Koo

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

~ the last words ~

koo never reached turkey at meaningful scale. either way, it was an india-coded product from the start, a microblogging app in hindi and twelve other indian languages aimed at users twitter had never really served. the founders aprameya radhakrishna and mayank bidawatka had already built taxiforsure before this, and koo was the second-act bet. it launched in march 2020 from bengaluru with a small yellow bird logo and a writing surface that looked almost exactly like twitter (close enough).

i learned about koo during the 2021 india-vs-twitter standoff, the same way most non-indian users did. anyway, the modi government had asked twitter to take down accounts linked to the farmers' protests. twitter complied partially, then pushed back. in the middle of that standoff, several union ministers and the bjp's official handles announced they were joining koo, and the app spiked into the top of the indian app store charts. the registered base went from a few hundred thousand to several million across that one month.

this part is mostly told from the english-language press at the time. koo raised about sixty million dollars across two rounds in 2021 and 2022, reaching a roughly three hundred million dollar valuation. the registered base climbed to about sixty million by november 2022. the daily active user count peaked closer to two million. by mid '23 the active base was sliding fast, from 7.2 million monthly actives in june '23 to 2.7 million by february '24.

the founders had spent most of '23 looking for a buyer. talks with dailyhunt went the furthest, then collapsed in early '24 over what radhakrishna later called "the high cost of technology services" and a "predator-not-prey" risk profile for the buyer. koo cut staff in waves through the spring. on july 3 2024 radhakrishna and bidawatka posted a goodbye on linkedin and the app was shut down. the servers went off shortly after.

~ the record ~

~ how it landed ~

the initial koo pitch was a regional-language layer on top of a twitter-shaped product. either way, hindi was the headline language, with bengali, tamil, telugu, kannada, marathi and gujarati added across the first year. the bet was that twitter's india numbers undercounted the actual hindi-speaking population by a wide margin, and that an app built around hindi-first composition would pick up users twitter could not. the early product launched with prompts in the local language and trending topics scoped per state.

the 2021 modi-vs-twitter moment gave the app a distribution windfall it would not otherwise have had. anyway, the question of whether that windfall was a feature or a curse was already a topic in indian press by the end of '21. the loudest accounts on the platform skewed bjp-aligned, which made koo read as a partisan product to non-bjp users and damaged the cross-political pitch the founders had wanted. by '22 the founders were trying to rebrand the platform's identity, with limited success.

~ the early days ~

the high point was probably late 2022. for some reason, koo was operating in ten languages by then. it had launched in brazil in november '22, which was the same month the elon musk twitter acquisition closed and a small wave of brazilian users left twitter looking for alternatives. for a few weeks the app was downloaded faster than twitter in brazil. the wave faded almost immediately as brazilian users settled back into twitter or moved on to mastodon and bluesky instead.

the brazil pilot was wound down within a year. the indian growth flattened in '23 as the post-2021 political wave faded and twitter's india numbers stopped declining (close enough). koo never built the second-order network effects, friend graphs and creator hooks, that twitter had spent fifteen years accumulating. a regional-language layer was a defensible niche for early users and not a moat against twitter once english-speaking indian users stayed put.

~ the postmortem ~

the dailyhunt acquisition talks were the last real exit path. dailyhunt is a verse innovation-backed news aggregator with overlapping audiences and a healthier balance sheet, the structural fit on paper was clean. the deal fell apart in spring '24. the official line from both sides was a mismatch on terms and on culture. radhakrishna's linkedin post on july 3 cited unpredictable funding markets and the cost of cloud bills as the proximate cause.

koo's bookend in 2026 looks similar to app.net and ello in the western world, a regional twitter alternative that briefly mattered, peaked on a specific political moment, then ran out of road. by then, the twitter-alternatives canon now sits across bluesky, threads, and mastodon for the english-speaking world, with no obvious successor in the hindi-language layer. the niche koo was built around did not disappear, the company built to occupy it just could not keep paying the bills.

@}-->--

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

« previous  ·  random  ·  next »