Make the promise usable where people encounter it

In 2021, while director of people operations, I supported an anonymity promise for an employee survey. The proposed manager report omitted names. I accepted that feature too readily as evidence that the promise was sound, without examining how the available filters would work together in a small shop or a distinctive role. The assurance was part of a people process for which I held responsibility.

The pre-release review combined store, role and tenure. Fourteen resulting cells contained one person. A recipient who knew the team could potentially infer who a response described even without a name field. We held the report before manager release. This fictional account does not claim that a respondent identity was published, that a particular inference occurred or that a legal breach was established. The finding was sufficient to challenge the promise before we relied on it.

I took the issue back through the CHRO and supported ₹24 lakh for report redesign, permissions and explanation. Qualified privacy review was necessary alongside the people-function work. The correction concerned the report's context and use, not merely the removal of another column. Managers also needed to understand why some apparently useful breakdowns could not be offered on the original terms.

The outcome was a held release and a funded correction. I do not claim universal anonymity from a revised design or a measurable increase in trust. My error remains the early endorsement of an assurance without adequate examination of the recipient's perspective. The case changed the question I ask before signing off a people promise: what can someone infer from the whole view, including what they already know about the people behind it?

In 2022, thirty-two adults returned to work after extended family-care breaks. A common programme offered a practical way to prepare managers and arrange cover, but it could not decide every person's hours or circumstances. I wanted the approval to support workable agreements rather than turn the word return into an assumption that everyone would resume the same pattern immediately.

Twenty-one colleagues resumed usual hours and eleven agreed phased hours. We allocated ₹28 lakh to a twelve-week programme of cover and manager preparation. Those numbers describe the agreed arrangements and programme period. They do not identify diagnoses, assign care responsibilities by gender or establish improved health, retention or productivity. The people concerned remained adults with different circumstances, and unnecessary personal detail did not belong in a management summary.

My team worked with operating managers and qualified employment owners to make responsibilities and review points clear. COO Sonal Deqvezi retained store operations. A manager needed to understand what cover was available and what had been agreed, while the individual needed a credible account of the arrangement. The existence of programme funding did not authorise an informal promise beyond what the company could deliver or replace the review required for an actual employment decision.

The immediate result was a set of supported return arrangements, not a universal model or a finding about longer-term outcomes. I would not present eleven phased schedules as evidence that the other twenty-one needed no support. Nor would I treat the programme as a statement of legal entitlement. The useful people-leadership judgement was to retain individual agreement inside a shared operating programme and to state what the evidence did, and did not, establish.

The annual team award used in twelve shops paid nothing below 100% of sales plan and ₹12,000 at or above it. For three hundred eligible adults, a very small difference around full plan could therefore create a large payment difference. In 2023 I brought a redesigned formula to the board with qualified reward and employment review, including a clear explanation of who could receive less as well as more.

The proposed formula paid zero at or below 90%, then ₹1,000 per percentage point above 90%, capped at ₹20,000 at 110%. At 98.8%, the new award would be ₹8,800 rather than zero. At 100.2%, it would be ₹10,200 rather than ₹12,000. Exactly 100% would produce ₹10,000, also below the old ₹12,000. Those examples made the tradeoff visible instead of describing gradual payment as an improvement for every participant.

The maximum annual envelope was ₹60 lakh: three hundred eligible people multiplied by the ₹20,000 cap. Board approval of that ceiling was not an actual payout, a profit result or evidence of improved retention. We needed managers to explain the calculation and the transition accurately, including what the scheme measured and the boundaries of the examples. This fictional award is separate from any conclusion about statutory bonus obligations.

I recommended the redesign because the previous cliff deserved scrutiny, while retaining responsibility for the choices inside the replacement. A smoother curve does not settle whether the performance measure or award level is appropriate. The result was an approved design and maximum envelope with explicit tradeoffs. The case demonstrates how a CHRO can make reward mechanics understandable before an attractive description conceals the experience of people whose payment may fall.

In 2024 we introduced a new employee-concern channel in twelve shops, supported by ₹33 lakh for casework and manager preparation. The next quarter contained thirty-one concern records, compared with twelve in the previous quarter. Management wanted to understand whether the change showed a healthier voice environment. I argued that the count alone could not answer that question.

At the reporting cutoff, nineteen records were closed and twelve remained pending. These were records rather than a verified count of unique people. Their closure status did not establish satisfaction, resolution on the merits or an absence of further concern. The matters described role and shift clarification, but the aggregate account was not a suitable place to expose an individual's circumstances or allow a small-shop identity to be inferred.

Several explanations for the higher count remained possible. Access to the channel may have changed; familiarity with its purpose may have changed; underlying concerns may also have differed. The comparison did not isolate those effects. I kept the figures in the management account while resisting a claim that the programme had increased trust or measured the frequency of harm. Qualified case owners retained responsibility for the actual matters and their treatment.

The funded programme established support for handling and explaining the channel. It did not earn a positive culture conclusion merely because more records appeared. My contribution was to distinguish use, status and experience so that the next enquiry could be chosen responsibly. A board or CEO should see what the channel records and what remains unknown, rather than receive a reassuring score created by attaching a favourable interpretation to an ambiguous movement in volume.

A 2025 communications review covered 240 colleagues in eighteen shops. We found that 176 had reliable access to the work app during paid shifts and sixty-four did not. A central publication count could not describe the experience of those sixty-four people. I asked that the communication plan address the route to the message before management treated publication as evidence that colleagues had been informed.

We approved a ₹46 lakh programme for shared terminals and paid briefings. The programme did not require employees to use personal phones or read work messages off duty. Operating managers and the people team needed to make the route usable within the working day, while respecting privacy and the purposes for which access was provided. The review was about access conditions, not a mandate to monitor personal devices or private activity.

The distinction between access and comprehension remained important. A terminal can be available without every colleague receiving a particular message, and receipt does not establish understanding. Approval of the programme therefore did not support a claim that all 240 people had read or understood every communication. The next review would need evidence appropriate to the actual message and audience, without turning a practical access check into intrusive individual surveillance.

My role was to make the missing condition visible and support a funded response alongside store operations. The outcome recorded here is the identified gap and programme decision, not a completed universal communication result. It illustrates why a people leader should examine how a central process reaches a working colleague. A policy cannot be considered usable solely because the organisation has successfully placed its words on a system.