Chief Product Officer India

The decisions behind a more understandable booking

Design and product work: 26 years; Career begins July 2000
Product colleagues: 46; Research and design included; four leads
Annual function expense: ₹12.6 crore; Engineering under separate leadership

Decisions & consequences

Giving eighteen changeover conflicts a visible cause

We prepared 240 bookings across twenty-six spaces with participating venue operators. Eighteen combinations revealed conflicts involving preparation or clearing time that the existing display did not represent clearly. The session times themselves appeared compatible, which made the omission easy to miss. I asked the team to treat the finding as a problem in the model of the booking, rather than propose a warning asking staff to remember extra time outside the calendar.

I sponsored ₹58 lakh of product and engineering work to represent setup and clearing windows separately from the session. Operators could explain the purpose of each window and see how it affected neighbouring bookings. We reviewed the behaviour with staff responsible for preparing spaces, not only the people who accepted enquiries. The work remained within a shared delivery plan agreed with our CTO; I did not present product sponsorship as sole authority over engineering resources.

Retesting addressed the eighteen known conflicts in the prepared set. We did not infer that every customer schedule was now free of conflicts or that the product had determined a safe interval for every type of event. Venue operators retained responsibility for the configuration and the physical work. We also kept the distinction between a system conflict and a staff decision to change an arrangement, so an override would not quietly become evidence that no preparation was needed.

The result was a clearer representation of the commitment surrounding a booking. I learned to ask whether a product's central object includes the work required to fulfil it. A session is only one part of using a venue. Making the other parts visible helped staff discuss a real constraint before accepting a neighbouring booking, while leaving them accountable for the operational judgement that the software could not supply.

Reporting the three keyboard users who still needed help

Thirty venue staff attempted representative booking tasks using only a keyboard. Nine completed them and twenty-one encountered barriers that prevented completion. The problems involved focus order, controls that were difficult to identify and steps whose next action was not apparent without a pointer. I asked researchers to preserve the task observations rather than reduce the exercise to a satisfaction score or a list of preferences about appearance.

We allocated ₹44 lakh to interaction changes and agreed the implementation sequence with engineering. Designers and researchers checked the proposed behaviour against the tasks that had exposed the problems. On repeat testing with the same thirty participants, twenty-seven completed and three still required assistance. I reported both numbers. The improvement was substantial within that exercise, but the remaining difficulties were not made irrelevant by the success of the other participants.

We did not describe the result as a formal accessibility certification or as evidence about every user, assistive technology and device. The same-group repeat also meant participants had prior familiarity with the task. That was a useful condition for inspecting the changes, but it limited what we could infer about first-time use. The three unresolved cases remained in the next research and delivery plan rather than disappear into a rounded completion percentage.

This work changed how I discuss inclusive product development with colleagues. A claim about access should identify the activity and the people whose experience supports it. The practical outcome was a more usable set of booking interactions for the tested group, with documented work still to do. My responsibility was to secure attention and resources for that work and to prevent an encouraging result from becoming a broader promise than the evidence justified.

Showing twenty-three holiday exceptions before a series changed

Operators reviewed 140 recurring booking series against the dates they intended to reserve. Twenty-three contained holiday exceptions that did not match that intention. Some staff expected a change to affect one occurrence, while others believed they were changing the remaining series. The existing confirmation did not show the resulting dates clearly enough to resolve those different interpretations before the edit took effect.

I approved ₹36 lakh of product work for a preview-and-confirm flow, with engineering capacity agreed through our normal planning process. The preview identified the occurrences affected by the selected action and made the scope of the change explicit. Researchers tested the language with operators who handled recurring classes and rehearsals. We avoided assuming that a public holiday always meant closure, because each venue and customer agreement could require a different arrangement.

The twenty-three reviewed mismatches were corrected with the operators. Existing customer commitments were not rewritten automatically merely because we had changed the interface. Staff remained responsible for confirming intended dates and communicating an alteration where necessary. The product retained the distinction between a single occurrence and a wider series action so the record could explain what the operator had approved.

The outcome was an inspectable change before application, rather than a promise that recurring bookings could never be misunderstood. I learned that confirmation is meaningful only when it exposes the consequence a person is being asked to accept. A short question with an affirmative button can feel efficient while leaving the central ambiguity untouched. Showing the actual dates gave staff a better opportunity to catch an unintended change and to account for the commitment they were making.

Making seventeen expired holds legible to the next colleague

A review with venue operators examined sixty-four tentative reservations. Seventeen remained visible after their stated expiry without an unambiguous expired status. Colleagues reading the calendar could not tell whether a conversation was still active, whether a customer had confirmed elsewhere or whether a new enquiry could be considered. I treated the issue as a failure to represent the status of a commercial discussion, not simply a housekeeping task to remove old entries.

We funded ₹29 lakh for explicit expiry labels, a named staff owner and a review action that distinguished an expired hold from a confirmed booking. Converting a hold required an operator decision. We did not silently turn every hold into a reservation or automatically release a space on behalf of a venue whose customer agreement might require contact first. Product researchers worked through the different local practices before we settled the shared behaviour.

The seventeen examples were reviewed and their status clarified by the responsible operators. The record of the original hold remained available. We made no claim that the product had recovered lost revenue or improved utilisation by a measured percentage, because this exercise did not establish those outcomes. Its evidence concerned the clarity of status and the ability of another colleague to identify who should resolve the outstanding question.

The lesson was that a calendar can display a time accurately while misrepresenting the strength of the commitment attached to it. A tentative conversation, an expired request and a confirmed booking should not look interchangeable. I wanted the product to help the next staff member ask the right question without assuming that software had authority to settle a customer relationship. The change supported a human decision at the point where the old display had obscured the need for one.

Discuss a leadership challenge