Ending a development candidate can feel like a judgement on the people who worked hardest to understand it. By the time a difficult review arrives, the programme may carry a budget history, a commercial expectation and a story about what it will contribute to the company. These commitments are real, but they are not all evidence for the next technical step. I try to make that distinction explicit. The work already undertaken deserves an accurate account; the proposed continuation still needs a reason grounded in what we now know.
This is a demanding position for a technical leader because continuing often appears more supportive. Another experiment offers hope and keeps a timetable alive. A stop recommendation makes the disappointment immediate. Yet support for a team should include protecting its effort from an indefinite obligation to rescue an increasingly weak proposition. The question is not whether the colleagues were capable or committed. It is whether the next programme has a credible purpose, whether its result could change the decision and whether that purpose justifies the additional resources requested.
Separate the work done from the reason to continue
In May 2026, we reviewed a formulation candidate after CNY 6.8 million of development expenditure. Its planned stability work had not supported the route needed for the next commitment. The sponsor proposed three further months at CNY 2.1 million. I wanted to understand what the new work would examine that the previous programme had not established. Repetition may sometimes have a justified investigative purpose, but repeating an approach because the earlier result is unwelcome is not, by itself, a technical explanation for spending more.
The review involved formulation, analytical and quality colleagues within their responsibilities. I did not treat a CTO recommendation as a substitute for those judgements. Their account informed my conclusion that the proposed continuation lacked a sufficiently grounded new hypothesis. We recommended stopping that development candidate. The decision did not establish that no future scientific route could ever work. It established that this proposed next commitment was not justified by the evidence and rationale presented to the executive review at that time.
The financial explanation needed equal care. The CNY 6.8 million had been spent on work performed. It could not be recovered by declining the next proposal. The unapproved CNY 2.1 million was an estimate of future work, not a cash saving already realised. Six specialists became available after completing programme closure responsibilities, but they remained employees and were reassigned. These distinctions prevented the stop decision from becoming an exaggerated efficiency story. Its value lay in making a responsible next allocation, not in inventing a financial benefit from every number in the paper.
There is a strong argument for patience in development. Difficult candidates can require repeated learning, and premature stopping can discard a valuable possibility. I take that argument seriously. The answer is not a general preference for ending programmes; it is a clear account of what another stage would test and how its result would be used. A plausible new explanation may justify work even after disappointing results. A familiar promise expressed with greater confidence does not become a new explanation simply because the organisation wants the opportunity to remain available.
Keep the decision from hiding inside the forecast
Commercial planning can make a technical decision feel predetermined. A candidate enters an internal scenario as a possibility, then appears in a later discussion as an expected contribution. Over time, removing it looks like a business loss that the development team must prevent. I want the assumptions behind that transition to remain visible. A planning scenario is useful when its conditions are understood. It becomes misleading when a conditional technical route is treated as a commitment while the evidence needed to support it remains unfinished.
When we stopped the candidate, the commercial scenario was revised to remove its assumed contribution. That was part of completing the decision. A stop that exists only in the laboratory while its promise remains in the business plan creates contradictory accounts of the same programme. Management then risks allocating other resources around an expectation the technical team no longer supports. The difficult conversation should include the forecast owner, so the implications travel through the organisation instead of ending at the function that made the recommendation.
This also affects how a successor is described. A different formulation concept considered in July required its own rationale and evidence plan. It did not inherit the earlier candidate's expected timing or probability of success. The same commercial need might motivate both concepts, but that common need did not make them equivalent technical propositions. I am particularly attentive when a new project name allows the old forecast to survive almost unchanged. Renaming can make a portfolio appear to have learnt while leaving its central unsupported assumption intact.
The board's useful question is broader than whether management was optimistic. Directors can ask how conditional development opportunities enter the investment and planning account, which decisions could change them and how those changes are communicated. That is a governance discussion about the integrity of the company's choices. It does not require directors to determine laboratory methods or substitute their scientific interpretation for authorised specialists. My contribution as an executive is to make the connection between technical evidence and commercial assumptions clear enough for informed challenge.
Close the route without closing the inquiry
Stopping does not mean discarding the work. Original observations, unsuccessful approaches and the reasoning behind the conclusion may help a later programme ask a better question. Programme closure therefore needs attention and resources. Material, documentation and continuing responsibilities do not disappear when the next budget is declined. In our case, the specialists completed that work before reassignment. Counting them as immediately available would have understated the obligation remaining from the decision and made the next programme's capacity plan less credible.
The record should also preserve uncertainty honestly. A conclusion that a particular continuation is not justified should not be rewritten as proof that an entire area of inquiry is impossible. Future conditions, knowledge or methods may support a different proposition. Equally, keeping that possibility open should not be used to avoid ending the current route. A company can retain its learning while closing a specific commitment. The distinction allows intellectual humility without leaving every project permanently active in case another reason to continue might someday appear.
I would improve our next stage review by agreeing continuation questions earlier. Before substantial effort accumulates, the team should understand what the stage is intended to establish and what kinds of outcome would require reconsidering the route. Those questions may change as knowledge develops, but any change should be explained. Earlier clarity will not remove disagreement or disappointment. It can make the eventual conversation more about the evidence and less about discovering, late in the programme, that different colleagues had been working toward different meanings of success.
The project can end while the learning remains useful and the people remain valued. That is the standard I want a development organisation to meet. It requires financial accuracy, a disciplined relationship with forecasts and respect for the specialists whose work informs the choice. A responsible portfolio contains opportunities we pursue, uncertainties we examine and routes we deliberately close. Its quality is not measured by how rarely we stop, but by whether the next commitment has a reason that survives careful examination.