Helium.com

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

~ the eulogy ~

i missed helium at the time. it was a US writing community at a moment my english output was still forum posts and homework drafts. helium peaked around 2008 to 2010, when i was deep in the engineering curriculum at university in ankara. the idea of writing fifteen dollar articles in english for a US site would have read as a strange freelance economy from somewhere far away. by the time i would have noticed the category, the panda update had already broken it.

mark ranalli launched helium in andover massachusetts around 2002 as a writing community where members both wrote articles and rated each other's work. the rating system was the trick. for every topic, members compared articles head to head, and the result fed a leaderboard that decided which piece sat at the top of the category page and which sat at the bottom. revenue share went with the rank. about seventeen million in venture money came in across rounds, and the contributor base grew past one hundred fifty thousand at peak with a million articles in the database.

rr donnelley took an early stake in 2008 and bought the rest in june 2011 for roughly seventy million dollars total. the donnelley acquisition put helium under a print conglomerate that wanted to use the content factory to feed customer publishing. the same year, google rolled out panda. helium's traffic fell because its long-tail commercial articles were exactly what the algorithm was built to demote. new articles per day dropped to about a fifth of pre-panda volume, then to a tenth. on may 14 2014, the helium team announced closure. the site stopped accepting articles on may 21 and went fully dark on december 15 2014.

~ the dossier ~

~ how the rating system worked ~

helium's distinguishing piece was the peer-rating layer. you submitted an article about a topic that already had multiple submissions, then you were asked to rate two other pieces against each other in a head-to-head comparison. enough comparisons over time ranked all the articles in a topic from top to bottom. the top piece got the largest share of search-driven ad revenue. lower-ranked pieces earned a smaller share or nothing. members were also asked to evaluate new submissions before they went live, so the writing community doubled as the editorial layer.

the design assumed members would behave honestly when rating peers. in practice the ranking drifted into politics and reciprocity, with small cliques rating each other's work upward. the rating layer was also slow. it took dozens of comparisons before a new article reached its true ranking, and during that time the piece earned almost nothing. by the late 2000s most contributors treated helium as a tier-three outlet behind better-paying platforms.

~ what panda did ~

when google rolled out the panda algorithm in february 2011, helium was exactly the kind of site it had been built to demote. long-tail informational articles in templated shape, written quickly for search traffic. helium's google referrals dropped, and the revenue share to writers fell with it. the company tried microsites and quality tiers. the final pivot was a freelance-marketplace called helium publishing services. none of it replaced the search-driven volume.

~ what got buried with it ~

the helium archive was never fully migrated. members were given a window between may and december 2014 to download their own articles before deletion. some pieces moved to medium or to private wordpress sites. most of the archive vanished. the rating-driven peer-evaluation idea reappeared in quora and stack exchange in cleaner form, but helium's specific version of pay-per-rank writing did not survive the panda cohort.

@}-->--

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

« previous  ·  random  ·  next »