The decision includes the expense and the people whose situation is not yet resolved.

In 2011, Samvetrik Digital received 147 requests for employees to move between delivery units. Fifty-eight releases occurred after the dates communicated to the individuals. The receiving managers saw missed staffing commitments; current managers saw unfinished work; employees saw a company that could not explain when its promise would become real. I was head of workforce planning and had helped produce the allocation reports that made the intended dates look more settled than they were.

I proposed that a release should be recorded as agreed only when the current manager, receiving manager and employee understood the handover. Account leaders objected that this would expose uncertainty in plans already shared commercially. I accepted that consequence. An uncertain date was not made dependable by placing it in a confident column. We separated requested dates from agreed dates and required an escalation when the two could not be reconciled.

Delivery leaders retained responsibility for customer commitments and technical handover. My team coordinated the review and tracked the individual agreement. We also obtained ₹62 lakh for planned overlap on selected transfers. That cost was preferable to assuming a receiving team could absorb the move without preparation, but it reduced the short-term financial attraction of several transfers. Some requests were delayed with an explanation rather than approved on an unrealistic timetable.

In the next tracked cohort, twenty-four of 142 releases were late. The cohorts involved different work, so I do not claim that the process alone caused the reduction. The stronger result was a record that distinguished a wish from an agreement and brought unresolved cases to a named decision-maker. I would have involved employees earlier in designing the communication. We initially concentrated on management agreement and learned that a technically workable handover could still leave the person unclear about their new responsibilities.

A planned 2020 client mobilisation was cancelled when 312 Samvetrik Digital employees had already been identified for the work. Some were finishing earlier assignments; others had recently joined. Delivery leaders wanted time to find alternative demand, while finance challenged an open-ended period without billable work. Employees needed a clear account of their position, not repeated reassurance that another project was close.

As CHRO, I recommended a sixteen-week paid transition programme with an ₹8.4 Cr funding envelope covering the employment and preparation costs identified for that period. The chief executive asked what evidence would justify extending support. I proposed individual assignment prospects and practical development needs, reviewed at defined points, rather than a general assertion that the market would improve. Finance accepted the bounded expenditure; account leaders accepted that tentative opportunities would not be counted as completed placements.

People partners met each employee, while delivery teams assessed available work and technical leaders advised on preparation. We described what the company could commit to during the funded period and what remained uncertain afterwards. The programme did not require employees to describe their waiting time as a development opportunity if the proposed learning had no credible connection to future work. It also did not remove the need for appropriate employment advice on individual decisions.

At the main review, 232 people had entered other client assignments, 36 had accepted external opportunities and 44 still needed an individual resolution. Those categories sum to the original 312; none was counted twice. We reported the remaining 44 prominently and assigned follow-up decisions. I do not present the programme as universal redeployment or permanent employment assurance. It bought a defined period to make better decisions. I would now set out the final review criteria at the opening conversation more clearly, because uncertainty becomes harder to manage when employees learn the decision standard only near the end.

In 2023, Samvetrik Digital planned to expand cloud-support work while many employees remained concentrated in established application services. A proposed learning programme measured success mainly through course completion. Practice leaders questioned whether that would tell them who could handle the actual support tasks. Employees wanted to know whether completing the pathway would lead to a relevant assignment. The two questions were related but not identical.

I asked for three distinct milestones: completing the learning, demonstrating the agreed capability and entering work where that capability would be used. Technical leaders defined the practical exercise and the level of supervision expected on an initial assignment. My team coordinated access, protected learning time and an explanation of how results would be reviewed. We approved ₹1.6 Cr for the programme, including assessment and preparation costs. Delivery managers objected to releasing experienced staff during busy periods, so the schedule was phased rather than based on an unfunded expectation of evening study.

Of 240 enrolled colleagues, 211 completed the learning. Among those completers, 167 demonstrated the required practical capability. Within six months, 126 of that qualified group had entered relevant assignments. The remaining 41 qualified colleagues were still waiting for suitable work; 44 completers needed further practice; 29 enrollees had not completed. We examined those categories separately. The figures did not establish that everyone was interchangeable or that assessment alone predicted performance on every customer engagement.

The programme changed the questions in our investment review. Additional teaching would not necessarily help a capable employee waiting for an opening, while a new assignment would not resolve an identified practice gap without support. The trade-off was a less flattering headline and more work tracking individual progress. I would have agreed the likely assignment sequence earlier with commercial leaders. We were right to separate capability from deployment, but some of the waiting period could have been anticipated before enrolment.

In 2024, a customer start date moved after 96 candidates had accepted employment offers from Samvetrik Digital. Several had already resigned from previous jobs. The commercial team believed the delay would be temporary, but it could not give a reliable revised mobilisation date. Finance proposed deferring joining where possible. Recruiters warned that another vague date would transfer the company's demand uncertainty into individual households.

I recommended retaining the agreed joining dates and creating funded interim work and preparation. This was a specific decision for the commitments already made, not a claim that every future offer could ignore commercial risk. We secured ₹2.7 Cr for the interim employment and preparation plan. The chief executive accepted the cost after reviewing alternatives and the evidence of commitments communicated. Legal and employment colleagues advised on the individual arrangements; I did not treat a general policy statement as a substitute for those reviews.

Delivery leaders identified useful internal assignments and relevant preparation, with clear supervision. We avoided assigning invented work merely to make utilisation appear stronger. Each new colleague received an explanation of the delayed engagement and the alternative plan. Some were interested in different opportunities within the company. We discussed those options without presenting the original project as certain to resume or implying that accepting a different role would have no career consequences.

Within twelve weeks, 74 colleagues entered the intended engagement and 22 agreed other assignments. All 96 joined on the dates the company had committed to. The interim cost remained visible in the account review rather than being described as a learning saving. The experience changed how we documented hiring assumptions before making an offer. I would now require a clearer contingency discussion earlier in the approval process. The lesson was not that a forecast can be made reliable through optimism, but that the company must know which personal commitments it is creating when it acts on that forecast.

In 2016, as vice president for people, I supported a delivery-management scorecard with a strong emphasis on utilisation. We expected it to improve commercial discipline and reduce unexplained unassigned time. Over two quarters, people partners reported that managers were postponing mentoring sessions and resisting internal releases because those decisions could weaken their immediate result. The measure was encouraging a narrower definition of good management than the company intended.

I initially believed better guidance would solve the problem. Conversations with twelve project managers changed my view. Several could show that they had followed the scorecard rationally while creating difficulties elsewhere. The issue was not simply an unwillingness to collaborate. We had asked them to support development and mobility while making those choices look like avoidable financial underperformance. I took responsibility for the people-function's role in promoting the design.

We suspended the single-metric emphasis for the next review cycle and worked with delivery and finance on a broader assessment. Utilisation remained relevant, but planned mentoring, agreed internal releases and known staffing dependencies were discussed explicitly. We did not invent a universal composite score that claimed to resolve every trade-off. Managers needed a documented conversation about circumstances and contribution. Preparing that review cost ₹38 lakh in management time, facilitation and revised reporting support.

Across the twelve teams examined, forty-six planned mentoring sessions had been postponed during the earlier period. Thirty-four were subsequently completed and twelve rescheduled with named owners. That is a limited observation, not proof of a company-wide productivity improvement. The more durable change was a requirement to test a proposed measure against plausible management behaviour before using it to judge people. I would have run that challenge before launching the scorecard. A metric can be mathematically consistent and still reward conduct that undermines the operating model it is meant to support.