The calendar ends the session before the room is ready

A calendar can be completely accurate about the start and end of a session while still misleading the people who run the room. The first group finishes at four and the next begins at four fifteen. On screen, the arrangement looks orderly. In the building, somebody must move equipment, check what has been left behind and prepare the space for a different activity. If the product records only attendance time, that work survives as a memory, a side note or a conversation that the next colleague may never hear.

I encountered this problem while reviewing a prepared set of 240 bookings across twenty-six spaces with venue operators. Eighteen combinations exposed conflicts involving setup or clearing time. We had not selected the bookings to calculate an error rate across every customer. We wanted to examine conditions that the product needed to represent. The finding was useful because it identified a missing part of the commitment, not because it produced a percentage that could be placed on a performance chart.

## Start with the people between sessions

The person accepting an enquiry may be several steps away from the person preparing the room. One is discussing a customer's preferred time; another is deciding how to arrange furniture or return equipment. Both can act reasonably while relying on different meanings of a booking. Product research needs to include that difference. Asking only whether the receptionist can enter a reservation quickly would miss the work that makes the reservation possible.

I ask researchers to follow a booking through the people who encounter it. What does each person think has been agreed? Which information comes from the system, and which comes from knowing a colleague or remembering a previous event? A useful observation is often an ordinary sentence: someone says that everybody knows the room needs longer after this kind of session. The next question is how a new employee learns it, and whether the product gives them any reason to ask before confirming the next booking.

The aim is not to convert every local practice into a universal rule. Two venues can use similar rooms differently. A small rehearsal may require little rearrangement, while a workshop in the same space may need a different preparation. The product should provide a way to represent those arrangements without declaring that one interval is correct everywhere. That distinction protects both usability and the operator's responsibility. We can expose the assumption; we cannot establish the physical facts from a calendar alone.

In the reviewed examples, we separated the session from setup and clearing windows. This made the additional time visible and gave staff a reason for an apparent conflict. It also changed the conversation with customers. A period could be unavailable even though nobody was attending a session at that moment. The explanation was no longer hidden in somebody's memory. It could be discussed as part of the service the venue was agreeing to provide.

## Describe the constraint before designing the warning

Warnings are tempting because they appear cheaper than changing the model. We could have added a reminder asking staff to allow enough preparation time. But a reminder would still leave the calendar showing a free interval that was not actually free for the intended use. It would ask the operator to correct the product's meaning every time they read it. I prefer to establish what the reservation represents before deciding how the interface should alert someone to a conflict.

We sponsored ₹58 lakh of work to make the windows explicit. Product, research and engineering agreed the behaviour together. A conflict needed to show which part of the arrangement overlapped and give the operator a route to examine it. Merely painting both bookings red would identify a problem without helping staff understand whether they needed to change a session, revise preparation or discuss another space. The explanation should support a decision, rather than simply demand that somebody make the warning disappear.

An override raises a further question. Operators sometimes have a legitimate reason to change an arrangement. If the system lets them do so, the record should not imply that the original constraint never existed. It should make the revised commitment understandable to the next colleague. Otherwise a one-time decision can look like a permanent rule, and the work that somebody agreed to absorb becomes invisible again. Flexibility is useful only if people can understand what was flexed.

Testing the change required returning to the eighteen known conflicts. We could inspect whether the new representation handled those examples and whether staff understood the cause. That gave us a defined result. It did not establish that all future schedules would be conflict-free, and it did not prove that the chosen preparation interval was sufficient for every event. I want those limits stated when the work is reported, because the language of completion can easily become broader than the examination that supports it.

## Keep the operator's decision visible

A venue's responsibility extends beyond time. Staffing, access, equipment and the permitted use of a space can all affect whether an event should proceed. A software supplier must be careful when describing a booking as valid or available. Those words may sound like a wider approval than the product can provide. Our role was to make the scheduling commitment clearer. We did not certify event safety or decide that a particular physical arrangement was suitable merely because the calendar accepted it.

This boundary also matters inside the product company. As CPO I sponsor priorities and explain the evidence behind them. Engineering has its own leadership, and venue operations belong to the customer. Treating a product decision as if one person controlled all three areas would make the account simpler but less useful. The work succeeds through an agreement about responsibilities, including who decides configuration, who implements behaviour and who checks what happens in the building.

The commercial benefit should be described with the same care. We had evidence of a clearer model and known conflicts addressed in a prepared set. We did not measure recovered revenue or declare that utilisation had improved by a certain amount. Those outcomes would require their own definitions and observations. A credible account can explain why a change matters without borrowing a financial result that the work has not established.

I now look for the equivalent of changeover time in other products. It is the necessary activity that falls outside the system's central object and therefore becomes somebody else's unrecorded responsibility. Finding it usually requires listening to the person who prepares, hands over or clears up. A good product does not make that work disappear from view. It gives people a more faithful description of the promise they are making, and a chance to resolve the practical consequences before the next person arrives.