What does it take so long to get listed in DMOZ?

A little bit of fun with huge numbers

[Dynamoo]

[Those of you in the UK who know who Peter Snow is will understand the fun of playing around with big numbers like this ;)]

A common complaint from site submitters to the Open Directory is "Why does it take so long to get listed when there are so many editors?"

The ODP lists 53,359 editors looking after 3.8 million sites. That gives each editor an average of just over 71 sites each to look afer. So how come it can take so long to get processed?

Well, of the 53,000+ editors mentioned probably only 5,000 are active as the larger total includes all editors ever, including those not longer editing.

And out of those, if we assume that 90% are "junior" editors looking after a rough average of 200 sites per category. That would mean that there are 4,500 editors looking after just 900,000 sites.

If that is so, then that would leave 10% of "senior" editors (500) looking after 2.9 million sites, i.e. 5800 sites per editor. (Sanity check - until I got some help recently I had 4800 sites to look after).

In other words, 10% of the editors would be looking after 76% of the sites.

And if the submission rate of 15,000 entries a day is to be believed, then 10% of the editors on average would be trying to process 76% of the entries, which would be 22.8 entries a day, but only 0.8 entries per day in the more junior categories.

OK, now these figures are partly based on guesswork, speculation and rumor, and if you feed different numbers in, then you get very different results, but all of them seem to point to there not being enough active editors to go around.

However, there is a definite split between editors who look after just one or two categories, and those who look after great chunks of the directory.. and although every part of the directory has an editor *somewhere* in the hierarchy, you can see that the spread of editors around the directory is fairly uneven, and that some of the more senior editors have a huge amount of work to do just to keep up with submissions.

What's the solution? A tricky one for the editors and staff to work out I think. The ODP is the largest directory because it applies to most person-hours to building it... another guestimated statistic.. if each site takes five minutes to review, describe and categorise then the directory has taken 36 person years of work just to process the EXISTING entries, never mind the hundreds of thousands of sites going offline that need to be processed and the huge number of sites rejected.

And here's a final statistic for you.. if each listing was charged out at a Yahoo-rate $299 per year, the turnover of the ODP would be $1.1 billion annually. If only they'd gone IPO back in 1999 :)

[JamesR]

>if each listing was charged out at a Yahoo-rate $299 per year, the turnover of the ODP would be $1.1 billion annually

If they charged that much they would only get 1% or less of the submissions :)

[rfgdxm1]

The length of time to get reviewed depends a *lot* where you submit. You can get in quick in many of the small hobby cats. The large commercial cats can take forever, because they are so heavily spammed, and finding people to want these cats is difficult.

[hstyri]

Sanity check - until I got some help recently I had 4800 sites to look after

If this implies you're an ODP editor, it's not a true statement. There's a number of experienced editors doing lots of work all over the directory. ;) (Some of that work being janitorieal work, other tasks more like quality assurance.)

What's the solution?

Note that it may be useful to take a step back and consider what the problem is. One explanation that people complain a lot is that they have too large expectations considering the word "submit" when submitting URLs to the Open Directory. (Why not look up that word in a dictionary?)

Note that the goal of each editor is to list useful sites, not to list all sites that can possibly be listed.

Even more imporant: It's a voluntary project. You may argue as much as you like, but it'll still be a voluntary project. :)

[Dynamoo]

I wasn't arguing :) Just pointing out the huge scale of the directory and why it sometimes doesn't quite work as smoothly as submitters would like.

Another statistic - 50,000+ editor must mean 1,000ish editor being approved per month. Wowsers.

[kctipton]

There's a kernel of truth in these numbers. It wouldn't be hard to do the statistics on a letter of the bookmarks, would it? ( [dmoz.org...] ) Just keep track of who's an editall and who isn't.

As for 1000 editors a month, that's also generally true - but some of those people never log in to make a single edit (and I don't know why they bothered to apply if they didn't really want to edit). Those folks still increase the editor count.

[g1smd]

A few people have guessed, in other threads and fora, that only about 10% of applications are approved. If true, then that makes over ten thousand applications per month. Sure must keep the limited number of people who review these, very busy indeed.

[rafalk]

The average is higher than that.

[Dynamoo]

The math is really more an illustration than anything else. If you feed different number in, you get different results.

However, I wonder how much meta-editor time is taken up with these new editor applications?

Also please note that I'm not actually criticising the directory, not do I have any magic wands to make it better.. but it is a *huge* undertaking and it works better than any of its competitors :)