Google Directory

» see directory.google.com as it lived, on the wayback machine

~ the last words ~

I have used Google since about 2000 and I cannot remember clicking the Directory link once. That is most of what needs to be said about why it died. It sat there in the row above the search box, next to Images and Groups, a door into a hand-built map of the web, and the box beside it was simply faster at answering the question I had walked in with.

It launched in March 2000, and it was not Google's directory in the sense of Google having made it. The contents came from the Open Directory Project, the volunteer-edited catalog Netscape had started in 1998 that everyone called DMOZ. At the point Google took the feed, DMOZ held roughly 1.5 million sites across more than 200,000 categories, kept by something over 22,000 unpaid editors who argued with each other about where things belonged.

What Google added was the ordering. DMOZ listed the sites inside a category alphabetically, which meant a page called Accelerated International MBA Program came before Harvard. Google reordered every category by PageRank, so the sites other sites linked to floated to the top. This was the first time PageRank was visible to the public in any form, and Larry Page described it then as working something like peer review for the web.

~ the record ~

~ the funny part ~

Google Directory was the company hedging against its own thesis. In 2000 it was not obvious that keyword search would beat a catalog, because Yahoo had been built on a catalog and Yahoo was the front page of the internet. Google ran both models side by side for eleven years, and one of them had been quietly winning since about week two.

The directory also had a use nobody advertised. Getting listed in DMOZ was a ranking signal, and a DMOZ listing that surfaced in Google Directory carried a visible PageRank bar next to it. That turned a volunteer library into a bottleneck for an entire industry, with SEO forums full of people waiting years for an editor to approve a submission and accusing those editors of taking money to move faster.

~ the actual end ~

There was no announcement. On 20 July 2011 the URL started returning a message telling you that web search is the fastest way to find what you need, and that if you would prefer to browse a directory you should visit dmoz.org. Google closed Google Labs the same day, which got all of the coverage, and the directory went out underneath it.

The phrasing of that goodbye is worth reading twice. Google did not say directories were finished. It said go use the one this had always been a copy of, which was true, and which was also a way of noting that Google had never owned the thing it was closing.

~ the long tail ~

DMOZ itself lasted another six years on inertia, closing on 17 March 2017 when AOL stopped paying for it. Google had already stopped using DMOZ descriptions in search snippets by then, which removed the last commercial reason anyone had to care about being listed in it.

What went with the directory was a claim about the web that had been true when it was made. In 1998 a few thousand volunteers could plausibly read and file the interesting parts of the internet. By 2011 that was no longer arithmetic anybody could do, and the last serious attempt at it had spent a decade running inside the company that made it unnecessary.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »