Table of Contents

Notes from Coworking Europe 2026: what the stage said vs what members actually book.

Everyone agreed on phone booths

Ask a room of coworking operators what AI will do to their floor plans and one answer comes before any other: more phone booths. That was the consensus on the AI panel at Coworking Europe in Paris last week. The logic is easy to follow. More calls, more video, more voice work with AI tools, and more need for somewhere quiet to do it.

It is a reasonable prediction. Our booking data does not show it. In the Flex Space Observatory, booths remain a small share of bookings, around 4.17% of the total resources bookings, and there has been no sharp rise.

I was only in Paris for a day this year. More than 500 people, a lot of content, and a lot to catch up on afterwards. I spent most of it in conversations rather than sessions, but made it to the two panels I wanted to see: ancillary revenue and AI in coworking. Both left me with the same thought. What operators say on stage and what members actually book are not always the same thing, and the distance between the two is worth paying attention to.

Ancillary revenue: who is actually booking

Ancillary revenue is not a new topic, but every year brings new ways to monetise space. The usual suspects came up in Paris, virtual offices and parking among them. So did meeting rooms, event spaces and the growth of hot desks.

We went into the Nexudus data to look at the question the panel kept circling: who books these resources. Across meeting rooms, event spaces and other bookable resources, 27% of bookings come from external contacts and 73% from members over the last 12 months. Contact bookings per location have doubled since 2016-2019 (+102%), while member bookings grew 56%.

External bookers also stay longer. In 2026, non-member bookings average around 232 minutes, against roughly 168 for members. That is a different customer, usually booking for a specific purpose, and one that most spaces still treat as an afterthought.

Hot desks are the other trend we have tracked in the Observatory for a while. In 2016 they were around 1% of bookings. By 2022 they peaked at about 17%, while meeting rooms fell from around 57% of bookings to the high 30s.

Part of this comes from larger companies and teams. They want the option to take open-area desks when headcount peaks, without committing to more private office space.

What works also depends on the space. F&B matters far more in large spaces than small ones, where footfall cannot carry it. Parking sells better in central locations where parking is hard. Copying another operator's ancillary mix without their size and location is not a guarantee to success.

Five ways to add revenue without adding overhead

Most spaces can offer a set of services with little implementation cost. Five worth looking at:

  1. Mail handling tiers. Receiving post is a commodity. Scanning, forwarding and parcel notifications are services members and virtual office clients will pay for, and the process already exists.

  2. Lockers and storage. Hot desk members have nowhere to leave things. A bank of lockers or a storage room turns unused corners into recurring monthly revenue.

  3. Opening resources to external contacts. If outside bookers are more than a quarter of bookings and stay longer, make it easy for them: a public booking page, card payment at checkout, no membership required.

  4. Out-of-hours event hire. Weekends account for about 4% of bookings in the Observatory data. Evenings and weekends are capacity you already pay for, and workshops, meetups and community events can use it.

  5. Paid add-ons at booking. Catering, AV equipment, extended hours or printing, offered at the point of booking rather than on a price list nobody reads.

None of these need new space or new hires. They need a product set up in the system and someone to decide the price.

Dynamic pricing is not a hotel problem

Dynamic pricing was the other debate in the ancillary session. Together with TWIC we dedicated a full edition of Undercurrents to it this summer, so I have a view.

Comparing coworking directly to hotels or airlines is a mistake. The inventory is perishable in the same way: an empty meeting room hour cannot be sold tomorrow. The customer is not. A hotel guest is a one-off transaction. A coworking member comes back every week, talks to other members, and notices when the room they booked last Tuesday costs more this Tuesday. Pricing that wins a single booking at the expense of that relationship could be a bad trade.

What operators do value is granularity: different rates by time, resource and customer type, and the ability to be transparent with their community about why prices differ. Many already do this [manually] without calling it dynamic pricing. In the Undercurrents research, one operator found only around 6% of its bookings happened at the base rate.

Formal adoption is still low. Around 9.5% of operators use a dynamic pricing engine, according to our data. For a topic that fills panels every year, that is a long way from mainstream.

Booths, micro offices and the deskless office

Back to the AI panel. If booths are in such demand, why does the data not show it? I see two explanations.

The first is naming. Operators call them booths, pods, call rooms or quiet rooms, and some of that demand may be landing in the "other" category, which has held at around 37% of bookings. The second, and more likely, is that most booths are not set up as bookable at all. They run first come, first served, and demand that never touches the booking system never shows up in booking data. That does not mean it cannot be measured. Occupancy sensors show how often a booth is used and for how long, without anyone having to book it. The panel may be right about usage. Most operators just are not measuring it yet.

The second shift I have now heard about at different conferences concerns large offices. Demand that used to be for a whole floor is being broken into micro offices for teams of two or three. Smaller teams are doing more with AI, need less space, and still want a door that closes.

That leads to a bigger question. Offices have always been designed around the desk and the computer. If AI lets us work in a more mobile way, through voice, on a phone, away from a fixed screen, do we end up with a deskless office? I wrote about this a while ago, and in issue #024 we looked at how AI is redrawing floor plans around focus, alignment and recovery.

Where the opportunity is

Operators on stage speak from what they see on their own floors. Booking data shows what happens across many spaces at once. Each sees something the other misses, and the gap between the two is where the opportunity is.

Booths are the clearest example. If the demand is real but unbookable, the operator has no data on it and no revenue from it. The same goes for external bookers nobody is courting, hot desk overflow nobody is pricing, and rates that already vary without anyone measuring the effect. Making a resource bookable, or at least measuring it, is the first step to knowing whether you built the right room. And bookable no longer has to mean friction. The usual objection is that nobody wants to open an app to grab a booth for a quick call. With AI agents, a member can ask in a chat or by voice and have it booked in seconds.

If you want to test your own assumptions against the numbers, the full resource demand analysis is on the Flex Space Observatory.

Reply

Avatar

or to participate