Chief Information Officer / India
The default carried more authority than it had received
- Technology organisation
- 26 colleagues
- Three heads within the CIO remit
- Annual technology expense
- ₹5.1 crore
- FY2026–27; ₹2.4 crore change programme separately approved
- Information-systems career
- 34 years
- Application support through campaign-services CIO responsibility
Decisions & consequences
01Correcting the calendar rule I had supported before it authorised too much work
In 2018 I supported a resource calendar that multiplied forty colleagues by thirty scheduled hours and displayed 1,200 hours as available for campaign work. The common view looked more orderly than the planning sheets it replaced. A proposed allocation totalled 1,120 hours, apparently within that capacity. Daniel asked where approved leave and training had been included in the calculation.
The review identified 240 such hours within the gross total. Subtracting them left 960 potential hours before assignments and skill fit. The proposed allocation exceeded that account by 160 hours. Some planning colleagues wanted to retain the simple display and leave individual managers to make adjustments outside the system. I rejected that as a sufficient answer because the shared calendar was already inviting people to rely on the broader figure.
We held the unsupported 160 planned hours and committed ₹32 lakh over nine weeks to correcting the account and the review process. The case does not assert that the hours had already been worked, that clients suffered cancellations or that the correction created a realised productivity gain. Nor did 960 become a universal allocation limit independent of skills and existing commitments. People and delivery owners retained those decisions; technology had to preserve their conditions rather than imply that an empty space established permission.
The trade-off was losing a convenient capacity headline and delaying part of the proposed plan while managers rebuilt the relevant account. I owned the rule I had supported. The result was a more defensible planning view with the remaining decisions still visible. The lesson was that a system can be numerically tidy and operationally misleading at the same time. Its simplicity should help colleagues understand the commitment they are making, not conceal the subtraction or qualification that would change whether they are entitled to make it.
Returning inherited workspace permissions to the owner’s intended boundary
In 2023 a review examined forty client-workspace folders. Twenty-eight matched the accountable owner’s intended permission boundary. Twelve inherited access more broadly than intended from the surrounding workspace structure. The inherited settings were technically consistent with the rule, but the business relationship required a more specific decision than the parent setting could supply.
The systems sponsor argued that a common inheritance model was easier to maintain and that exceptions could multiply administrative work. Daniel wanted the client-work boundary made explicit before the system invited further reliance. I supported holding the twelve affected folders for review with the authorised owners. The question was not whether common settings were inherently unsuitable, but whether this particular setting accurately expressed the permission the business had approved.
We used ₹47 lakh over eleven weeks to correct ten folders and leave two held while their owners clarified the intended arrangement. Qualified security, contractual and client colleagues retained their respective responsibilities. The record does not assert that an unauthorised person accessed material, that a disclosure occurred or that a legal breach was established. The bounded result was ten corrected boundaries and two unresolved decisions kept visible, instead of treating technical consistency as evidence that every business permission had been settled.
The trade-off was additional owner review and a less uniform workspace model. I wanted the enterprise to see the cost of preserving the intended relationship rather than describe the exceptions as avoidable inconvenience. A folder’s location in a hierarchy can make an access rule easy to inherit, but cannot by itself explain why a particular person should receive the material. The CIO’s contribution was to make that distinction practical in the system while leaving the underlying permission with the people authorised to decide it. The two held cases mattered because closure was not a stronger outcome than a question still awaiting a responsible answer.
Holding archive projects whose next action lacked a settled contractual account
In 2025 an archive proposal treated completed campaign projects as ready for the next removal process. The completion status was useful for production administration, but it did not establish every obligation attached to the records. We reviewed sixty projects with the accountable business and contractual owners before allowing the proposed process to rely on that status alone.
Forty-five were eligible under the owner review. Fifteen had unresolved contractual questions and were excluded from the proposed action. The operating sponsor wanted a uniform rule to reduce accumulated archive work. Qualified colleagues asked that the unresolved relationships remain visible rather than be converted into a technical instruction. I supported the hold and a revised workflow that recorded the separate authority needed for the next action.
We used ₹53 lakh over twelve weeks to revise the process and its ownership account. The example does not prescribe a legal retention period, claim that the fifteen projects could never be removed or assert that completion automatically required continued storage. Actual contractual and information decisions remained with qualified owners. The result was a bounded eligibility review and a system that could carry an unresolved question without treating it as an error to be cleared merely to finish the workflow.
The trade-off was continued management of the held records and a less convenient rule than selecting every completed project. I wanted that consequence acknowledged rather than hidden inside an apparently efficient automation. The important distinction was between an operational milestone and permission for a later action. A completed campaign tells us something about production; it does not answer every question about the material the enterprise still holds. The CIO should help those decisions remain connected and attributable, so a routine status cannot quietly acquire authority that the business and its professional advisers have not actually supplied.