Two questions inside one margin
A project manager once brought me a revised completion forecast and an unsigned change request in the same spreadsheet. The estimate of the remaining work had risen. The proposed fee increase almost restored the original margin. Looked at together, the project appeared to have suffered only a small setback. Looked at separately, it had a substantial delivery problem and a possible commercial response. Those were different facts, supported by different evidence and controlled by different people. My first responsibility was to keep them distinguishable long enough for management to make a useful decision.
The implementation contract carried EUR2.8 million in agreed fees. Its original direct delivery-cost estimate was EUR2.14 million, implying EUR660,000 of direct contribution before central overhead. After customer testing exposed work that the plan had understated, the revised cost was EUR2.59 million. Against the existing fee, the forecast contribution therefore fell to EUR210,000. These were completion forecasts, not a statement that the whole contribution had already been recognised in the accounts or received in cash. The distinction mattered because several service and acceptance stages remained open.
Commercial colleagues had good reasons to believe that part of the additional work should be paid for. They could point to customer requests and decisions that had changed the implementation. Delivery colleagues could also identify effort that reflected our own planning weaknesses. Neither explanation justified postponing the cost estimate until the fee discussion was resolved. Work could consume people and money while the negotiation continued. A forecast that required the negotiation to succeed before it became believable would conceal the decision management actually faced.
There is a human difficulty here. The sales team may hear a lower margin as a judgement that the account was sold badly. The delivery team may fear that a candid estimate will be used against it before responsibility for the changes is understood. Finance can make those reactions worse by presenting the new number as a verdict. I prefer to begin with the work: what remains, who has estimated it, what assumptions changed and which parts are still uncertain. Attribution is necessary, but it should not obstruct the first honest view of completion.
The estimate needs an operating owner
Finance did not invent the revised effort requirement. Delivery leaders broke the remaining implementation into defined activities and named the people responsible for estimating them. Some activities depended on a customer decision; others depended on our own rework. We retained those dependencies instead of converting every uncertainty into one percentage contingency. Finance checked the arithmetic, cost rates and relationship to recorded spending. The operating team remained responsible for explaining why the work was necessary and how it would be completed. That separation made challenge possible without pretending that an accountant could replace the delivery judgement.
I asked the team to show the movement from the original estimate, rather than merely supplying a new total. A director reviewing the project could then distinguish completed work that had cost more than planned from work still to be done. The distinction affected the choices available. Money already spent could not be recovered through a more optimistic schedule. Remaining effort could sometimes be reduced through a narrower agreed scope, a different sequence or a clearer customer decision. Each alternative required an operational explanation, not just a lower cell in the forecast.
The commercial team wanted the expected change order included because excluding it seemed excessively pessimistic. I understood the concern but proposed two views. The first used the fee already agreed and the latest defensible completion cost. The second showed the commercial proposal, its supporting evidence and the additional work that accepting it would entail. Neither view was a substitute for the applicable accounting assessment. They were management views intended to reveal the choices. The board received the difference openly, with the commercial outcome still described as unresolved at that point.
The eventual agreement added EUR380,000 in fees and approximately EUR100,000 in direct delivery cost. The revised contract value became EUR3.18 million and the forecast cost EUR2.69 million, producing EUR490,000 of forecast direct contribution before central overhead. That was better than the position under the unchanged contract, but it did not restore the original EUR660,000. We kept the shortfall visible. Calling the whole negotiation a recovery of margin would have overlooked the extra work accepted in return and the deterioration that remained after the agreement.
Recovery is a separate decision
The case did not establish that every estimate should exclude every possible commercial change. Different contractual evidence can support different treatment, and specialist accounting advice may be necessary. My narrower point is that the operating estimate should not be adjusted to make an uncertain commercial outcome seem inevitable. Management needs to see which part of the forecast rests on work assessment and which rests on a customer's future agreement. When those assumptions are blended, a missed negotiation can look like an unexpected operating loss even though the underlying effort was known earlier.
A useful review also asks what happens if agreement takes longer. People may remain assigned to the project, invoices may await a milestone and other customers may lose access to scarce skills. None of those consequences is captured by the final contract value alone. In this case, we discussed the staffing sequence while the scope conversation continued. We did not describe potential future customer work as compensation for a current shortfall. A relationship can be commercially valuable without making every present concession financially sound. That judgement belongs in an explicit discussion with the account owner.
I would now establish the review earlier, at the first material change in the remaining effort. The initial debate took too long because the team had treated commercial resolution as the prerequisite for financial clarity. A short account of the current contract, revised work and possible agreement would have made the tradeoff available sooner. It would also have reduced the pressure on project managers to defend an outdated margin until somebody else approved a new one. The point is to make revision a normal part of responsible management, without making every revision painless.
For a board reader, the most useful question is not simply whether the project is back on plan. It is which assumptions now support completion and who is accountable for the next decision. I want directors to see the cost already incurred, the work still required, the commercial terms actually agreed and the uncertainties that remain. That account is less elegant than a single restored percentage, but it gives the executive team a firmer basis for allocating people and negotiating with the customer. A services margin earns credibility when it survives an honest account of delivery.