Outcome statements
Describe the affected decision, current friction, target condition, and accountable sponsor.
Signed outcome and baseline registerEnterprise AI · Strategy
Turn an AI ambition into a sequenced portfolio of accountable decisions.
Strategy work connects operating priorities to feasible workflows, data obligations, technical dependencies, and adoption capacity. It produces choices and investment gates rather than a list of technology themes.
- Input
- Operating goals, workflow evidence, constraints, and change capacity
- Decision
- Fund, investigate, sequence, defer, or stop
- Output
- Portfolio thesis, gated roadmap, and measurement plan
02
01 · Direction
A strategy should state which operating constraint will change and how anyone will know.
We translate broad aspirations into measurable workflow outcomes, while recording constraints such as regulation, data access, integration lead time, and tolerance for automated action.
Constraint register
Capture policy, data, architecture, procurement, and workforce constraints early.
Owned constraints with review datesNon-goals
State what the portfolio will not automate or promise during the planning horizon.
Approved exclusions and rationale03
02 · Prioritisation
Value, feasibility, consequence, and learning value belong in the same comparison.
Candidate workflows are assessed with the same questions. High-value ideas with weak evidence become discovery work; high-consequence ideas require stronger controls and proof before scale.
Screen
Remove ideas without a decision owner, usable evidence, or credible adoption path.
Triage decisions with reasonsInvestigate
Time-box uncertainty around data, integration, evaluation, and user behavior.
Discovery findings and unresolved risksSequence
Order initiatives by dependencies, reusable foundations, and organisational capacity.
Dependency-aware portfolio roadmap04
03 · Governance
Each stage earns the right to consume more money, data, and organisational attention.
Investment gates define the evidence needed to move from discovery to bounded pilot, operational trial, and production service. A stop decision is treated as a valid portfolio outcome.
Discovery gate
Confirm a credible workflow, source boundary, sponsor, and evaluation approach.
Problem brief and feasibility recordTrial gate
Demonstrate representative quality, user fit, safe failure, and an operable architecture.
Trial report and exception logService gate
Confirm ownership, support, security review, cost envelope, and controlled release.
Readiness decision and runbook05
04 · Decision cadence
Portfolio reviews should expose trade-offs, not produce status theatre.
A common review surface links each initiative to its evidence, unresolved assumptions, dependencies, spend envelope, and next decision. Sponsors can redirect investment before delivery momentum hides weak foundations.
Illustrative portfolio review: initiatives are compared through common evidence, not enthusiasm.
Comparable evidence
Initiatives use shared definitions while retaining workflow-specific measures.
Visible assumptions
Unknowns are dated, owned, and tied to the next investment decision.
Dependency pressure
Shared data, identity, and operating foundations are planned as portfolio work.
06
05 · Mobilisation
The roadmap becomes real when teams know the next evidence they must produce.
We leave a sequenced backlog, decision calendar, ownership map, foundation plan, and templates for future workflow briefs. Strategy remains revisable as evidence changes.
90-day decisions
Near-term work is expressed as decisions to unlock, not large undifferentiated programmes.
Foundation backlog
Cross-cutting identity, data, evaluation, and operations work has its own owners.
Review cadence
Sponsors revisit evidence, constraints, and stop conditions on an agreed schedule.
The next step is a bounded working session: one workflow, its evidence sources, the decision owner, and the conditions under which the system must stop or hand over.
Run a strategy working session