Vending machines and other overlooked data sources in hotels

Hospitality has a tech literacy problem hiding in plain sight. Service robots and generative AI were among the most discussed themes based on a bibliometric review of tourism and hospitality journals (To & Yu 2025). Meanwhile, there’s a much older form of automation that generates purchase data in hotel lobbies, the vending machine.

In Haaga-Helia’s HosByte project, we explore smart omnichannel hospitality sales opportunities. The vending machine has recently come to our attention as an untapped sales channel, while revealing a recurring problem in hospitality operations. The industry generates large amounts of operational data, but the difficulty sits after collection. Lamest and Brady (2019) studied five hotels chosen for being high-performing users of data and dashboards, and still found managers short of the time and the interpretive skills to turn those dashboards into decisions.

Hotel CityBox Hakaniemi operates without a restaurant which is a rare scene in the Finnish hospitality industry. Instead, guests rely on vending machines and a self-catering kitchen. Operationally, that changes more than food service. It turns a small revenue stream into a forecasting tool.

Amelia Lintu, who manages the property, described it during a HosByte interview:

“Vending machines, they offer great data. It’s quite interesting to see different purchase patterns with guests in the types of products that they buy, weekends versus weekdays or time of the day. Now that we’re going into summer, we can already predict that beverage sales will go up considerably and we’re able to adapt the selection of the beverages.”

The machine is a source of data. The data shows what sells, when demand changes, and how guest behavior shifts between weekdays and weekends. CityBox is an unusual case. Without a restaurant, the vending machine serves as a key purchase point. That’s why its data offers a clearer picture than it would in a hotel where most food and drink purchases happen over the counter.

Decisions that data can support exist in every hotel with a vending machine. Someone chooses what is being sold in the machine, the right amount of the product, and when to change the selection. In a full-service hotel that choice is a smaller line item, and smaller line items are the ones that are often made by instinct and are rarely revisited.

The data hotels already pay for

The standard hotel approach to vending is passive. The supplier stocks the machine, the supplier collects the revenue, and the hotel takes a percentage. At the same time, the machine generates information on purchase timing, seasonal demand patterns, price sensitivity, and product substitution. Guests generate that data inside the hotel’s own building, and by default it stays with the supplier who owns the machine and the terminal in it. Whether the hotel ever sees the data comes down to one question: has anyone asked to get it?

Research on hospitality automation has concentrated on the guest-facing end. Anwar, Deliana and Suyamto (2024) studied automated check-in, check-out and chatbots, and found that moving routine transactions out of staff hands freed time for guest interaction. The vending machine sits at the other end of the operation, where nobody is serving the guest in the first place. The savings come from reducing the guesswork behind stocking decisions and seasonal selection changes this far been made by relying on memory.

Other underused data sources

Across HosByte project partners, the same pattern repeats when we talk about using data for smart omnichannel sales. Vending machines are only one example of underused operational data sources. Laundry services, coffee machines, mini-bars, parking systems, and paid Wi-Fi tiers all generate behavioral data that hotels could actively use.

Lintu noted that CityBox already receives operational reporting from its laundry partner. Many hotels have similar supplier relationships, and it’s worth checking whether those include a data feed at the next contract renewal.

Several practical steps follow from this. Ask vending suppliers what operational data they collect and could share. Audit the smaller technology systems inside the property, the coffee machines, the laundry, parking and the Wi-Fi tiers, and for each one ask the supplier what data it collects, in what format and how often they can send it. Discuss with PMS providers how point-of-sale integrations could support non-room revenue tracking.

Generative AI and service robots take up most of the research attention in hospitality and tourism sector (To & Yu 2025). AI or service robots won’t help operators who still treat their vending machines and pre-existing data sources like laundry services and parking systems as black boxes. Even though that is often where the more useful operational data already exists.

Image: Shutterstock