Chief Information Officer / India
The completed programme still had qualifications to explain
- Manufacturing systems career
- 30 years
- First appointment July 1996
- IT colleagues
- 28
- Four managers; reports to the CEO
- Annual IT expense
- ₹7.8 crore
- Separate ₹3.6 crore approved programme capital
Decisions & consequences
01Delaying a transfer until product history could be explained
The separation schedule assumed that transferring 420 product-history records would complete a defined part of the receiving company's systems readiness. Review found thirty-seven provenance exceptions. Some records could be read and copied but did not carry enough supporting history for the receiving business to rely on them in the intended way. I recommended that the transfer decision distinguish technical availability from the business acceptance of that history.
The programme office preferred to move the entire group on schedule and resolve exceptions after handover. That would have reduced the immediate delay, but it also risked presenting the receiving organisation with a completed milestone whose limitations were difficult to reconstruct. Operations leader Sanjay Belqathi supported a narrower acceptance account. We funded ₹64 lakh for the provenance work and delayed this transfer by six weeks. The cost included specialist review and the continued separation support required during the delay.
Thirty-one exceptions were resolved through the agreed process. Six remained restricted and were identified explicitly in the handover. The resulting account did not claim that every product history had been validated. Technical and business owners retained authority over the permitted use of the restricted records. IT's role was to preserve their status and the supporting evidence rather than remove a warning because the project was ready to close.
The choice sacrificed a cleaner completion date and required additional support expense. It also made the receiving company's obligation understandable after the original team dispersed. I would apply the same question to another transfer without assuming that six weeks or the same review method would fit it. The practical test is whether the recipient knows which records it can rely on, for what purpose, and which questions still require an authorised answer. A successful copy operation does not settle those questions on its own.
Keeping two measures outside a uniform group total
A reporting review found that eight of twenty-eight measures used similar labels for different quantities. Some local records expressed linear metres; others described saleable rolls whose lengths varied with the product and order. A consolidated display could therefore look precise while combining unlike things. The issue was not a broken calculation. It was an insufficiently defined comparison that the system could reproduce very efficiently.
Group reporting colleagues proposed applying a standard conversion so every measure could appear in the same dashboard. Operations wanted the product and order differences retained. I supported an agreed-definition programme with finance and operations ownership, costing ₹38 lakh for mapping, report changes and follow-up validation. That choice took longer than adding a conversion factor and required readers to accept a less uniform presentation.
Six of the eight measures became comparable under approved definitions. Two remained separately labelled because the underlying processes still differed. We did not conceal them in an aggregate or describe the programme as complete harmonisation. The report explained the scope of the comparison and preserved the original local quantities where they were needed. Finance retained accounting policy, while operations retained responsibility for the meaning of the production measures.
The outcome was a more defensible comparison within the reviewed set, not proof that every group report was reliable. I wanted the residual difference to remain visible even when it complicated the executive view. A systems leader should be able to explain why a common chart sometimes needs separate series. The cost of a uniform appearance can be a mistaken business inference, especially when readers assume that a shared label means a shared unit. Our decision kept two exceptions in view and made the basis of the other six explicit.
Retiring software without calling an estimate a realised saving
After organisational responsibilities changed, we reviewed 186 software entitlements rather than renew them solely from the prior year's list. Forty-two required reassessment of ownership, need or transition arrangements. A usage report offered one perspective, but it could not establish the contractual position or whether an infrequently used application retained an important business record. I asked procurement, legal and business owners to take those questions separately.
The cheaper proposal was to cancel all forty-two and reinstate any that users later demonstrated they needed. That would have made the renewal reduction visible quickly, but it shifted the burden of discovering a dependency into live operations. We approved ₹27 lakh for review, transition work and supporting arrangements. Thirty entitlements were retired. Twelve remained with named owners and documented reasons for continuing them.
Estimated annualised avoided renewals were ₹19 lakh. We reported that as an estimate associated with the changed renewal position, not as cash already received or a first-year net saving. The twenty-seven-lakh programme expense remained in the account. Procurement and legal colleagues retained contract interpretation, and business owners confirmed the operational consequences before removal. The work did not establish a wider software-licensing compliance opinion.
The programme left us with a smaller unresolved ownership problem and a clearer basis for subsequent renewals. It also showed that a responsible retirement can cost more initially than the annual amount it avoids. That is not a reason to reject it automatically, but it is a reason to state the economics accurately. I would rather present the transition cost, retained tools and estimate together than win approval with a headline that assumes every unused-looking entitlement is disposable. The board should be able to see what management stopped paying for, what it continued and why the timing of the benefit differs from the timing of the expense.
Assigning custody before acting on a storage-reduction proposal
A proposed storage-reduction exercise led us to review sixty-four business-record categories. Seventeen had no agreed business custodian for retention decisions. The systems team could identify where the material sat, but location did not establish who could authorise retention, restriction or disposal. I separated that ownership problem from the technical ability to remove files. The distinction mattered before any claimed storage benefit entered the investment case.
One proposal was to apply an age-based rule across the unresolved categories and handle objections afterwards. Legal and privacy colleagues asked for purposes and obligations to be reviewed first. We funded ₹43 lakh for classification, ownership decisions and supporting system changes. This route delayed the simpler clean-up and used specialist time that had not been included in the original estimate. I accepted that cost because the system should implement an authorised decision rather than make the policy by default.
Fourteen categories received approved custodians and handling instructions. Three remained restricted pending qualified review. The programme did not claim that the company had achieved complete legal compliance or that the remaining categories could be disposed of. Relevant Indian data-protection obligations were considered by the qualified owners alongside other applicable requirements; this case does not prescribe a retention period for another organisation.
The result was an accountable decision route for a defined group of records. It was not a dramatic reduction in stored volume, and we did not invent one to make the programme easier to celebrate. I wanted the unresolved categories to stay on the management account with a responsible review path. A CIO can make policy decisions executable and observable, but cannot replace the business and legal authority needed to decide them. The work improved the basis for later action while preserving the questions that still required an answer.