De'Longhi paid 662 million euros for a service annuity. Water dispense operators own one and rarely price it.

Most of the coverage filed this deal under coffee. I read it as a price tag on the service contract under the machine. De'Longhi paid 662.5 million euros last week for a bigger share of La Marzocco and Eversys. Its stake rose from 61.56% to 88.02%.

What the 662 million euros actually bought

The division sells professional coffee equipment. La Marzocco builds barista machines for cafes. Eversys builds automatic machines for offices, hotels and contract caterers. In the year to 30 June 2026 the division made 569 million euros of revenue, up 38%. A business built on placed equipment and service visits earned a 30.2% adjusted EBITDA margin.

The growth matters as much as the margin. Since the two brands were combined in late 2023, revenue has risen by more than half. Adjusted EBITDA has doubled over the same stretch. Equity value has climbed from 1.28 billion euros to about 2.5 billion. That is what a serviced equipment business looks like when it compounds. At that price the buyer was paying for a book of customers who keep coming back.

Why a 30% margin changes the POU water conversation

POU, or point of use water coolers plumbed into the mains, runs the same shape. The operator places a machine. Income then comes from filters, sanitisation and service visits. Many dispense operators still describe themselves by fleet size. De'Longhi's price puts the value in the service annuity ahead of the machine count.

The reason is simple enough. A machine sale happens once and carries a margin set by competition. A service contract renews, and the cost to switch sits with the customer. Filters, sanitisation and compliance records tie a customer to the operator who installed the unit. That stickiness is what earns a 30% margin. Dispense operators already own it, and many of them give it away on price.

The ownership structure carries its own signal. Minority holders had the right to ask for a stock market listing from March 2028. De'Longhi bought most of them out before that window opened. A buyer who pays early to avoid an IPO usually rates the asset above the market's price.

Where the bottled model keeps an edge

Bottled water dispense, shortened to BWD, keeps the supply in a container the operator owns. That container gets filled off site and swapped when it empties. The model holds an advantage that rarely gets argued. Every delivery leaves a record. The customer sees what arrived, when, and who signed for it. Proof of service is built into the bottled model and has to be engineered into the mains-fed one. A filter change is one line on an invoice. The customer has little way to check that it happened. Auditability matters more as buyers professionalise. A procurement team defending a hydration budget can put delivery notes on the table. That is one reason buyers keep bottled on the list long after they install their first tap.

What workplace hydration operators can take from a coffee deal

Zenith's pan-European model shows the three formats lean on employers to different degrees. Mains-fed is the most employer-dependent of the three. Its median workplace share sits well above bottled's and above taps'. That makes POU revenue the most sensitive to occupancy of any format. A service annuity is only as steady as the sites it is attached to.

There is a practical test for any operator reading this. Ask what a buyer would see if they opened your accounts tomorrow. If they find one revenue line and a machine count, you are selling a box business. If they find contracted service income with renewal history, you are selling an annuity. The same company can look like either, depending on how it reports.

Operators who want De'Longhi's multiple need to report like De'Longhi. That means recurring revenue on its own line, with contract length and renewal rates beside it. Buyers and lenders can price that. A fleet number tells them almost nothing. The coffee market has now shown what a serviced installed base is worth when it is reported clearly. Water dispense has the same model and far worse disclosure. The operators who fix the disclosure first will set the price for everyone else.

P.S. We break every market down by format, channel and operator. Unit bases, pricing and cancellation rates are there to model against. You will find the full set in the 2026 Zenith Water Dispense Market Reports. Built on the world's largest water dispense database. See the 2026 reports