A spare part belongs to a revision before it belongs to an order

A familiar description can make a replacement request feel straightforward. The customer names a product family, the service colleague recognises it and a component appears in the catalogue. Each step seems to reduce uncertainty. Yet a distinction may still be missing: which revision of the product is actually in use? If that revision changes the component required, the familiarity of the name can conceal the question that matters most. I learned to ask whether our service process preserved that question before it moved an order towards dispatch.

We reviewed 170 spare-part requests at Kelthavi. Twenty-nine selections referred to an incompatible product revision. The finding did not mean that every request was wrong, and it did not tell us the position of every product in the field. It identified a defined set of selections that needed correction and a process whose distinctions deserved examination. I wanted the response to be more specific than a reminder to take care. Colleagues needed a reliable way to obtain the compatibility decision, not simply a stronger instruction to make it correctly.

## Recognition is not the compatibility decision

Product families are useful for many purposes. They help people describe what they own and organise the way a company sells or supports its range. But the category that makes a conversation easy may be broader than the category needed for a technical replacement. A revision can matter even when the outward appearance or commercial name remains familiar. The operating process must preserve the distinction that engineering considers relevant, rather than assume that the most convenient label is sufficient for every later decision.

We funded ₹54 lakh of compatibility-record and selection work. Engineering retained authority for the relationships between product revisions and components. Service needed a way to identify the product concerned, while operations needed to make an unresolved question visible before dispatch. Those responsibilities were connected but different. A service colleague could gather the necessary facts without becoming the engineer who approves a technical relationship. Operations could improve the selection path without declaring compatibility through the act of processing an order.

The difficult cases were those in which the history could not be established immediately. A process designed only for complete records may work smoothly in a demonstration and become unreliable when it encounters an older product or an unclear description. We needed a route for uncertainty, with a responsible owner and an explicit pause where the missing fact mattered. That pause should not be treated as an employee's failure to keep the queue moving. It is part of doing the work correctly when the available record does not yet support the next decision.

The commercial account should recognise that work. A held request can look inefficient if the measure rewards only the speed of dispatch. Yet a quick order that sends the wrong revision may create another service problem and leave the customer waiting longer. I would not solve that tension by inventing a universal cost for an error or by claiming that every pause is justified. I would ask management to examine the reason, the facts required and the owner capable of resolving it. The measure should support that judgement rather than erase it.

We rechecked the twenty-nine known cases against engineering-approved records and resolved their selections. That was a concrete result. The corrected cases helped demonstrate that the revised arrangement could carry the distinctions those requests required. They did not establish that every future request would be right. I wanted the report to retain that boundary because a known-case correction and a broader reliability claim are different statements. One should not silently become the other as the account moves from a service discussion to an executive summary.

## The known correction is a starting point

There are further questions after the identified cases have been resolved. Can colleagues use the revised records when the product history is incomplete? Does the process make the responsible technical owner clear? Are new revisions reflected in the service material before the next request arrives? Those questions concern the continuing operation of the arrangement. They cannot all be answered by showing that the first twenty-nine cases were corrected. Management should identify which evidence it has and which work still requires examination, rather than use one successful check to close the entire subject.

The same discipline applies to the way the customer is informed. A person should not be told that a replacement is suitable merely because it has been located in a catalogue. The explanation should reflect the decision actually reached and the conditions that remain relevant. Where the product needs professional assessment or the record is uncertain, the next step should be clear. Operations should help make that response understandable and timely without filling a gap in knowledge with confidence. A convenient assurance can be harder to repair than an honest account of the details still needed.

For a COO, the case is about organising responsibility around product history. Manufacturing may know why a revision changed, engineering may hold the technical relationship and service may receive the question years later. If those pieces do not meet in a usable process, the customer experiences the gap as a failed replacement. The solution is not necessarily to put every decision into one department. It is to make the handover between responsibilities explicit enough that the needed distinction survives beyond the team that first understood it.

A board can examine the quality of that account without approving individual parts. It can ask which product populations depend on revision-specific service, how uncertainty is handled and whether management's measures make an unresolved technical question visible. Directors should understand the resources behind the response and the limits of the result being reported. They should not use a general service improvement claim as a substitute for the relevant professional assessment. Oversight is more useful when it tests whether the organisation has connected the necessary responsibilities than when it asks for an attractive completion percentage alone.

I would describe our work as a funded correction to compatibility selection, with twenty-nine known cases resolved and a clearer route for uncertainty. It is a practical account that leaves room for continuing examination. The lesson extends beyond mobility products: the name people recognise is not always the fact that should decide what happens next. An operating process needs to preserve the distinction that makes the decision valid. A spare part belongs to the product revision it can serve before it belongs to the order the business would like to complete.