Microsoft Tag

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

~ the goodbye ~

microsoft tag was an attempt to replace the qr code with something prettier. instead of black-and-white squares, tag used colored triangles in a 5x10 grid. the format was called hccb, high capacity color barcode, built inside microsoft research by gavin jancke. on paper it could hold more data per square inch than a qr code. in practice almost no one outside a microsoft marketing meeting ever decoded one.

tag launched at ces in january 2009. magazines began printing the little triangle blocks in ad layouts a few months later. you were supposed to install the microsoft tag reader app, point your phone at the page, then watch a url resolve via microsoft's servers. that last clause was the problem. tags were not standalone data. they were pointers into a microsoft-hosted directory, which meant if microsoft ever turned the directory off, every tag printed anywhere in the world would become a colored decoration.

tag was a us-coded product at the magazine-and-advertiser layer. it lived in the kind of print ad campaigns american magazines were still running in 2010 and that mostly never reached turkey at scale. by the time i was scanning anything off a page it was a qr code with a built-in os scanner on android. i remember reading about tag in 2009-2010 tech blogs and thinking it was the kind of thing microsoft research did, then losing track of it the way you lose track of small product announcements that have no follow-on news.

~ the chronicle ~

~ the breakout ~

tag had a real marketing window. magazine publishers liked it because the triangles could be made very small and still scan, which let them fit a tag inside an ad without taking up the white space a qr code needed. for about eighteen months you could find tags in pop-up ads, restaurant menus, real estate flyers, and the inside pages of business-magazine cover stories. microsoft pitched it as the format that would finally make camera-based barcode scanning a consumer behavior in the us.

it did not become that. the underlying problem was that qr codes were free and open, with working open-source decoders on every os. tag required a proprietary stack. for any publisher choosing between two roughly equivalent technologies, the open one wins.

~ the decline ~

by 2012 the format was clearly stalling. iphone and android scanners had absorbed qr decoding as a default feature, and the camera apps could parse codes without any third-party reader. tag still needed the microsoft tag app. that is a single extra step in a product category where every extra step kills adoption.

on august 19, 2013, microsoft announced a two-year sunset. the tag service would keep resolving urls until august 19, 2015, after which any printed tag would resolve to nothing. scanbuy, a barcode-services company, picked up the underlying tag infrastructure and migrated paying customers onto scanlife. the consumer side was dead either way.

~ the inheritance ~

tag is now a small footnote in the qr code's victory lap. the actual lesson is more interesting than the product, though: it is one of the cleaner examples of a closed format losing to an open one in real time, in a category where the closed format was actually a little better on paper. microsoft research had a slightly higher-density code with a slightly nicer aesthetic. it lost anyway because the entire stack required microsoft to keep paying for a resolver.

the qr codes you scan in 2026 on a restaurant menu are direct descendants of the format tag was supposed to displace. tag's colored triangles still exist on old print runs of magazines from 2010-2012, sitting next to coupons and phone numbers, pointing at servers that no longer answer.

@}-->--

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

« previous  ·  random  ·  next »