People decisions with a cost, a challenge and a consequence

In late 2021, Neralith's executive team reviewed complaints about how several relationship managers had pursued business. The existing incentive scheme gave financial production considerable weight. Risk colleagues wanted stronger conduct assessment, while the commercial head argued that any change would be treated as a reduction in earning opportunity. One proposal was to reinterpret the previous year's rules so that the new approach could affect payments already being discussed. I opposed that route. It would have made a desirable policy change depend on an explanation employees had not received during the relevant performance period.

I worked with chief risk officer Daniel Tersavan on terms for the 2022 cycle covering 320 relationship managers. We described the conduct factors, the decision process and the route for an employee to challenge an assessment. The remuneration committee considered the proposal within its remit. Managers then rehearsed conversations using different combinations of financial results and conduct concerns. Legal and risk colleagues handled the separate questions arising from specific prior conduct; changing future pay rules was not a substitute for dealing properly with those matters.

The commercial head remained concerned about retention. We did not pretend the scheme was costless. Twelve senior employees left during 2022, and exit accounts contained several reasons. I would not attribute every departure to remuneration or dismiss every departure as unrelated. The bank retained the new prospective terms, reviewed difficult applications quarterly and revised the manager guidance where the explanation had been unclear. What we gave up was the convenience of one immediate mechanism that appeared to solve both past concerns and future design. The resulting process was slower to introduce, but employees could distinguish the rules under which they had worked from the rules that would apply next.

The 2023 operations programme changed work performed by 160 employees. The initial timetable assumed that a training offer would be enough to support internal mobility. I asked receiving managers to identify specific duties, supervision and vacancies before we described the programme as a redeployment plan. Operations wanted to complete the move within nine months. Several business units wanted experienced recruits rather than employees learning a different activity. Those positions could not be reconciled by asking affected colleagues to show more enthusiasm for training.

We allocated S$2.8 million to training and transition support, excluding ordinary salaries. I negotiated temporary release arrangements with the sending managers and paired each proposed move with a receiving supervisor. Employees received a written account of the role, location, working pattern and pay implications. Some declined an internal option because it did not fit their circumstances. We treated that as a decision to discuss, not proof of unwillingness to change. Qualified local advisers supported the employment terms and separation process in each country.

By December, 118 colleagues had accepted internal appointments, twenty-seven had left through agreed separation and fifteen remained in time-limited roles while further options were considered. The fifteen were a continuing responsibility, not a success hidden in the placement figure. Finance accepted a longer period of overlapping cost than the original plan envisaged. We also stopped using course completion as the headline outcome because an employee could finish every module and still lack a suitable job.

The most useful change was the receiving manager's obligation to participate before the announcement. It made the transition a business decision involving real work, rather than an exercise delegated to human resources after the operating plan had been settled. The programme did not avoid every departure. It made the available choices more concrete and left a clear account of who still needed an answer.

During my Tersali appointment, a 2009 integration brought 240 employees onto a common job framework. A first proposal removed several inherited allowances at the conversion date. The spreadsheet showed simpler administration and an annualised S$900,000 reduction in expense. Branch managers objected that employees on late shifts had arranged transport and family commitments around those payments. I initially regarded the issue as one for individual consultation after the framework was agreed. Listening to the shift patterns changed my view: the allowance was connected to the way the bank scheduled the work.

I led six compensation specialists and worked with operations to separate job evaluation from the treatment of existing allowances. We retained transitional payments for eighteen months, with an estimated S$640,000 cost over that period, and reviewed the underlying shift arrangements. The finance director challenged the additional cost and the risk of creating a permanent exception. We agreed an explicit end date, individual written explanations and quarterly review of the remaining arrangements. Local advisers checked the employment implications; I did not present a group policy preference as authority to override existing terms.

The common framework went live three months later than planned. Twenty-two employees requested a formal review of their mapping, and eight mappings changed after further assessment. The delay and transitional cost remained in the integration account. We did not call the original S$900,000 estimate a realised saving. The experience taught me that harmonising a category does not necessarily make its consequences uniform. Two employees placed in the same grade may be affected differently by the timing of a change.

I still favour intelligible pay structures. What changed was my willingness to ask whether administrative simplicity was being purchased through a sudden burden on a small group of employees. A transition can be temporary, specific and affordable without pretending that the initial design was already fair in every practical circumstance.

In 2024, the nomination committee asked management to examine twelve critical positions. Seven had a candidate assessed as ready now; five did not. The development plans for several potential successors had appeared in successive annual papers with little change. Their managers said they were indispensable in their current jobs. I understood the operating concern, but it meant the bank was simultaneously calling these colleagues future leaders and refusing them the experience on which that judgement depended.

I proposed four nine-month assignments with defined responsibilities outside each person's usual reporting line. The S$780,000 programme allowance covered temporary backfill, travel and assessment support. It did not represent a salary increase for the participants. Two business heads opposed paying for cover, arguing that their teams could stretch. We reviewed the workload and rejected a design that simply required the participants to perform both jobs. One assignment began six weeks later so the replacement manager could complete a handover.

At the end of the assignments, two participants were assessed as ready for a broader role, one needed further exposure to commercial decisions and one decided that the envisaged position did not suit the kind of work they wanted. That last result was valuable, even though it did not improve the succession chart. We retained external search as an option for the uncovered roles. None of the four was promised an appointment at the start.

My contribution was to connect the committee's question to the operating resources needed to answer it. The programme cost more than circulating development objectives, and it required business heads to tolerate temporary unfamiliarity in important positions. In return, the bank learned something about actual responsibility, including a reason not to proceed with one anticipated succession path. I would rather present that mixed outcome than preserve a reassuring name whose willingness and readiness had never been tested.

In 2025, a senior manager at Neralith was being considered for a broader appointment when concerns about their treatment of colleagues were raised. The executive sponsor wanted a quick conclusion so the appointment timetable could continue. Another manager argued that delaying the process would itself imply guilt. I advised separating the appointment decision from the inquiry and communicating that the bank had not reached a finding. The person concerned was entitled to a fair process, and employees raising concerns needed protection from pressure to withdraw them.

We used a qualified external investigator with a defined brief and restricted access to the case material. I coordinated process support, while legal advisers addressed procedural questions and the designated decision-makers considered the findings. I did not conduct the inquiry or publish witness accounts. The appointment remained unfilled for eleven weeks, with an acting arrangement and clear limits on delegated authority. That created inconvenience for the business and an additional management burden; those costs were preferable to allowing the proposed promotion to determine the inquiry's pace or conclusion.

The final employment actions were handled confidentially. This public account does not identify a finding or reconstruct the individuals involved. The outcome I can describe is procedural: the appointment timetable did not control the investigation, the acting authority was documented and employees had a separate route to raise concerns about retaliation. We reviewed the adequacy of that support after the case closed.

The decision was difficult because every available action could be interpreted as a signal. I could not remove that uncertainty through a carefully worded announcement. I could insist on a defensible process, limit unnecessary disclosure and ask the executive team to carry the temporary operational cost. A people leader sometimes contributes most by resisting an artificial deadline rather than producing an immediate answer that satisfies the organisation's preferred narrative.