A technology proposal can be financially affordable and still be impossible to execute responsibly in its proposed period. That statement often meets an understandable objection: if the benefit is worthwhile, why not add people? Sometimes that is exactly the right response. In a medicines manufacturer, however, a programme may depend on experienced colleagues whose contribution includes knowledge of the process and authority to assess the proposed change. More implementation capacity does not automatically supply either. A portfolio that ignores this distinction can look fully resourced while several projects quietly depend on the same unavailable people.
I learnt to ask about that dependence before taking a request to the investment discussion. The question is not whether quality, laboratory or manufacturing colleagues support the idea in principle. It is whether named responsibilities have been understood, time has been protected and the organisation can perform the necessary work alongside its continuing duties. A sponsor's enthusiasm is valuable, but it does not create hours in another team's calendar. Nor should a technology leader interpret attendance at a workshop as acceptance of every later review obligation implied by the programme.
Find the decisions behind the schedule
A project plan often lists a review as a short activity near a milestone. I ask what decision that activity is meant to support. The answer may reveal several different tasks: explaining an existing process, agreeing the intended use, examining test evidence, considering exceptions and deciding whether a next step is appropriate. Those tasks need preparation and sometimes iteration. If they are compressed into one unexplained line, the plan gives a misleading impression of both the effort and the authority involved. The review then becomes a date the team hopes to reach rather than work the organisation has committed to perform.
In preparing our 2026 programme, I allocated EUR24 million of change authority across four purposes: EUR7 million for instrument-record stewardship, EUR9 million for enterprise renewal, EUR5 million for access and continuity, and EUR3 million contingency. A separate EUR4.2 million analytics proposal offered a potentially useful way to examine manufacturing information. Its financial case deserved attention. Its schedule also depended on specialists already engaged in the approved programme and their continuing roles. The missing resource was not simply a developer or an additional project coordinator; it was experienced participation in decisions that could not be transferred without examination.
We considered whether a smaller first stage would solve the problem. That was a serious alternative, not a rhetorical step on the way to rejection. A narrower technical build might have reduced expenditure and delivery effort, but the proposed stage still required the same specialists during the same period. It therefore left the main constraint intact. I recommended deferral with a clearer capacity plan for reconsideration. I did not argue that analytics was unnecessary, and I did not record the unapproved EUR4.2 million request as a saving achieved by my team. Neither claim would have described the decision accurately.
Protect the work that cannot be delegated casually
There is a difference between work that can be supported externally and responsibility that must remain understood internally. An experienced supplier can contribute substantial expertise, testing and documentation. That does not mean the manufacturer can stop knowing why a configuration is acceptable for its own use. The right external contribution depends on the task, the competence available and the oversight arrangement. I resist both extremes: assuming every review must be performed by the same few internal people, or assuming that purchasing capacity resolves every question about ownership and judgement.
The practical response is to separate tasks and examine them. Some preparatory work may be moved, shared or scheduled earlier. Some knowledge can be developed through planned learning and supervised participation. Other decisions may still require particular experienced colleagues. That analysis can increase capacity over time, but it should not be represented as already achieved merely because a proposal includes a training line. Building capability has its own effort, timetable and evidence. A board or executive committee should be able to distinguish current capacity from a future capability on which a programme hopes to rely.
Continuing operations also deserve an explicit place in this discussion. A specialist does not become available for a project because the project has a senior sponsor. The person's ordinary work may include investigations, reviews and support that cannot all be predicted at the start of a quarter. A credible plan leaves room for that reality and states how priorities will be reconsidered when demands change. I do not expect perfect foresight. I do expect the project to acknowledge the obligations that its schedule would otherwise make invisible, and to identify who can resolve a conflict when it appears.
Contingency is relevant but not magical. Our EUR3 million portfolio provision could address some uncertainty within the agreed programme. It did not automatically authorise every new idea or create specialist attention on demand. I find it useful to ask what kind of uncertainty a contingency is intended to cover. A financial reserve can support additional work where capable people are available. It may not solve a timing conflict involving a small group of experienced reviewers. Treating all uncertainty as a money problem leads to investment reports that are precise about euros and vague about execution.
Make deferral an examinable choice
Deferral should have an account as clear as approval. I state what is being postponed, why the proposed period is not credible, which alternatives were considered and what would allow reconsideration. I also state the opportunity cost. A useful capability may arrive later; the existing way of working may remain inconvenient. These are real consequences, even when they are difficult to value precisely. A decision should not gain apparent certainty by omitting them. Equally, a sponsor should not be required to invent a dramatic loss simply to have the cost of waiting heard.
For the analytics proposal, the next useful step was a capacity requirement connected to the actual decisions, not another version of the same financial presentation. The sponsors could then discuss sequencing with the operating and quality leaders whose teams were involved. That work preserved a route back to the proposal. It also made clear that my recommendation was about the executable plan before us, rather than a permanent view of the technology. I want teams to understand that a well-supported later proposal is more valuable than an immediate approval sustained by assumptions nobody has accepted.
The broader lesson is that an investment committee should see the organisation it is asking to deliver. Money, engineering effort, operating participation and decision authority interact, but they are not interchangeable. A CIO can help by making those relationships visible before a commitment becomes public or politically difficult to change. I judge the quality of that contribution by whether colleagues can challenge the schedule and still recognise the value of the idea. A credible portfolio leaves room for worthwhile work to wait until the enterprise can actually perform it.