What really could change the directory into something that would be used by real people was if they adopted an editor model similar to the Wikipedia editor model, or if they went ahead and got rid of the whole editor community and replaced them with professional editors.

It would also be useful for the searcher if the directory also presented data from other sources whenever appropriate, atleast when the DMOZ data alone can not provide a good user experience.

Ok, this was my first blog post in 14 months. But that does not mean that I will start blogging again. Then again I might.