The archive · Commerce & Marketplaces · Strategic decision · 2015–2026
Karrot bets hyperlocal trust beats marketplaces; ₩270B revenue, global bet bleeds
Karrot bet proximity builds trust: free hyperlocal listings paid by local ads; Korea made ₩270.7B revenue and two years of profit, overseas lost ₩52.5B.
Karrot (Danggeun Market / 당근마켓)
What the business is
Karrot (Danggeun Market) is South Korea's dominant hyperlocal community marketplace: a free secondhand-trading app that only shows listings within a ~6km radius, with neighborhood groups, job boards (Danggeun Alba), business profiles and payments layered on, monetized mainly through location-targeted advertising.
Starting capital：$205M raised through Series D ($162M round at $2.7B valuation, Aug 2021, led by DST Global); cumulative ₩92.1B invested in Karrot Canada from 2021 to Q1 2026.
How it started
Karrot was founded in June 2015 by ex-KakaoTalk product people Gary Kim and Kim Yong-hyun, who watched South Korea's dominant secondhand site Joonggonara stay web-only and anonymous, with scams and no mobile app. Karrot restricted listings to a ~6km radius, verified users by phone and location, and pushed in-person trades; by April 2020 it was Korea's #2 shopping app with 7M MAU, and it grew roughly 3x per year through 2021.
What happened
Series C of $33M (June 2020, Goodwater/Altos) and Series D of $162M at a $2.7B valuation (Aug 2021, DST Global) funded Karrot Pay, O2O tools for about 300,000 local SMBs, and the overseas push: UK (Nov 2019), Canada (Sep 2020), US (Oct 2020), Japan (Feb 2021). At home, hyperlocal ads produced 2024's first meaningful profit, then 2025 standalone revenue of ₩269B with operating profit ₩67.1B (+78%), 190M transactions and 21M+ MAU — two consecutive profitable years.
How it ended up
2025 consolidated results: revenue ₩270.7B (+43%), operating profit ₩14.6B (+481%) and net profit ₩23B, but the global bet is still in the red — overseas subsidiaries lost ₩52.5B in 2025 on just ₩1.6B revenue, Karrot Canada alone has absorbed ₩92.1B of cumulative investment, and Karrot exited the UK, its first overseas market, on 2026-04-30 to concentrate on North America. Investor estimates put the company's valuation at about ₩2.5T, below the ₩3T of the 2021 round.
Background
Karrot (Danggeun Market) was founded in June 2015 by ex-KakaoTalk product people Gary Kim and Kim Yong-hyun, who saw South Korea's dominant secondhand website Joonggonara stay anonymous, scam-prone and web-only. Their counter-bet was that proximity creates trust: the app showed listings only within a ~6km radius, verified every user by phone and location, and encouraged face-to-face trades. Instead of charging fees, Karrot kept trading free and sold hyperlocal advertising to nearby businesses — landlords, tutors, car dealers and shopkeepers paying to reach users a few streets away.
The model compounded on Korea's density: 7M MAU by April 2020 (#2 shopping app after Coupang), then roughly 3x MAU growth per year through 2021, when TechCrunch reported 21M+ registered users and 14.2M MAU. A $33M Series C (2020) and a $162M Series D at a $2.7B valuation (2021, DST Global) funded Karrot Pay, O2O tools for ~300,000 local SMBs, and launches in the UK (2019), Canada (2020), US (2020) and Japan (2021). At home the ad machine produced 2024's first meaningful profit, then 2025 standalone revenue of ₩269B and operating profit ₩67.1B, with 190M transactions and 21M+ MAU.
The overseas half of the bet is still unproven. In 2025, overseas subsidiaries earned only ₩1.6B and lost ₩52.5B, Karrot Canada absorbed the bulk of ₩92.1B in cumulative investment, and on 2026-04-30 Karrot shut its UK service — the symbolic first international market — after seven years, citing a strategic focus on North America. Consolidated 2025 results were still positive (revenue ₩270.7B, operating profit ₩14.6B, net profit ₩23B), but by August 2026 investor estimates put Karrot's valuation around ₩2.5T, below the ₩3T of its 2021 round, with MAU slipping from a 21.6M December 2025 peak.
What has to be true
- A 6km listing radius plus verified identities and in-person trades turned proximity into a trust layer that anonymous national marketplaces could not match.
- Free listings plus hyperlocal ads aligned incentives: about 300,000 neighborhood SMBs paid to reach nearby buyers, and advertiser count grew 37% in 2025.
- Density was the core assumption, and Korea's urban structure supplied it — the model compounded, roughly tripling MAU every year through 2021.
- The same assumption failed to travel: low-density, franchise-heavy North America left little local-ad inventory, so overseas units lost ₩52.5B on ₩1.6B revenue in 2025.
- The UK exit shows disciplined selection: after seven unprofitable years, Karrot cut its first market and doubled down on North America rather than spreading thin.
What can be applied
Hyperlocal won where density held — Korea's neighborhood structure fueled a ₩270B ad business — but the UK and Canada had no equivalent local-ad base; home dominance did not make the model exportable.
Aftermath
As of the latest verified reports (2026-08-22), Karrot operates in South Korea, Canada, Japan and the US after ending UK service on 2026-04-30; co-founder Kim Yong-hyun is based in Canada leading North America. MAU slipped from a December 2025 peak of 21.6M to 20.6M in February 2026, and investor estimates put the valuation near ₩2.5T — below the ₩3T of the 2021 Series D — while the company says it will prioritize North American growth and 'financial growth' over an IPO. Korea's standalone business remains profitable, with 2025 revenue of ₩269B and operating profit of ₩67.1B.
Sources
- Danggeun Market, the South Korean secondhand marketplace app, raises $33 million Series C
- South Korean online secondhand marketplace Danggeun Market raises $162M at a $2.7B valuation
- Secondhand market app Karrot hits 2 million users in Canada
- 당근, 2025년 연간 실적 발표
- 당근서 작년 중고거래 연결 1억9천만건…2년 연속 흑자 확대
- 당근, 7년 만에 英서 철수…'밑빠진 독' 해외 투자 재편 시급
spotted an error? The archive wants to know.
Your turn
You just read one. Describe what you are building, and see who is betting on the same thing.
Free account · 3 free questions · no card