## The rule carried a judgement
In 2016 I approved a recognition scheme that appeared straightforward to administer. It used attendance over three months and described the resulting award as recognition for reliability. At the end of the assessment, 104 colleagues were excluded. Sixty-eight had only taken approved leave. The records were not missing and the calculation did what its rule required. The mistake was in what the rule allowed us to say about those people. I had treated the convenience of a measurable condition as sufficient support for a broader judgement about their contribution.
Approved absence and an unexplained failure to attend are not the same account of what happened. Our programme placed them on the same side of its eligibility boundary. Once an award was described as recognition for reliability, that boundary acquired a meaning beyond the attendance record. A colleague could reasonably hear that the organisation had authorised leave and then used it to withhold recognition of a valued quality. The contradiction did not become smaller because the calculation was consistent. Consistency can reproduce an unsupported interpretation just as efficiently as it can apply a sound one.
I was head of people operations, with responsibility for the proposal I endorsed. It would have been easy to describe the problem as a detail that an analyst had failed to flag. That would have misrepresented how the decision reached approval. The relevant question was available to me: what does exclusion communicate, and can the record justify that communication? I had concentrated on whether managers could explain the rule simply. I had not adequately examined whether its simple explanation would remain credible to someone whose authorised circumstances placed them outside the award.
The number sixty-eight gave the problem scale, but the principle did not depend on that being a large share. Even one case could have exposed the mismatch between the promise and the criterion. Conversely, identifying those sixty-eight did not validate the remaining thirty-six exclusions. We had no basis to label them all appropriate simply because their records differed. The arithmetic separated categories that needed examination; it did not complete that examination. This distinction matters whenever a review discovers a clear error inside a larger decision and is tempted to declare everything outside it correct.
## Cancellation had to explain the mistake
We stopped before awards were made. Design and administration had already cost ₹18 lakh. The expenditure remained, and managers had prepared for an event that would no longer happen as planned. There was a practical incentive to preserve the programme by changing its title or adjusting the explanation at the last moment. I did not think that would address the underlying issue. The workforce had already been told what the recognition meant. A less ambitious label might have protected the event while leaving the original message and the people affected by it largely untouched.
The explanation therefore needed to name the design error without turning it into a statement about every person's entitlement. We had used approved leave in a way that did not support our claim about reliability. That was enough to stop this award. It was not a complete legal analysis of leave, a finding about every exclusion or a promise that all forms of attendance recognition were impossible. Actual arrangements and professional advice still mattered. Keeping the explanation bounded made it more accountable, because it described the decision we had made and the reason we could no longer defend it.
Cancellation also created an uncomfortable asymmetry. The original proposal had been easy to summarise; the correction required more care. A short sentence about celebrating reliability had to be replaced by an explanation of why attendance and reliability had been treated as interchangeable when they were not. Leaders sometimes interpret that additional complexity as evidence that the correction will confuse people. I think the more useful question is whether the original simplicity was earned. If it depended on removing circumstances that changed the meaning, restoring those circumstances is necessary work rather than an avoidable communication problem.
We did not claim that stopping the scheme had improved trust or attendance. Those would have been separate outcomes requiring evidence. Nor could the ₹18 lakh be recovered by presenting the cancellation as a learning investment after the fact. The money was a cost of an approved programme we chose not to complete. I retain it in the account because an error becomes too easy to admire if its correction is described without inconvenience. The decision was better than proceeding, but it still reflected a failure in the judgement that had allowed the proposal to reach that stage.
## A replacement needs a defensible meaning
A replacement proposal should begin with the contribution the organisation wants to recognise, rather than the field in a system that is easiest to count. The distinction changes the discussion. If the purpose is to recognise colleagues who help a team deliver dependable work, attendance may be one piece of context, but it does not automatically describe all the relevant conduct. If the purpose is something narrower, its name and explanation must remain correspondingly narrow. The organisation owes employees an intelligible relationship between the award's stated meaning and the evidence used to decide it.
That relationship must be examined from the position of someone excluded as well as someone selected. Selection is often easier to celebrate than exclusion is to explain. Yet the latter can reveal whether a policy depends on assumptions its sponsors have not acknowledged. I would want a reviewer to describe an ordinary, foreseeable circumstance that places a capable colleague outside the rule and then explain why that result is appropriate to the stated purpose. The exercise is not a demand that every scheme include everyone. It is a test of whether a boundary is defensible in its actual context.
Managers also need a way to raise a contradiction before an event becomes too expensive or visible to stop comfortably. A programme can acquire momentum from its presentation, suppliers and calendar long before the final list is published. Review should therefore occur when the purpose and criteria are still open to change. The fact that a late objection is inconvenient does not make it weak, but earlier scrutiny gives the enterprise more choices. HR's role includes creating that opportunity and accepting that a challenge may concern the meaning of an approved design rather than a clerical mistake.
For directors, the case is a reminder that a people metric can become an institutional judgement when attached to reward, progression or public recognition. Oversight should examine that transition. What claim is being made about a person, and what evidence supports it? What commitments has the employer already made that the new rule needs to respect? Management must design and administer the programme, while the board can ask whether its account is coherent. My own example begins with a rule I approved and ends with a question I should have asked before the rule became an award.