Although institutional building plans jumped 33.4%, the water cooler market still sells to the office
By Zenith Water Dispense Team ·
Commercial building plans fell 17.6% in October. Schools and hospitals rose 33.4%. Water dispense sells into the half that is shrinking.
The trade is reading the October building figures as a slowdown. I read them as a change of address. Commercial building plans in the United States fell 17.6% against the same month last year. Institutional plans rose 33.4%.
Dodge Construction Network published its Momentum Index for October on 6 November 2026. The index came in at 260.4, down 9.0% on September. Dodge counts nonresidential building projects going into planning. The firm says the index leads spending on those buildings by a year to 18 months. Commercial covers offices, warehouses and data centres. Institutional covers schools, hospitals, universities and public buildings.
The planning index leads the water order by about a year
A water order is one of the last things a new building places. Here is why. The machine arrives once the floor is finished and the counter is fitted. Somebody then has to hold a budget code for refreshments. A plan drawn this autumn becomes a dispense enquiry years later. Planning is the earliest honest signal a supplier gets. Everything downstream of it is already committed.
Dodge puts its own lead time at a year to 18 months ahead of construction spending. Add the build and the fit-out and the gap widens again. The planning numbers published last week describe an order book for 2028.
Commercial planning has been falling through this year. Institutional planning has not. An operator whose pipeline is mostly offices is reading a market that shrinks at the design stage.
Where each format stands decides who can buy it
The simplest way in is to ask where each one stands. On the floor beside a socket stands the bottled cooler running on 15 or 19 litre bottles (BWD). Against a wall with a pipe behind it stands the mains-fed cooler (POU). Cut into the counter sits the tap that pours boiling, chilled and sparkling water (ITS).
That matters because building types carry different fabric. A school corridor has sockets and no spare counter. A hospital ward has neither to give. The building decides the shortlist before the buyer has an opinion.
An office fit-out is the one place all three formats compete on level ground. A designer there can give any of them what it needs. That is also the half of the market whose plans are falling.
Why a school buys water on different terms from a floor of desks
Institutional buyers run frameworks. A school group or a hospital trust buys for many sites at once. The contract is measured in years. The money works differently here. A longer term lowers the quit rate. So each win is worth more over its life. The choice also moves back to the design stage, months before anyone takes a water call.
Occupancy is the second difference. An office with hybrid working is busy for part of the week. A hospital runs every hour of every year, so the same machine pours far more water. More throughput means more filter changes, more sanitisation visits and a bigger service line. The machine stops being a rental and starts being a route.
The institutional win is worth more and takes far longer to land. Most dispense sales teams are paid the other way round. Commission lands on the quick office close. So the pipeline fills with whatever is easiest to reach.
The French model already shows the office flattening out
Zenith's France model carries unit numbers for seven commercial building types, split by format. The 2025 reading is blunt. Mains-fed units in the French workplace barely grew at all last year. Healthcare, education and leisure each grew several times faster. Bottled units fell in every vertical, and the workplace was among the steepest. Leisure sites held up best. So the office is the one place where bottled is falling and mains-fed has stopped growing.
Follow the install. A workplace account is won from another operator, because almost every office already has something. A hospital adding a wing is a fresh placement. There is no incumbent to displace. A saturated vertical gives you a share fight. A building that did not exist last year gives you a placement.
Four French building types have no instant-tap base at all
Bottled takes criticism for its cost per litre and its carbon, and both complaints are fair. Set them against what the model shows about reach. Four building types in the French model carry no measurable instant-tap base. Education, retail, transport hubs and leisure are the four. Every one of them carries bottled.
A format that has never entered a building type cannot be placed there next month. Demand is now moving toward buildings the newest format has never served. Bottled is what an operator can put on the floor that quarter. The cost sits elsewhere. The route still has to be driven and the bottles still have to be carried. That is the real charge against bottled, and it is a heavy one. A finished office with a fitted kitchen will keep choosing the tap, and it should.
Where we think the office book goes from here
Here is our call. The commercial office account is a share fight through 2028. The growth placements sit in institutional buildings instead. Operators who build no institutional route to market will be defending price inside two years. That route is a different build. It needs bid writing, framework registration and a longer cash runway.
The Momentum Index publishes a commercial line every month. That line is where our call lives or dies. It fell 17.6% against last October. Two straight quarters of commercial planning growth would put the office back in the growth column. We would say so if it happened. Meanwhile the plans being drawn are for buildings that buy water differently. Those are not the buildings this trade knows best.
🌍 Read the market before you book the meeting
Every country we publish has its own page, and the daily analysis sits next to it. Most questions about a market are answered before anyone has to email us.
P.S. When demand moves to a building type you have never sold into, guesswork gets expensive. The vertical splits that tell you whether a market's schools and hospitals already buy, and in which format, are modelled market by market in the 2026 Zenith Water Dispense Market Reports. Thirty-one European markets have a file each, and the list stops there because the continent does. Where the market is Turkey, South Korea or the UAE, the same build is already on the shelf. Nothing is left to commission when the file you need is Mexico or Japan. Unit numbers and revenue run side by side to 2030. Homes are counted apart from workplaces in every market file. The commentary is sent with the model, or left out, whichever suits the job. We get these figures from the operators direct, in interviews we run every year. https://waterdispenseinsights.com/reports