Europe's lowest water cooler churn sits in its fastest-shrinking bottled market

One figure changed how I read this market. Western Europe’s bottled cooler base grew last year. Almost every operator I speak to says their own bottled book shrank.

Zenith’s published 2026 European data puts bottled dispense up 3.4% in Western Europe in 2025. Europe’s whole installed base passed 6.6 million units, up about 2% on the year. Both the market number and the operator complaint are correct at the same time. The gap between them is a number most boards never see.

Three machine types run this market. BWD (bottled water dispense) means coolers fed by 19-litre bottles. POU (point of use) means coolers plumbed into the mains. ITS (instant taps) means counter units that pour boiling, chilled and sparkling water.

What a net number leaves out

Every fleet figure you read is a net figure. It is gross installs in, minus cancellations out. Two separate engines produce one headline. A book with a 12% quit rate has to install 12% of its base a year. That only holds it level.

Operators track the quit rate closely. It comes out of the service system and it lands on every board pack. The install rate rarely gets the same treatment. Sales owns it and reports it as a raw count. Almost nobody divides it by the losses it has to cover.

The best retention sits in the fastest-shrinking book

Germany has the lowest bottled cancellation rate in Western Europe. It also lost close to a quarter of its bottled base in five years. Finland has the highest cancellation rate in the region. Its bottled base is close to flat across the same period. Retention rank and growth rank barely line up.

Those two readings are directional cuts from our own database rather than published figures. The pattern behind them is plain enough. Germany’s bottled decline came from the sales side. Customers stayed. New ones stopped arriving.

Why the install rate goes unmanaged

Retention spending is easy to approve. It is defensive, it has a clear owner, and the saving shows up inside a quarter. Install capacity costs money now and pays back across a whole contract. So it loses most budget arguments.

There is a measurement problem too. Cancellations are counted by a system that has to log them. Gross installs are counted by people who are also counting upgrades and replacements. Mixing those together produces a figure nobody trusts, so it gets left off the pack.

What the growing market actually did

Iberia carried Western Europe’s bottled growth last year, and it did it in homes. Spain’s residential subscription model has no unusual loyalty advantage. Its quit rate sits in the middle of the Western European range. It grows because it sells new households faster than it loses old ones, week after week.

That is worth saying plainly, because bottled coolers get written off too easily. In the right channel, with a dense enough route, bottled is the growth line. Spain has now grown for thirteen straight years.

The ratio to put in front of a board

One number fixes most of this. Divide gross installs by cancellations over the same twelve months. Above 1.0 the book grows. Below 1.0 it shrinks, whatever the quit rate says. A business with excellent retention and a replacement ratio of 0.8 is shrinking on autopilot.

Buyers should ask for it too. A target with best-in-class churn and a falling install rate has an ageing customer base. The price should reflect that. Low churn is evidence about your customers’ age as often as it is evidence about your service.

We can see this because operators submit their own unit counts to us. We also work with local data partners in each market. The markets that grew last year sold harder. Holding on tighter came second.

For 2027 planning, set the install target first and let the retention target follow it. A fleet plan that starts with churn is a plan to shrink slowly. The question to put in the room is simple. How many machines do we have to place next year before we have grown at all?