The committee that declared everything a priority
In this illustrative 2020 scenario, Élodie leads an IT portfolio committed before the crisis. Sales will not move its CRM, Finance wants invoicing improvements, and the factory wants its ERP wave. Security requests an urgent fix. The committee declares every project a priority. Karim, the delivery lead, knows the same specialists are expected in simultaneous meetings. The missing element is not a scoring method; it is a decision that actually removes work.
Reprioritisation compares commitments with feasible capacity and decides what continues, narrows, pauses or stops. Money may not be the bottleneck: architects, business owners and acceptance teams can constrain the entire portfolio. Written in 2026, this retrospective addresses the 2020 context. Names, workloads and costs are teaching examples. Current references are clearly identified and do not retrospectively create new obligations.
Count capacity by skill and delivery window
Élodie examines the next four weeks because an invoicing deadline and rollout decision fall within that window. Karim subtracts absences, unavoidable support, on-call duties and work that cannot safely stop. He separates functional, application and data capacity. A free data specialist cannot replace the business owner required to approve acceptance. Suppliers provide actual availability rather than contractual staffing assumptions.
Demand means remaining work, with ranges where preparation is uncertain. Effort already spent is not future capacity. Project leads identify indivisible activities, dependencies and the minimum work needed to preserve a usable asset. Preparation and deployment are separated: configuration can be preserved without launching a rollout. This gives sponsors meaningful alternatives to complete continuation or cancellation.
A matrix that leads to a choice
| Four-week commitment | Functional / application / data | Illustrative decision |
|---|---|---|
| Available capacity after operations and absences. | 24 / 40 / 16 days. | This is the planning ceiling. A larger budget does not create an available skill. |
| Security fix for the exposed scope. | 8 / 8 / 0 days. | Retain necessary scope; Security confirms urgency and operations prepares the change window. |
| Stabilise invoicing. | 12 / 12 / 4 days. | Retain with targeted acceptance and reconciliation; Finance supplies the authorised business owner. |
| Make critical remote access reliable. | 4 / 8 / 0 days. | Retain the critical population; defer convenience requests and confirm support and rollback. |
| ERP rollout to the next country. | 10 / 18 / 10 days. | Pause deployment; preserve prepared configuration and data, and renegotiate dependent commitments. |
| CRM extension and new reporting dashboard. | CRM: 8 / 10 / 4 days; dashboard: 4 / 6 / 8 days. | Pause both extensions. Spare data capacity cannot compensate for unavailable functional approval. |
In practice: genuinely pause the ERP wave
The three retained activities consume 24 functional, 28 application and 4 data days. Two skills appear to have spare capacity, but no functional capacity remains to accept another wave. Élodie therefore refuses to restart ERP merely because developers could begin. She authorises only preservation of prepared assets within a separately checked allowance. Preservation must not become silent expansion of the paused scope.
The sponsor records no country deployment, a frozen accepted configuration, an inventory of prepared data, committed licences, a supplier contact and restart conditions. Unnecessary workshops are cancelled. Project leads release calendar slots and stop asking the functional team for approvals. Continuing the same meetings, tickets and validations means the portfolio changed colour rather than priority. Business teams, Procurement and suppliers receive the same decision.
Put obligations and dependencies before scores
A value score cannot compensate for a missing prerequisite. Élodie first identifies interruptions creating unacceptable exposure or preventing essential service. Qualified functions confirm obligations and deadlines; commercial preferences must not be relabelled as regulatory requirements. She then identifies enabling activities protecting several flows. Discretionary improvements are compared using remaining benefit, required capacity and reversibility.
A score supports discussion rather than making the decision. Equally scored projects may need different skills or incur very different restart costs. A nearly completed project may still depend on an unavailable operating organisation. Past expenditure is documented but does not independently justify more spending. The question is what further money and effort would achieve or put at risk.
Price the four alternatives consistently
Continuing funds completion and service transition. Narrowing needs a smaller useful outcome with its own interfaces and acceptance criteria. Pausing incurs preservation and restart costs. Stopping can involve exit costs, contractual commitments and asset recovery. Élodie requests all four perspectives for CRM, not just its initial investment figure. Procurement verifies suspension and termination conditions with competent stakeholders rather than inferring them from a budget line.
The register distinguishes certain cost, estimates and unknowns. Licence savings require an actual contractual reduction mechanism and date. Released effort is capacity, not automatically an accounting saving. Postponing ERP protects the functional team without mechanically reducing payroll. Reallocating that capacity to invoicing creates delivery and risk benefits, not invented savings.
Define restart conditions instead of a hopeful date
A paused project does not restart because a month has elapsed. Élodie requires confirmed functional capacity, valid preparation data and supplier availability in an accepted window. The business owner revalidates need if commercial conditions changed. The project lead estimates refresh work before reusing the old schedule. Ignoring these differences causes immediate delay and encourages teams to blame every later problem on the pause.
The portfolio decision log records the chosen alternative, preserved scope, decision maker, next condition and holding cost. Capacity changes trigger focused review. A three-day absence need not reopen the whole portfolio; review commitments using the unavailable skill. Limiting started work prevents supposedly exceptional projects from gradually reversing the decision.
Sector-specific priorities
In manufacturing and pharma, production windows, Quality availability and previously qualified changes can make postponement more complex than shifting dates. Protect continuity and traceability work; establish whether preparation requires revalidation before deployment. Industrial management participates in assessing the real cost of delay.
In banking and insurance, distinguish essential service, confirmed obligations and commercial improvements. Dependencies across data, controls and suppliers can make small enabling tasks more important than a visible programme. In services and public administration, contact capacity and backlog reduction may take priority over a new feature. Affected populations and alternative channels influence decisions; an average benefit must not conceal serious deterioration for one user group.
Evidence that the committee actually decided
The following week, Karim should see less simultaneous work, business owners available at agreed checkpoints and suppliers following revised commitments. Élodie checks execution against the adopted matrix. If a paused project still consumes the same specialists, the sponsor corrects its mandate or explicitly accepts a new decision. A governable portfolio makes no visible in calendars, purchase commitments and work queues, not merely in a shorter presentation.
Sources and method
Primary sources checked on 4 October 2026. The year identifies the period being examined; this retrospective was written in 2026. Later documents provide present-day comparisons, not knowledge attributed to that period. People, scenarios and numerical examples are illustrative teaching material, not results of a CYTIZEN engagement.