I once defended a cost calculation before checking whether it answered the question in front of us. In 2018, our operating-planning comparison divided ₹24 crore of spend by 600,000 production starts. The result was ₹400 per start. The proposal allowed that number to be read as spend per accepted set, although the accepted quantity was 480,000. On the same planning basis, the second calculation was ₹500 per accepted set. The arithmetic in both divisions was straightforward. My mistake was to allow the first label to carry the meaning of the second when colleagues were preparing a commercial decision.
Name the count before defending the calculation
A production start is a useful event to count. It can help explain the work entering a process and the resources required along the way. An accepted set describes another point in that process. The two measures need not be equal, and neither becomes meaningless because they differ. What matters is the question being asked. In our proposal, the commercial discussion concerned the economics of accepted output. Using the start count without preserving its meaning gave a more favourable unit figure, but it did not make the business capable of selling every start as an accepted component. The count's label was part of the calculation's usefulness.
Arun challenged the comparison during the operating review. My initial instinct was to point back to the expenditure and the division. Both reconciled. That response missed his concern: reconciliation to source figures does not prove that the selected figures belong together for a particular purpose. A finance team can work carefully and still provide a misleading answer if it has not understood the operating event. I had to return to the meaning of the quantity before asking the commercial team to rely on the result. The correction concerned the framing of the proposal, not merely another check on a spreadsheet formula.
The 480,000 accepted sets were eighty per cent of the 600,000 starts in the reviewed account. That proportion helped us understand the relationship between the two measures. It did not establish a permanent acceptance rate for every product, factory or future period. The scope mattered because a specific review can easily become a general assumption once it enters a concise presentation. We needed to say which work the figures represented and where further analysis was required. Otherwise, correcting the denominator in one place could still leave the organisation using the resulting ratio more broadly than the evidence supported.
The ₹24 crore also needed a clear description. It was the spend used in this operating-planning comparison. We did not claim that dividing it by accepted output produced a statutory inventory valuation or that every item belonged in an accounting asset. Financial reporting has its own requirements and assessments. A planning question can legitimately examine a broader economic use of resources, but it should not borrow an accounting label that suggests another conclusion. I wanted the proposal to be understandable to commercial colleagues while remaining precise enough that finance specialists could see what had and had not been calculated.
Keep acceptance outside the price preference
Technical colleagues retained responsibility for the component acceptance criteria. A preferred cost figure could not determine whether a component met those criteria. This boundary sounds obvious when stated separately. It becomes less comfortable when a commercial deadline is near and the financial plan already assumes a particular unit answer. I did not want the technical team placed in a position where accepting more output appeared to be the easiest way to preserve a proposal finance had prepared incorrectly. The correct response was to change our account and the decision it supported, not to make the acceptance process accommodate the desired number.
Our customers manufacture ophthalmic instruments, while we supply non-optical housings and mechanical subassemblies. That position in the supply chain requires care about the authority we describe. Finance does not decide the clinical use of a finished instrument, and a planning comparison does not certify the component or the instrument. My contribution concerns the business resources and commitments around the accepted product. I need enough understanding of the operating process to interpret its evidence correctly, while leaving product approval with qualified colleagues. A CFO's influence is more useful when it respects that boundary and makes the financial consequences explicit.
We withdrew the initial comparison before price approval and rebuilt the proposal with both quantities visible. We also supported a separate ₹3.2 crore process programme. The programme was an additional commitment, not a deduction that made the original spending account more favourable. Its approval did not prove that acceptance had already improved or that the company had recovered an equivalent amount. Keeping those distinctions visible prevented the correction from becoming a new success claim before the operating evidence existed. We had identified a problem, corrected the decision paper and funded work; each statement described a different stage.
There was a commercial cost to reopening the discussion. Sales colleagues had expected a prompt answer and had organised their work around it. I needed to explain why the revised basis mattered without treating their deadline as an unreasonable demand. They were entitled to expect finance to provide a useful comparison the first time. Owning the error meant acknowledging that our clarification created more work and could change the proposal's attractiveness. It did not mean defending the original answer until somebody else accepted its consequences, nor presenting technical review as an obstacle that had unexpectedly complicated an otherwise settled decision.
Let the correction change the decision
A corrected number should be allowed to alter the next choice. If the organisation insists on retaining the same conclusion regardless of the revised basis, the review becomes an exercise in rewriting the justification. I wanted the price and process discussion to reopen with the accepted-output economics visible. That did not dictate one universal commercial response. The company might consider price, scope, investment or another combination of actions. The point was that the decision should be made with an account that answered its actual question, rather than with a more convenient denominator left in place to preserve momentum.
For a board, the lesson is not to inspect every production count directly. It is to ask whether management can explain the measures carrying an important proposal. What event does the denominator represent? Does it match the unit being priced, funded or compared? Which professional decision determines that event? Where does the measure stop being representative? These questions test the account without asking directors to become production engineers. A clear answer should help the board see the commercial proposition and the remaining uncertainties together, including where a plan depends on output that has not yet reached the relevant accepted state.
I also changed how our finance team discussed review. Checking a calculation includes checking the language around it. A heading, unit or short explanatory line can determine how the next reader uses a number. If the text says accepted output while the count says starts, the reconciliation is incomplete even when all the cells add correctly. This is a practical responsibility shared by the person preparing the analysis and the executive signing it. I had signed the proposal's interpretation, not only its arithmetic, and needed to make that responsibility clear when explaining the correction to colleagues.
The experience made me slower to defend a figure and quicker to ask what it describes. ₹400 per start and ₹500 per accepted set were not competing opinions about the same unit. They were answers to different questions using the same planning spend. Once that distinction was restored, the enterprise could reconsider the commercial choice and the process work on a clearer basis. A useful financial number does more than reconcile. It carries enough meaning for another person to make a decision without silently changing the question that produced it.