The decision changed when the next process was included

In 2017 I approved a sheet arrangement intended to improve material yield on a pharmaceutical-leaflet job. I was head of multi-site operations and gave the number of pieces obtainable from the sheet more weight than the grain direction required by the later fold. The resulting problem appeared during conversion. Of 1.2 million printed leaflets, 900,000 had the unsuitable orientation; 300,000 were separately reviewed as suitable. We held the entire quantity before any dispatch.

The production team initially wanted to recover the schedule by adjusting the folding operation. Irene Zeqthuli required the component to meet the agreed physical specification rather than accept a weaker result because the printing had already been done. The commercial team also needed a customer explanation. I had approved the arrangement, so it would have been incomplete to describe the event only as an issue discovered by quality. My decision had created the work that the next process could not complete as intended.

We remade the 900,000 affected leaflets and released the 300,000 suitable ones after separate review and the customer's agreed disposition. Rework cost ₹38 lakh. The replacement quantity arrived five working days after the agreed date. Those consequences remained in the case; the successful hold did not erase the delay or make the first run a productive use of resources. No leaflets had been sent before the hold, and the example makes no claim about a clinical event.

I changed the approval discussion for sheet arrangements to include the later fold and the person responsible for confirming it. That was a change in how we prepared a job, not evidence that all future defects had been eliminated. The lesson for my leadership was specific: a locally efficient use of paper can be a poor production decision when it ignores the next operation. I needed to reopen my own judgement and pay for the consequence rather than ask another team to make the approved mistake work.

Our 2022 capital discussion began with a ₹6.4 crore proposal for additional printing capacity. Existing monthly print capability was twenty-four million leaflets, matching the current demand total. The job mix mattered: six million were supplied flat and eighteen million needed folding. Available folding capability was sixteen million. The apparent pressure to print more was therefore not the complete account of the work waiting to be delivered.

I asked the planning team to trace the demand through its required route. The two-million monthly gap sat in folding for that mix. Adding a press would create more printed work ahead of the same constraint. The proposal had support because the press was visible and the printing department was busy. I had to explain why activity at that stage was insufficient evidence for the proposed capital allocation, without dismissing the department's genuine workload or assuming that one new asset would settle every operating problem.

With the CEO and technical colleagues, I put forward a ₹3.8 crore folding programme. The board approved that scope instead of the press proposal. It was installed in late 2023 with a rated monthly folding capability of twenty million leaflets for the reviewed mix. In the subsequent specified review month, completed folding was 17.2 million. We reported the actual output separately from the rating; we did not describe twenty million as an achieved monthly result simply because the equipment had been commissioned.

The two proposals had different purposes, so the ₹2.6 crore difference was not presented as a like-for-like saving. Nor did the investment remove the need to manage changeovers, staffing and approved work arriving from print. My contribution was to make the capital choice follow the process that constrained the promise. The remaining difference between capability and completed work stayed visible, giving the next operating review a real question rather than a premature declaration that the factory's capacity problem had been solved.

In a 2023 review, our dashboard showed 96 of 120 orders delivered by their agreed dates. The 80% result was correct at order level. The orders were not equal in size, however. The ninety-six on-time jobs contained fourteen million leaflets, while the twenty-four late jobs contained sixteen million. Across the thirty-million-leaflet cohort, approximately 46.7% of the units were on time. The single order percentage gave a more favourable impression of the volume customers were waiting for.

Commercial colleagues challenged the account because several large customers were still dealing with substantial shortages. The operating team defended the dashboard: it used the same measure as earlier reviews and had not miscounted the orders. Both points were true but insufficient. I asked for the quantity view beside the order view, with the original agreed dates preserved. Changing a customer's promised date after the event would have made neither account more useful.

We restated the cohort with both measures and examined the concentration of late volume. I did not replace one number with the other as though quantity alone captured every customer consequence. A small urgent order could still matter greatly. The paired account allowed the management discussion to ask why a few large late jobs dominated the shortfall while retaining each order's commitment. The correction changed the report and the priorities for investigation; it did not retrospectively make any shipment on time.

The immediate cost was a less comfortable performance discussion and the need to reopen assumptions about scheduling. I resisted attaching an invented improvement rate to that change. The case ends with a more accurate account of the reviewed cohort, not a claim that the next month's delivery had already recovered. My operating lesson was that a familiar percentage needs its unit and distribution. A leadership website should show the decision to expose a weakness without turning the act of exposure into an unmeasured business success.

A 2024 customer programme comprised eighteen language layouts and three million leaflets. Fourteen layouts, representing 2.1 million units, had the required customer approvals. Four layouts covering 900,000 units remained pending. The commercial request described the programme as a single urgent order. Treating that description as permission to print all eighteen versions would have asked our production organisation to resolve a content decision it did not own.

I worked with Irene and the customer team to separate the approved and pending parts. The alternative was to hold the whole programme until every version was ready, which would delay material whose content had already been approved. We proposed a partial release with an explicit quantity and layout list. The customer agreed that scope. The pending versions remained outside the release until their approvals arrived, rather than being treated as minor variations of the approved files.

The 2.1 million approved leaflets proceeded under that agreement. The remaining 900,000 followed later after the required approvals. The revised sequence added ₹28 lakh of expediting expense to the programme. We retained that cost in the commercial review; a partial release was not a cost-free workaround. The case does not claim that printing staff checked the clinical meaning of the translations or that a delivery decision established compliance of the finished medicine. Those responsibilities remained with the appropriate customer authorities.

My contribution was to make the production permission narrower and clearer than the urgent request. It required negotiation about the schedule, quantities and extra expense rather than a general instruction to move faster. I also wanted the record to show what remained pending after the first release. Completing one agreed part should not make the overall programme sound finished. The experience reinforced a distinction I bring to operating discussions: responsibility for arranging the work includes recognising the decisions that the supplier cannot legitimately make for its customer.

In 2025 we reviewed the protective arrangement for a 600-carton shipment format. The existing plan put forty cartons on each pallet, requiring fifteen pallets. A proposed lower stack held thirty cartons, requiring twenty pallets for the same shipment. At the agreed illustrative rate of ₹1,100 per pallet space, the additional five spaces cost ₹5,500 per shipment. For a hundred shipments in a quarter, that meant ₹5.5 lakh more transport expense.

The dispatch team was concerned about losing loading efficiency. Qualified packaging colleagues wanted the physical conditions assessed before the higher stack continued by default. I supported a trial and a two-quarter expense envelope of ₹11 lakh for the hundred-shipment-per-quarter planning volume. The amount was an approved allowance for the changed arrangement, not proof of the eventual number of shipments or a full-year cost commitment.

The reviewed trial covered forty-two shipments. Thirty-eight met the agreed packaging acceptance conditions; four remained unresolved and required further examination. I kept those four in the report. The result supported a bounded operating decision under the reviewed conditions, but it did not establish that the lower stack would prevent every form of damage across every route. We did not infer a clinical benefit from a packaging trial or compare it with an incompatible historical sample to manufacture an improvement percentage.

I asked the customer and packaging teams to retain the conditions of acceptance and the questions still open. The additional expense made the consequence visible in the operating plan, while the unresolved shipments prevented the trial from becoming a universal claim. My role was to connect the physical arrangement with its commercial cost and limits. A higher loading rate would have been cheaper on the transport sheet, but that sheet alone could not decide whether the agreed component would arrive in an acceptable condition. The decision remained open to further evidence rather than becoming a slogan about quality.