If Coway's new accounts jumped 26.7% while home revenue rose 2%, the POU water market is reading the wrong number
By Zenith Water Dispense Team ·
Coway's home revenue grew 2.0% while new rental accounts grew 26.7%. Zenith's models show why format mix tells you almost nothing about earnings quality.
I keep asking operators the same thing, and almost nobody has an answer ready. If your revenue grew 2% and your new accounts grew 26.7%, which figure describes your business? That pair of numbers came out of Seoul on 5 November 2026.
Coway is the market leader in South Korean water purifier rentals. It published its third quarter that day. Group revenue came to 1,426.1 billion won for the quarter, a rise of 13.7%. Operating profit was 260.8 billion won, up 7.3%. Home-market revenue rose 2.0%. Net new rental accounts in that same market rose 26.7%, to 214,000.
Why a rental book hides its own growth
A rental business reports its good news late. An account signed in September bills three months of rent this year. The revenue line is a record of decisions customers made two and three years ago. Coway's home revenue grew 2.0% because the accounts behind it were signed in 2023 and 2024. The 26.7% belongs to 2027 and 2028.
Overseas the same engine ran harder, with revenue there up 26.2%. Malaysia grew 21.5% and the United States grew 45.8%. Those markets are earlier in the same cycle, so more of their selling still shows up as growth.
Earnings quality does not follow the format mix
Look at a customer's invoice and the format names itself. On a bottled water dispense (BWD) bill, litres are most of what the customer pays for. Monthly rental dominates a point of use (POU) bill. An instant taps (ITS) bill puts its biggest number at the start, in the price of the unit.
Zenith's own models say something awkward about format mix. The mix tells you very little about the quality of a market's revenue. South Korea earns almost all of its dispense money from mains-fed machines. The UAE earns almost all of its dispense money from bottles. Both earn close to nine tenths of their dispense revenue on repeat. One is near that mark and the other is well past it.
Two markets built on opposite formats arrive at nearly the same answer. So the bottled-versus-mains-fed argument fills conferences for the wrong reason. Earnings quality is not what divides the two. Revenue per machine is, and that is a different problem.
Where capital and revenue move together
Across the models we have recomputed, bottled revenue never depends on a new installation. Every pound of it comes from water delivered to a machine someone installed years ago. Mains-fed revenue depends on an installation for a share in the teens or less.
The tap line is the single place where capital and revenue rise together. Roughly half to four fifths of tap revenue in a given year needs a machine fitted that year. An operator growing a tap book is buying its own growth each January.
Bottled gets paid in the year it sells
Bottled has one advantage this week's numbers make plain. Water goes out, an invoice follows, and both happen inside the same month. The effort and the revenue land in the same year. A rental book defers its own good news by design.
Some operators need this year's selling to show in this year's accounts. A bottled round serves them better. That is a real commercial argument, and the conversion case rarely answers it. Bottled is also the only format of the three that earns before it asks for capital.
What 2027 pricing will hang on
Operators who can show which part of the book renews will set their own price in 2027. The ones who cannot will be judged on a revenue line that is already two years old. Coway's quarter is a reminder. In a rental business the account book and the revenue line disagree. They always will, inside any one year.
🎯 Bring us the decision while it is still open
Everything above came off the same models our client work runs on. Say which market you are in and what you are deciding, and we will come back on how big the job is, where we would start and what it costs.
P.S. Two markets built on opposite machines can earn money the same way. The line a buyer actually prices is the revenue that renews without a fitter, and that split sits format by format inside the 2026 Zenith Water Dispense Market Reports. Europe's share of the series is thirty-one separate files. Should Mexico or Japan be the market you are pricing, that file is already written. Pick any of South Korea, Turkey and the UAE and you get the same build. Nothing from a transport hub is folded into the office total. The forecast runs to 2030 for units and money alike. The study and the workbook travel together unless you ask for one on its own. Built from direct interviews with local operators and data partnerships in every market we cover.