Kandian

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

~ the obit ~

kandian never reached me. i was on the anglophone side of the internet in ankara through the whole 2019-2022 window, and chinese short-video apps live behind a language wall and an app-store wall that make them invisible unless you are looking. tencent launched kandian in 2019 as a defensive move. bytedance's douyin (the chinese domestic version of tiktok) had already become the app that owned people's attention, and every quarterly earnings call at tencent had to explain why qq and wechat were losing time-spent to it.

the pitch was social media plus news plus short video, with baked-in editing tools so people could produce as well as scroll. by march 2020 tencent said the kandian family of apps had reached 240 million daily active users, with about 58 minutes per user per day. those are enormous numbers by any western baseline. they were still not enough to stop douyin.

tencent tried to buy its way in with money. through 2020 it distributed roughly 3 billion yuan (about 443 million us dollars at the time) as creator subsidies to about one million kandian contributors. that is the sort of number you spend when the strategy is clear and the market is still moving. within two years the calculation flipped. on 15 august 2022 tencent posted the shutdown notice for the standalone app, with a data-backup window and a note that some kandian functionality would move into the mini world short-video feature inside qq.

~ the rap sheet ~

~ what tencent was actually defending ~

the thing to understand about chinese consumer tech in 2019-2022 is that time-spent-in-app was the metric everyone tracked. tencent had won the previous decade with qq and then wechat. bytedance quietly built a recommendation engine that made douyin the app you opened first when you sat down. every minute that migrated to douyin was a minute not in a tencent property, which meant an ad impression tencent could not sell.

kandian was one of a series of counters tencent tried. weishi, yoo, wesee, kandian, and mini world inside qq were all pointed at the same problem from different angles. none of them broke through. the shutdown of kandian in 2022 was part of a consolidation move that folded the surviving pieces into properties tencent already had large audiences in, which is the standard move when a rival brand has run out of ways to catch up.

~ how the shutdown actually read ~

the shutdown notice was blunt in the way chinese consumer-app announcements often are. adjustments in business development strategy. a specific date for data export. a pointer to qq mini world as the place remaining features would continue. no long farewell letter, no last-livestream event, just a quiet turning-off. the western equivalent would be a google graveyard sunset post with slightly less dressing.

the 240 million daily actives number that tencent had quoted in 2020 did not appear in the shutdown post. by 2022 the audience was smaller and less durable than the peak headline had suggested, though tencent never disclosed how much smaller. the kandian creators who had lived on subsidies drifted to douyin or kuaishou, which were paying less but reaching more.

~ what it says about that era ~

kandian is a good marker for a specific tencent problem in that period: the company kept trying to fight a recommendation-engine competitor with a portfolio of experimental apps and cash subsidies, and the recommendation-engine competitor kept winning. bytedance had built the underlying model well before douyin was branded, and every kandian-shaped counter had to compete against that infrastructure with a two-year handicap.

in 2026 the douyin/kuaishou duopoly is still the shape of chinese short video. tencent's contribution to that market lives inside qq as mini world and inside wechat as channels, both of which run on audiences tencent already had rather than on new brands tencent tried to build. kandian is the clearest small monument to the era when tencent thought a fourth or fifth brand launch might still change the answer.

@}-->--

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

« previous  ·  random  ·  next »