The decision includes what the company accepted, spent and gave up.

In 2020, two Valserenne Logistics depots reviewed the working arrangements of 286 drivers. Managers shared a weekly rota, but late changes were common enough to weaken its usefulness. In one eight-week period, 114 changes were communicated after the original arrangement had been shared. Operations colleagues explained the variability in freight demand. Drivers described the practical cost of being asked repeatedly to change plans, even when each individual request appeared reasonable.

I asked the review group to distinguish exceptional disruption from recurring demand that the staffing model had not funded. The proposal combined twelve relief positions, at EUR 740,000 annual employment cost, with a clearer route for approving and explaining late changes. Depot leaders worried about paying for cover that would not always be fully used. Finance wanted evidence that the additional provision addressed an operating requirement rather than an unspecified expectation of improved morale.

We examined the pattern of changes with operations and employee representatives. The executive team accepted the cost as part of the service arrangement. Managers retained responsibility for the operational plan and applicable transport requirements. My team worked on the employment terms, communication and review of exceptions. We did not tell employees that the published rota could never change, or treat the new process as a replacement for the relevant local agreements and professional advice.

The next eight-week period recorded forty-seven late changes. Freight conditions differed, so I do not attribute the entire reduction to the intervention. We reviewed the remaining changes individually to understand whether the relief provision was reaching the right part of the operation. Some arrangements still needed adjustment. The decision established that predictability had a cost management was willing to fund. I would involve the dispatch teams earlier in a repeat exercise, because they could identify where the proposed cover would be difficult to use. A central relief total is only useful when it matches the work and location in which the uncertainty occurs.

A 2022 proposal for an overnight distribution service included EUR 4.8 million in estimated annual labour cost across eight depots. The commercial team had worked hard to make the offer competitive. Before the final submission, operations raised questions about relief, supervisory coverage and the working patterns assumed in the model. The visible hours had been costed, but several practical requirements were either understated or treated as something existing teams would absorb.

As CHRO, I brought the people, operating and finance colleagues together to examine the proposed employment arrangement. We looked at the work required, the availability of qualified staff and the assumptions about cover. Local advisers reviewed relevant employment questions; transport managers retained responsibility for operational and transport compliance. My contribution was to make those dependencies part of the bid decision, not to certify the service plan or set the selling price.

The revised provision was EUR 6.1 million annually. Commercial leaders argued that passing the difference into the offer could lose the opportunity and that some uncertainty might be managed after launch. I asked which part of the revised requirement they believed could actually be removed, and what service change would make that possible. The executive team chose to reprice rather than accept the original employment assumptions. The customer did not accept the revised offer, and the company lost the bid.

That outcome belongs in the account. We did not produce a revenue win, and I cannot prove what would have happened had the original price been accepted. We did establish that the people provision described a service management was prepared to fund. The missed opportunity was the cost of that decision. I would bring the joint review earlier next time, before the original estimate became the reference point for a long commercial effort. A late challenge can look like a functional veto when the real issue is that an employment assumption should have been examined as part of the offer from the beginning.

In 2010, while I led French labour planning, a recurring group of pay queries pointed to differences between depot records and the variable inputs reaching payroll. The central team could explain the calculation it had performed. Employees and supervisors sometimes had a different record of the working arrangement on which the input should have been based. Closing a query with a formula explanation did not resolve that disagreement.

I sponsored a joint review covering 1,400 employees, with payroll specialists, depot managers and employee representatives. We traced selected discrepancies back to the original records and then extended the reconciliation across the affected input categories. The work identified EUR 184,000 of corrections. Qualified payroll and employment colleagues reviewed the basis for payment; I did not treat a management preference as a substitute for the applicable entitlement or advice.

Operations managers were concerned about the time required to reconstruct older records. The payroll team wanted a stable input format, while employees wanted an answer they could check. We agreed a clearer explanation of variable entries and a named route for resolving differences before the next cycle. The company accepted both the corrections and the administrative effort. We avoided presenting the exercise as a voluntary benefit, because the purpose was to establish the amount that should have been paid under the relevant arrangements.

Queries fell from 164 to forty-three across two later payroll cycles. That was a service measure, separate from the correction amount and not proof that all employment concerns had disappeared. We retained a route for further challenge and reviewed whether fewer queries reflected better explanations rather than a loss of confidence in asking. I would have involved depot supervisors in designing the input explanation sooner. A form that is clear to payroll can still be ambiguous to the person recording a late change at the end of a demanding shift. The durable result was a more traceable connection between the work agreed and the pay record, rather than simply a smaller queue of unanswered questions.

A 2015 integration proposal affected 114 roles across two French depots. The business case assumed that moving work into a revised network would simplify supervision and reduce duplicated facilities. For employees, the implications varied. Some could move to the receiving location, some could continue in a redesigned arrangement and others faced a change that did not fit their circumstances. A single group announcement could not answer those different questions.

As group employee-relations director, I worked with the operating sponsor, local advisers and employee representatives to make the proposal and alternatives clear. We kept the consultation connected to executives who could revise the operating plan. The sponsor wanted a firm completion date; employees needed to understand the work, travel and employment implications before responding. I recommended separating the decision about the network from the individual arrangements required to implement it.

The final plan involved forty-two transfers, fifty-eight colleagues remaining in revised local roles and fourteen agreed departures. These categories account for the 114 affected roles in the exercise; they are not a claim that every person preferred the result. The company approved EUR 980,000 for the employment transition, including agreed support and departure costs. Relevant terms and processes were reviewed locally. We did not use group agreement on the operating direction to imply that each individual circumstance had already been settled.

Implementation took three months longer than the original planning assumption. The business accepted that delay while the individual arrangements were resolved. I would have explained the distinction between the consultation timetable and the operational-change timetable more carefully at the outset. Their early combination created avoidable pressure on managers to describe a provisional decision as final. The experience reinforced a practical lesson: a distributed organisation needs a way to carry individual consequences back into a central plan. Otherwise consultation can become a communication exercise around a decision the business no longer believes it is able to change.

In 2024, I approved a pilot intended to simplify shift-swap requests for 312 warehouse employees. The proposed process used a personal-phone application as the required entry point. We expected fewer lost requests and a clearer record for supervisors. During the pilot, forty-seven colleagues could not or would not use that route. Reasons varied, including device access, confidence with the application and a preference not to use a personal phone for employment administration.

My team initially proposed additional demonstrations. Employee discussions showed that training would address only part of the problem. We had made access to a routine employment process conditional on a personal-device arrangement the company had not properly examined. Supervisors also reported that some requests continued informally, leaving the supposedly complete digital record incomplete. I accepted that the pilot design, rather than employee reluctance alone, needed to change.

We withdrew the phone-only requirement and retained three routes: the application for those choosing it, a workplace terminal and a recorded supervisor-assisted request. The EUR 96,000 already spent on the pilot remained in the project account. A further EUR 54,000 funded the alternative access and revised administration. The technology team was concerned about supporting parallel routes; operations wanted one record. We therefore required each route to enter the same request log, with a visible status and a way for the employee to check it.

The result was access through more than one channel, not proof of a universal improvement in employee satisfaction. We reviewed whether requests from different routes received comparable handling and kept the pilot limited while that evidence developed. I would ask about access conditions before selecting the interface in a future project. A process is not simpler merely because it is convenient for the team administering it. My responsibility was to acknowledge the assumption I had approved and ensure that a colleague could make an ordinary employment request without having to adopt a personal technology arrangement as the price of being heard.