Sesam

» see sesam.no as it lived, on the wayback machine

~ the goodbye ~

Five hundred million kroner is a strange number to find attached to a search box. That is roughly what Schibsted, the company that owns most of the newspapers Norwegians actually read, spent between 2005 and 2009 establishing that a country cannot have its own search engine. Its own chief executive said the figure out loud afterwards, which is more candor than most of these stories get.

Sesam launched on 1 November 2005 and it was not a toy. The crawling and indexing came from Fast Search and Transfer, the Norwegian search company good enough that Microsoft bought it three years later for something like $1.2 billion. Schibsted wired the box into VG Nett, which is where the country went for news, and open-sourced its own layer on top under the name SESAT.

The plausible part of the strategy was the archive. Google could crawl the open web and so could anybody, but Schibsted owned the newspapers, which meant Sesam could search Norwegian news coverage back to 1983 and nobody else could. That is a real advantage and it is the correct kind of thing to build a challenger on. It just answers a question people ask about twice a year.

In 2005 I was at university in Ankara with a TTNet line at home, and the search argument had been settled for me around four years earlier without anybody making a case for it. That is the part these ventures kept underestimating. Nobody switched to Google after weighing the options. They stopped opening the other ones, gradually, and then could not have told you when.

~ the chronicle ~

~ the money ~

Search is expensive in a way that does not shrink with ambition. You need machines crawling continuously and an index deep enough that the odd queries still return something, and then you need enough marketing on top to get people typing a different address out of habit. Schibsted paid for all of it and employed around a hundred people to keep it running.

The thing it could not buy was the auction. Google's money came from advertisers bidding against each other for keywords, and that auction pays better the more bidders are standing in it. Sesam had roughly 150,000 visitors a day and a fraction of the advertiser pool, so every search it served was worth less than the same search served by Google, permanently, no matter how good the results were. You cannot out-engineer that. You can only outspend it until you stop.

~ the decline ~

By 2007 Sesam.no had 480,000 unique users and sat among the twelve biggest websites in Norway. In a country of under five million that reads like success, and inside the building it must have felt like traction. It was not enough to cover what the index cost.

The Swedish edition is the stranger case. In November 2006 it was being described as the leading search engine in Sweden, and thirty months later it was switched off. Both statements are true and the gap between them is the whole business problem in one line: market position in search is not the same as revenue in search, and Schibsted had bought the first one at the price of the second.

~ the day it died ~

Sesam.no closed on 30 March 2009. The Swedish site followed on 17 June. A hundred people lost their jobs, which for a search product is most of the people who knew how it worked.

The technology did not go in the skip. Schibsted moved the search machinery into Finn.no, its classifieds site, where it went to work on apartments and used cars instead of the open web. Finn is one of the most profitable things any European newspaper group has ever owned. The index that could not sell a keyword was perfectly good at finding a flat in Oslo.

~ what replaced it ~

Losing Sesam is arguably the best thing that happened to Schibsted. The company stopped trying to be a search engine and put its weight behind marketplaces and classifieds instead, which is a large part of why it is one of the few old European media groups that is genuinely rich in 2026 rather than genuinely nostalgic. The retreat was worth more than the campaign.

Sesam.no still resolves. What answers is a Wix error page telling you the domain has not been connected to a site yet. Somebody registered the name meaning to do something with it and then stopped, which is a fairly precise summary of the original venture as well.

@}-->--

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

« previous  ·  random  ·  next »