Blekko

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

~ the goodbye ~

blekko was the kind of search engine developers and seo people tried once and then forgot to come back to. the slashtag pitch was clever in the abstract. you typed your search query and appended /news or /tech, and the index would only return results from a curated source list rather than the open web. in theory it solved the spam problem that google was fighting with panda updates and ranking changes. in practice the average user did not want to learn a new search syntax to get cleaner results.

i was a working engineer in '10 by the time blekko opened to the public, doing a fair amount of search benchmarking out of professional habit. i tried blekko a couple of times the same way i had tried cuil two years earlier, then kept using google. that pattern was the blekko user funnel in miniature: tech-press attention at launch, a curious trial, then attrition back to whatever search bar was already in the browser address bar.

rich skrenta had the credibility to build this. he had co-founded dmoz, the original open directory project that became the volunteer-edited spine of search relevance in the late 1990s. he had founded topix, a news aggregator that gave him a working model of how editorial curation could improve over pure algorithmic ranking. blekko was the third move in that career arc, an attempt to bring that curation philosophy to general web search at scale.

the launch on 1 november 2010 got the press it deserved. blekko cracked time magazine's top 50 sites of 2011. it raised over $50 million from yandex and the usual venture set. at peak the index covered the top three billion pages of the web and the slashtag library had a few thousand user-curated topics. it was, by any reasonable measure, a working search engine. it just was not a google replacement, because being a google replacement turns out not to be a problem you can solve from the outside.

~ the chronicle ~

~ what it was ~

the slashtag idea was the differentiator. you could append any one of thousands of community-built tags to a query. /conservative or /liberal applied a political-leaning filter to news results. /healthtrusted limited a medical query to mayo clinic and a small set of vetted sources. /spam (more accurately, the implicit anti-spam filter that ran on every blekko query by default) used a manually-built blacklist to suppress content farms.

the anti-spam stance was the loudest pitch at launch. blekko marketed itself as the search engine that did not return demand media or ehow garbage. that resonated in the months after google's may 2010 mayday update because content farms were a visible problem. but in february 2011 google shipped the panda update, which slammed the same content-farm sites blekko had been blocking by hand. that took the wind out of the differentiator. by mid-2011 blekko's anti-spam pitch was no longer unique.

~ how it actually ended ~

the 27 march 2015 ibm acquisition was a soft landing. ibm wanted the crawler and the engineering team for watson, with the categorization tech as a secondary asset. the press release framed it as a technology pickup, not a search-engine purchase, and the public blekko.com search service was discontinued the same day with redirects routing former users to ibm's watson developer materials. the acquisition price was undisclosed and most accounts put it well below the $50-plus million blekko had raised.

skrenta and parts of the team moved to ibm watson. some of the blekko engineering ended up surfacing in later ibm watson search products. the watson brand itself had a difficult mid-2010s after the famously oversold healthcare deployments, and the blekko technology became a footnote inside a footnote. by 2026 the bigger story around ibm watson is its long decline through the watsonx rebrands, not the blekko absorption.

blekko sits next to [[cuil]] in the 2008-2015 cluster of pre-deep-learning search startups that pitched themselves as a fix for one specific google problem and got bought for parts. cuil tried to fix index coverage. blekko tried to fix spam. neither had a real path to a sustainable user base because the average searcher does not actually rank either of those problems above default-browser convenience.

@}-->--

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

« previous  ·  random  ·  next »