A quiet report can hide an expensive silence

In 2019, as director of people operations, I supported a site-management score that rewarded the absence of formal employee concerns. Eleven managers were included. I believed the measure would recognise early resolution and encourage managers to deal with issues before they escalated. Nirmal Verqashni questioned whether a zero count established either result. He warned that a manager could improve the score simply by making the formal route less welcome. I treated that objection as a communication problem rather than a reason to examine the measure itself.

A later listening exercise involved thirty-six conversations. Twelve employees described nineteen concerns they had not brought into the formal route. The accounts needed appropriate review; they were not nineteen substantiated findings of misconduct. Nevertheless, they showed that the absence of a record could not support the conclusion I had attached to it. I recommended removing the zero-count component from all eleven managers' reviews and acknowledged my role in approving it.

One regional manager objected that the revised account would reward sites that generated complaints and punish those that genuinely resolved problems. I agreed that a raw high count was no better a performance judgement than a raw low count. We replaced the count-based reward with review of access, response and unresolved matters, with appropriate confidentiality and responsible case handling. The ₹26 lakh programme funded listening, administration and follow-up; it did not purchase a finding that every concern was valid or a promise that disagreement would end.

The next quarter recorded twenty-seven concerns. Eighteen had received a response and nine remained open at the review point. A response was not labelled an agreed resolution, and the increased count was not presented as proof that working conditions had deteriorated. Managers had to explain more than a single reassuring total. The immediate consequence was a less flattering but more usable account and a cost for correcting a framework I had supported. I would now examine how a measure affects the willingness to use a process before allowing its result to stand for the health of the employment relationship.

In 2022, a review of field travel involved 132 eligible employees whose work required journeys between service locations. The policy allowed reimbursement of approved expenses, but some colleagues said that paying first was difficult within their household cash position. The account showed eventual reimbursement and therefore appeared complete to the central team. I asked finance and operations to examine the interval before the claim was settled, rather than treat an employee's ability to fund that interval as part of the job.

Finance preferred to shorten the reimbursement cycle instead of introducing advances. The concern was reasonable: advances needed reconciliation, unused balances could remain outstanding and managers might approve journeys without clear cost limits. We retained those control questions and designed a bounded ₹18 lakh advance pool for the reviewed travel cycle. Approved business purpose, spending limits, supporting records and return of unused amounts remained necessary. An advance was not an additional salary benefit or an unrestricted payment to the employee.

Of the 132 eligible colleagues, 124 used the arrangement in the first reviewed cycle and eight did not. The ₹18 lakh issued reconciled to ₹16.4 lakh of supported business expenses and ₹1.6 lakh returned. Those figures describe this cycle, not a permanent annual saving or proof that every future claim would be complete. The administration programme cost ₹9 lakh separately. Finance retained responsibility for recording and reconciliation, while HR checked whether the process was practically accessible to the employees it was intended to support.

The change required more work before travel and made the company carry funds earlier. Operations also had to provide better notice of planned journeys. It did not remove every exceptional expense or eliminate the need to question unsupported claims. The defensible outcome was that these reviewed journeys no longer depended on employees first finding the full amount from personal resources. I would have examined that condition when the travel policy was written. A reimbursement entitlement can be formally consistent while imposing a very different burden on people asked to use it.

Our 2023 internal supervisor-selection round received forty-eight applications. The first proposed format required an English presentation to a central panel. Some candidates regularly explained work and coordinated colleagues in Bengali, Assamese or Hindi, while the role also required enough communication with central teams to perform defined tasks. I asked operating owners to specify those tasks before HR treated presentation polish as a general proxy for supervisory capability. The question was not whether communication mattered, but which communication the job actually required.

A senior manager preferred the common English format because it appeared easier to compare and prepared successful candidates for later corporate roles. Nirmal and the operating assessors separated the requirements of the current appointment from possible future responsibilities. We provided structured discussion in an appropriate working language, with translation where needed, and retained explicit assessment of the central communication tasks. Technical competence remained with qualified assessors; changing the interview language did not change plant authorisation or remove the evidence required to supervise work.

Thirty-six candidates completed the revised process and twelve withdrew before completion. Twenty-two met the stated selection criteria, of whom sixteen received appointments against available vacancies. Six remained eligible for later consideration without a promise of appointment. Fourteen completed candidates did not meet the criteria and received an explanation of the relevant gaps. Withdrawals were not recorded as failed assessments. The process development and administration cost ₹21 lakh, and panels needed more preparation and time than the original presentation format.

The result was sixteen appointments made through a comparison tied more closely to the current work. It was not evidence that every successful candidate would perform well or that plant outcomes had improved. Managers still had responsibility for induction, support and review after appointment. I would establish the actual communication requirement before choosing the format next time. A standard interview can make administration feel fair while rewarding a skill whose relevance has never been properly examined; the remedy is a clearer standard, not the absence of one.

In 2024, urgent recruitment for eighteen operating-coordination roles led to offers above the prevailing range used for experienced colleagues in comparable responsibilities. The hires addressed a real staffing need. After they joined, twenty-seven existing employees raised questions about the comparison. The recruiting manager argued that the market had required the offers and that reopening existing pay would undermine the budget agreed for filling the vacancies. I accepted the market pressure but did not think it made the internal comparison self-explanatory.

HR and finance reviewed the actual responsibilities, experience and current pay across forty-five relevant positions. We did not assume that every matching title represented identical work or that every difference was unjustified. Qualified employment advisers reviewed the applicable requirements. The assessment identified twenty existing roles for an adjustment, with seven other cases requiring individual explanations rather than an automatic match. The CEO approved ₹88 lakh of additional annual employee cost for the twenty adjustments, separate from the original recruitment expenditure.

Finance asked whether the change could be postponed until the next ordinary review. That would have preserved the immediate budget but left the company asking people to wait after it had acted quickly for new recruits. We implemented the approved adjustments and retained the cost in the workforce account. The seven other colleagues received the basis of the decision and a route to question factual errors; receiving an explanation was not treated as agreement that the outcome was fair. The eighteen new hires were not blamed for accepting the offers made to them.

The review did not establish that all pay differences across the enterprise had been resolved. It addressed a defined group and changed the approval paper for later urgent offers to include the likely internal consequence. I would ask that question before the offer is approved, rather than wait for colleagues to force it into view. An external market explanation may justify a hiring choice, but leadership still has to account for what that choice does to the employment proposition it has already made to others.

Local recruitment near the plants frequently brought applicants with connections to existing employees. In a 2025 review, twenty-seven supervisors completed relationship declarations for a defined set of hiring decisions. Eleven disclosed a relevant family connection. The initial management response proposed excluding every connected applicant to avoid questions about favouritism. I opposed that automatic result. A disclosed relationship could affect who should take part in a decision without establishing that the applicant lacked merit or that misconduct had occurred.

The operating team was concerned about delays and the availability of assessors familiar with local work. We designed a process in which the relevant supervisor stepped out of the affected decision, another suitable assessor reviewed the evidence and the reason for the arrangement was recorded with restricted access. Qualified advisers helped review the applicable employment and privacy requirements. HR did not publish a family list or invite colleagues to infer improper conduct from the existence of a connection. Technical requirements and the responsible assessment remained unchanged.

The process covered sixteen recruitment decisions involving the eleven disclosed relationships. Nine applicants met the relevant selection criteria, five did not and two decisions remained pending because evidence was incomplete. Meeting the criteria was not represented as a confirmed appointment in this account. The programme cost ₹14 lakh for administration, assessor coordination and review. Some decisions took longer because the usual local supervisor could not simply complete the familiar route. We accepted that inconvenience rather than describe blanket exclusion as the only practical control.

The outcome was a documented way to examine a conflict in the decision process without turning kinship into an allegation. It did not prove that every undisclosed connection had been found or that the revised process could never be misused. We retained a route for questions about the actual conduct of a decision. I would make the disclosure and participation expectations clear before the next recruitment cycle. Employees should not have to choose between concealing a relationship and assuming that honest disclosure automatically ends a relative's application.