The question is who holds the responsibility, and how they exercise it.

In 2021, Thalverin Retail had twenty-four store-manager appointments to make. Seventy-two candidates entered the final review, drawn from inside and outside the company. Existing selection discussions relied heavily on trading results and the confidence of the sponsoring district leader. Those were useful inputs, but they did not consistently explain how the candidate developed colleagues, dealt with disagreement or exercised authority when the commercial answer was not the whole answer.

As CHRO, I proposed a more explicit account of the work the appointment would carry. Operations colleagues did not want a process that separated store leadership from selling. Internal candidates worried that an unfamiliar assessment exercise could discount years of contribution. I agreed that neither concern should be dismissed. We retained trading evidence and asked candidates to examine management situations alongside examples from their own work. The purpose was to add relevant evidence, not replace one narrow measure with another.

We funded USD 420,000 for assessment and the operating cover needed for participation. Selecting managers had to record the responsibility they had examined, the evidence supporting their judgement and the uncertainty that remained. A panel could question the account, but the process did not pretend to remove managerial responsibility for the appointment. Qualified people colleagues reviewed the fairness and consistency of the approach. We did not publish personal candidate material or rank employees publicly.

The appointments comprised eighteen internal and six external candidates. That partition describes who was appointed; it is not proof that the process produced better managers. We arranged support after appointment and reviewed actual work separately. Some unsuccessful internal candidates still found the explanation difficult, particularly where strong sales had previously been treated as an implicit promise of progression. I would explain the change in appointment criteria earlier in a future exercise. A company cannot quietly change what it recognises and expect a new assessment form to repair the expectations created by its earlier messages. The durable improvement was a more inspectable appointment decision, with a clearer distinction between past contribution and the responsibility being offered.

A 2023 pilot across thirty-six stores used customer feedback as an automatic input to individual performance ratings. I had supported the proposal because the company wanted frontline behaviour to receive more attention, not only sales. During the pilot, eighty-three ratings were contested. Some accounts concerned service within the employee’s control. Others mixed product availability, store policy and an interaction with several colleagues. The number looked more precise than the judgement behind it.

Operations wanted customer criticism taken seriously and worried that changing the measure would weaken that commitment. Store managers wanted a process they could explain without becoming investigators of every comment. Employees wanted the opportunity to respond to an account that affected their record. I asked the review team to examine the disputed cases rather than argue only about whether customer feedback was good or bad in principle. The question was how it was being used in an employment decision.

We revised twenty-nine ratings and retained fifty-four after reviewing the available evidence through the appropriate management process. I recommended withdrawing the automatic input, while keeping relevant customer accounts as evidence that a responsible manager could consider alongside other material. The company approved USD 160,000 for redesign, manager preparation and review support. Specialist colleagues advised on individual cases. The revision was not a finding that every customer account was unreliable or that every contested rating had been wrong.

The pilot exposed a mistake in my own reasoning: I had accepted consistency of calculation as a proxy for consistency of judgement. We retained the customer perspective, but required the employment conclusion to have an accountable author and a basis that could be explained. I would test the difficult cases before adopting a similar measure again, including situations where the employee had little authority to change the source of the complaint. The outcome was a changed decision process and corrected individual records where justified. It was not proof of higher customer satisfaction, lower employee turnover or a universal solution to performance assessment. Those would require different evidence.

In 2016, a network review led Thalverin Retail to close nine stores. The decision affected 286 employees. I was vice president for field human resources and worked with operations, finance and qualified employment advisers on the transition. The property and trading timetable was clear before every individual employment arrangement was settled. That difference created a risk that the project would be described as complete while some colleagues were still waiting for an answer.

The initial arrangement comprised 177 transfers, 81 agreed departures and 28 colleagues who stayed to support the closing period. The company approved USD 2.1 million for the employment transition, including agreed support and departure costs. We treated those groups separately because they faced different questions. A transfer required a real role and a workable arrangement, not merely a vacancy elsewhere in the network. A colleague staying through closure needed to know what would happen after that work ended.

Operations wanted experienced people available until the final trading and stock work was complete. Employees understandably wanted certainty about their next position. I opposed counting the twenty-eight temporary stayers as a settled outcome simply because their immediate work had been agreed. We retained individual follow-up and a named decision route beyond the closure date. Local requirements and personal circumstances were reviewed with qualified colleagues rather than inferred from a group announcement. We also kept unresolved choices visible in the executive transition account.

The remaining group eventually resolved into nineteen transfers and nine agreed departures. The final employment account was therefore 196 transfers and 90 departures, covering all 286 employees. That arithmetic does not make the experience painless or establish that every person preferred the result. It makes the company’s account complete enough to examine. I would build the post-closure follow-up into the original project plan more explicitly next time, with responsibility surviving the end of trading. A store can stop operating on a particular date. The obligations and explanations created by that decision may continue, and the people function should ensure they are not lost when the property project moves on.

In 2009, while I led field development, sixty-four supervisors were due to complete practical management assignments. The programme assumed they could combine the work with their ordinary store responsibilities. Several store managers supported development in principle but repeatedly postponed participation when trading needs increased. The central team initially treated the problem as weak commitment. Store visits showed that the plan had allocated learning activities without allocating the cover needed to undertake them.

I worked with operations to cost a limited cover arrangement. Finance questioned whether the programme should be redesigned to take less time. Store leaders wanted to know who would carry the immediate work, while supervisors wanted participation to be a real part of their job rather than something squeezed around it. We simplified some activities, but retained the assignments that required observation, preparation and a management conversation. Removing all of that time would have changed what the programme could teach.

The company approved USD 118,000 of operating cover. District managers agreed the practical arrangements with participating stores, and my team made the assignment requirements more explicit. The funding did not give us permission to ignore trading conditions. It made the tradeoff visible and allowed a manager to raise a real constraint before another deadline passed. We also distinguished incomplete work from poor capability: a supervisor who had not been released to practise had not necessarily failed to learn.

Fifty-one supervisors completed the assignments within the original period; thirteen required an extension. I kept both groups in the report and did not describe completion as proof of readiness for every management role. The assignments produced evidence for a further conversation about responsibility. I would start with the store-cover question earlier in a repeat programme, rather than treating it as a delivery problem discovered after launch. The experience changed how I described development expenditure. The central materials and facilitators were only part of the cost. If the business wanted practical learning during paid work, it also had to make a deliberate choice about the work that someone else would carry while it happened.

Eighteen experienced store managers declined consideration for district roles in 2024. The succession discussion initially treated the number as a problem of motivation or willingness to move. I asked for a more careful account of their preferences. Eleven wanted a broader contribution while remaining in store leadership; seven did not want a role change at that point. Those were different answers, and neither could be resolved by repeating the advantages of the district position.

Operations was concerned about creating expensive titles without additional value. Finance asked how a new store role would differ from work already expected of a strong manager. The managers themselves did not want an informal mentoring burden added without clarity about time or authority. I worked with the chief operating officer on a limited advanced-store mandate involving support for nearby managers, defined participation in appointment reviews and responsibility for a small number of practical development assignments.

We approved six pilot appointments with USD 96,000 annual pay effect. Selection considered the actual mandate and the evidence available, rather than automatically appointing everyone who had declined the district route. The other five colleagues interested in a wider store contribution did not receive a pilot role at that stage. We explained the limited scope and avoided presenting the arrangement as an entitlement to future promotion. District leaders retained their responsibilities; the advanced-store role was not an unofficial layer of district management.

The result was six funded appointments with a defined purpose. It did not establish that the pilot improved retention or that every store needed the same role. We reviewed the work and the effect on the appointees’ own stores before considering any extension. I would involve the receiving district leaders earlier in another trial, because they needed a shared understanding of where advice ended and decision authority remained. The wider lesson was about the question we had asked. A manager could be willing to contribute more while declining the particular career change we had offered. The company still needed an economic and operating case for an alternative, but it first needed to hear the answer accurately.