Make the institution ready for the promise it invites

In 2021 I approved the announcement of twelve branch openings. Premises, launch arrangements and dates were sufficiently advanced to create confidence, and I accepted that visible progress as an account of readiness. Seven locations had the necessary specialist-servicing assignments; five did not. The gap belonged in my decision before the announcement, not merely in an operating update afterwards. I had allowed the number of planned openings to become more persuasive than the organisation behind them.

The commercial preference was to preserve the public timetable. We instead revised it, opened the seven ready locations and deferred the other five. Standby expense of ₹2.1 crore remained in the account. We did not describe it as an avoided loss or assume that an eventual opening would recover it. The revised discussion required a named service assignment, its actual remit and workable capacity alongside existing duties. Adding a person's name to several location plans would not by itself resolve competing demands on that person.

I explained the correction to the board and retained ownership of the premature announcement. Branch, people, finance and technical leaders then worked through their respective responsibilities. The case does not invent a customer claim that was denied, a legal finding or a regulatory breach. It describes an enterprise commitment that had run ahead of readiness and the cost of correcting its timing. The later opening dates and commercial performance of the deferred locations are not supplied.

My judgement changed from asking whether a branch could open to asking what the institution was equipped to support once it did. That distinction is practical rather than rhetorical. It can change an assignment, a timetable or a decision to proceed. It also requires the chief executive to accept that a more accurate account may make a previously confident announcement look premature. Preserving that admission is more useful than writing a success story in which the original decision disappears.

In 2022 our industrial-park review examined thirty-seven insured sites. Twenty-four relied on the same two access roads and one power substation. Their addresses were distinct, but part of their operating exposure was connected. The combined sum insured across all thirty-seven sites was ₹486 crore. That amount was not a probable loss, a reinsurance recovery or a calculation of the capital required. The important finding was the dependency that a location-by-location account had made difficult to see.

I approved ₹86 lakh for mapping and engineering review. Risk head Anjali Qevzemi coordinated the enterprise question with the relevant technical owners. Underwriting, actuarial, engineering and reinsurance colleagues retained their assessments and decisions. The executive task was to ensure that the connection received attention and resources, not to derive terms from a map or declare every site equally vulnerable. The roads and substation could matter differently to individual operations, and that difference required work rather than an assumption.

There was a real tradeoff in adding another layer of investigation to business that appeared geographically dispersed in routine reporting. Colleagues needed time to reconcile site accounts and understand the shared infrastructure. We retained that cost and the limits of the available evidence. No actual catastrophe, claim total, premium saving or avoided financial loss is included in this fictional case. A fuller map is useful, but it does not by itself establish the frequency or severity of an interruption.

The result was an explicit concentration question in the enterprise account. Directors could ask how technical owners were examining common dependencies without substituting their own judgement for those disciplines. I learned to treat distinct locations as the beginning of a diversification discussion, not its conclusion. What connects sites may be as consequential as the count of sites, and the CEO must help the organisation see that connection before separate summaries make it disappear.

In a 2023 rehearsal of a ten-working-day CEO absence, we tracked forty-six routine decisions. Thirty-one stalled while colleagues waited for my personal sign-off, although formal authority sat elsewhere. This was not evidence that colleagues lacked ability or that every delayed decision should have been made faster. It showed a gap between the written allocation of responsibility and the practical permission people believed they needed. My own habits were part of that gap.

We funded an eight-week programme of delegation coaching and escalation practice at ₹64 lakh. The work examined the circumstances in which I had offered extra reassurance, revisited decisions or become the safest route through uncertainty. Reserved matters remained reserved, and underwriting, actuarial, claims, risk and other technical mandates were not reassigned through an executive workshop. Colleagues needed to understand both the authority they could use and the limits that still applied.

The tempting response would have been to issue a stronger instruction to act independently and then count decisions closed. I wanted the programme to examine why a person might reasonably expect a second approval despite the document. Senior behaviour had to support an authorised decision even when the CEO would have preferred another defensible option. Coaching could clarify that expectation, but it could not establish by itself that every future absence would be handled well or that a particular successor was ready.

The observed result remains thirty-one stalled decisions in the initial rehearsal and a funded change in practice. No later completion rate or operating saving is invented. The lesson for me was that delegation is partly an account of the chief executive's conduct. An institution cannot become less dependent on one person if that person continues to reward the habits that make personal permission feel necessary. The work therefore included changing how I participated, rather than asking everyone else to become more confident around an unchanged centre.

In 2024 we reviewed an outsourced after-hours service arrangement involving six supplier agents alongside our internal team. Sixty simulated notices were used to examine first-contact conversations. Eighteen responses wrongly promised coverage before an authorised assessment. The exercise was simulated; it does not establish eighteen affected customers, denied claims or actual settlement delays. It did show that the service conversation could create an assurance the person answering was not authorised to provide.

I sanctioned a ₹48 lakh programme covering the script, supervision and the route to our responsible internal colleagues. The supplier could acknowledge a notice, explain what would happen next and help the caller reach the proper process. It could not decide the contract's coverage by sounding reassuring. Claims and legal professionals retained their assessments, while operating leaders owned a workable handover. A caller should not have to infer those responsibilities from a vague disclaimer added after a confident promise.

The decision accepted more work in service design and supervision. We did not assume that removing one sentence would make every conversation appropriate, or that an outsourced colleague should provide no useful help. The revised brief required clarity about what could be said, what remained undecided and who would take responsibility for the next step. Its effectiveness would need further evidence. The funding was not presented as a measured reduction in complaints or a financial return.

For me, the case concerned the enterprise boundary of an outsourced service. A contract can describe that boundary without making it easy to respect during an anxious conversation. The CEO has to ensure that the operating arrangement, incentives and supervision support it. That is different from taking over a claims decision. I wanted an honest and useful first response whose reassurance came from a clear next step, rather than from a premature conclusion about an obligation that had not yet been assessed.

In 2025 a proposed cyber-insurance line carried a ₹120 crore first-year gross written premium ambition. Its service plan assumed one incident coordinator. We examined three concurrent fictional scenarios and identified four distinct response roles that could not credibly be treated as one person's interchangeable tasks. The proposal had described the commercial opportunity more fully than the organisation needed to respond when obligations arrived together.

I held the proposal and sanctioned ₹1.6 crore for readiness design. The work had to distinguish coordination, technical incident response, legal advice and customer communication, including the professional responsibility retained by authorised providers. One coordinator might connect those activities, but the title could not supply every competence or create simultaneous capacity. The exercise did not use an actual attack, expose customer records or claim that the organisation had already delivered the proposed response.

The tradeoff was to postpone a visible commercial opportunity while investing in the conditions on which it depended. The ₹120 crore figure remained an ambition, not booked premium, revenue or a forecast validated by the exercise. The ₹1.6 crore was a funded programme, not the full cost of an operating service or evidence that the gap had closed. Product, underwriting, risk, actuarial, legal and operating owners retained the further work necessary before any actual market decision.

I regard the outcome as a more demanding proposal rather than a successful launch. The board could now examine a service organisation alongside the opportunity, including what concurrent demand might require and which capabilities were still unproven. The case does not imply that a small provider is always unsuitable or that four roles necessarily mean four permanent employees. It asks whether the work and professional boundaries have been understood before a headcount or job title is used to make a complex promise look ready.