← Focus Areas — Data & AI

The Data & AI Opportunity Matrix

Stage 1 of 3 — Discover

Before you can evaluate whether a specific AI use case is worth building, you need a shortlist of candidates worth evaluating. This matrix is how you get there: a structured way to surface where data and AI can create value in your sector, before committing to any one initiative.

1. Discover
This matrix
Surface and prioritise candidate initiatives
2. Evaluate
Feasibility Evaluation
Score a specific named use case in depth
3. Enable
Portfolio Enablement
Sequence and resource multiple evaluated use cases

The framework

The matrix crosses two axes.

Five strategic focus areas (the columns) — what kind of value you're trying to create:

  • Experiencehow the service feels to the people using it
  • Communityconnection between the people the service touches
  • Sustainabilityenvironmental and long-term operational responsibility
  • Flexibilityhow well the service adapts to changing needs
  • Resiliencecontinuity when things go wrong

Five business activities (the rows) — what kind of business outcome you're targeting: Cost Efficiency, Value Generation, Customer Satisfaction, Risk Mitigation, Revenue Growth.

Each of the 25 intersections is a prompt: what would a data- or AI-driven initiative look like here, for this sector? Not every cell needs a strong answer — the value is in seeing which intersections are rich with opportunity and which are genuinely thin for your sector, not in forcing 25 initiatives.

How to use it

  1. Populate the grid for your sector. Work across each row, one strategic focus area at a time. Aim for one plausible initiative per cell where it’s genuinely applicable — leave weak cells blank rather than padding them.
  2. Size each initiative. For each one, estimate whether it’s expanding into new markets (Market Development), improving what you already do (Market Penetration), building genuinely new capability (Diversification), or enhancing an existing offering (Service/Product Development). This tells you what kind of investment and risk profile you’re looking at.
  3. Plot for prioritisation. A simple bubble view — size by expected value, position by implementation cost vs. complexity — makes it fast to see which initiatives are worth carrying forward.
  4. Carry your top candidates forward. Whichever 2–4 initiatives come out with the best value-to-complexity ratio go into the Use Case Canvas in the Feasibility Evaluation, one at a time.

Worked example: commercial real estate & property management

To show the framework in action rather than leave it abstract — a partial grid for a commercial real estate operator (illustrative, not exhaustive):

Cost EfficiencyCustomer SatisfactionRevenue Growth
ExperienceSmart building systems that tune HVAC and lighting to real occupancy, cutting energy spend without a manual scheduleA tenant portal that predicts and pre-empts maintenance requests before tenants notice the problemData-driven amenity personalisation that lets premium space command higher rents
FlexibilityDynamic space-utilisation data that flags underused floors before a lease renewal, avoiding over-committing to unused square footageOn-demand meeting-room and shared-space booking that adjusts to actual demand patternsFlexible lease structures priced using real utilisation data instead of flat-rate assumptions
ResiliencePredictive maintenance on critical building systems (lifts, HVAC) to avoid emergency call-out costsClear, data-backed communication during service disruptions, maintained through a single source of truthReliability as a sellable differentiator for tenants who need guaranteed uptime (data centres, healthcare tenants)

Note what's not here: Community and Sustainability columns, and Value Generation and Risk Mitigation rows, are left out of this partial example — in a full exercise you'd work through all 25 cells, but a sparse grid with genuinely strong ideas beats a full one padded with weak ones.

Adapting to other sectors

The five-by-five structure doesn't change. What changes is the texture of each cell — a healthcare provider's Experience row looks like patient communication and appointment prediction; a manufacturer's looks like operator-facing shop-floor tools. Use the sector shortlist on the Data models & ontologies page as a starting point for which standards and systems a given sector's initiatives would actually need to integrate with.

Once you've shortlisted a candidate, evaluate it in depth here.

← Back to Data & AI