EPF IN PRACTICE
How EPF works in your sector.
Three sectors. Three different operating realities. The same structured system — from initial diagnostic through to measurable, sustained performance improvement.
SECTOR USE CASES
Enterprise decison management is sector-shaped.
The governance constraints, economic drivers, and decision structures differ remarkably. EPF adapts to each — with a consistent approach but completely tailored, client-specific engagement.
Portfolio value creation and exit readiness
PE-backed businesses operate under compressed timescales with direct sponsor accountability. EPF diagnoses the portfolio company's cost base, operating model, and governance weaknesses, then designs the value creation plan the investment thesis depends on.
Typical products: EPF-Fusion / EPF-Core
Regulated transformation and cost accountability
FCA/PRA-regulated firms carry governance obligations that constrain every cost decision, operating model change, and commercial structure. EPF overlays regulatory requirements onto the transformation lifecycle.
Typical products: EPF-Core / EPF-Cost IQ / EPF-Halo
Investor reporting, cost-to-serve, and operational efficiency
Property firms manage fragmented data across lease administration, investor reporting, and property management systems. EPF establishes the cost-to-serve baseline and designs the operating model that makes the numbers trustworthy.
Typical products: EPF-Cost IQ / EPF-Core
AI GOVERNANCE & CONTROL
Why EPF matters for AI at scale.
AI investment decisions carry the same governance risk as any capital deployment. EPF provides the evidence discipline, ownership accountability, and cost controls that separate credible AI strategies from expensive chaos.
The AI programme with no case behind it
Failure: No structure, no blueprint
The Scenario
A bank stands up an AI function under pressure to 'have an AI story.' Six months in, there are eleven pilots running across retail, ops, and risk. Each has a sponsor, a budget line, and a demo. None has a defined case for why it exists, what it changes, who owns the outcome, or how anyone would know it worked. Spend has committed ahead of intent.
How EPF Fixes It
Operating model design and the Five Case discipline. Nothing funds or deploys without a defined case tested across strategic, economic, commercial, financial, and management dimensions.
What Changes
Every AI initiative has a named owner and a measurable outcome before a pound is committed. The eleven pilots become three funded cases and eight that never should have started.
The AI capability that grades its own homework
Failure: AI can't audit its own work
The Scenario
The team that built a fraud-detection model reports on whether it's working. Their dashboard says accuracy is up and false positives are down. Eighteen months later, the true picture surfaces: the model was quietly re-tuned to hit the metric it was measured on, operational cost rose, and the value case was never independently tested.
How EPF Fixes It
Separation of duties, enforced through TBM and ITFM. The whole reason independent cost transparency exists is that the function spending the money can't be the function certifying its value.
What Changes
No AI capability signs off on its own value. Every material claim gets a second, disinterested opinion from people who didn't build it and have no stake in it being right.
The AI spend that drifted off course one decision at a time
Failure: Intention drift
The Scenario
No single decision was wrong. A team in London picks one LLM vendor; a team in Singapore picks another; three business units each buy their own AI-copilot licences; a fourth builds in-house. Every choice was defensible in isolation. Aggregate them across a 27-country estate and you have duplicated tooling, four vendors doing one job, shadow AI nobody governs.
How EPF Fixes It
Demand and allocation governance — the same mechanism that catches duplication and sprawl across any distributed estate. EPF surfaces the drift at portfolio level before it compounds.
What Changes
The four vendors become one. Shadow AI becomes visible and governed. Spend reconnects to intent, and the drift gets caught while it's still cheap to correct.
The AI deployment with no permanent guardrails
Failure: No permanent record of intent
The Scenario
An insurer deploys a customer-facing AI assistant. The original business case set clear boundaries: no use of special-category data, a defined risk appetite, hard regulatory limits, a cost ceiling. Those boundaries were written once, into a document, and never reopened. A year later the tool is doing things nobody signed off.
How EPF Fixes It
Governance guardrails held as standing constraints, not one-time business-case clauses. Data boundaries, risk appetite, regulatory limits, cost ceilings — the musts and must-nots an AI deployment is continuously tested against.
What Changes
The boundaries that mattered at approval still bind a year later. Nothing drifts past its risk, regulatory, or cost envelope without governance seeing it.
The verification everyone assumed was done
Failure: Verification skipped because it feels done
The Scenario
A transformation programme's AI components are 'validated' — the delivery team says so, the sponsor believes it, the steering committee moves on. The material error — a mis-calibrated model quietly producing wrong outputs into a downstream process — surfaces months later, at the point where it's expensive to unwind.
How EPF Fixes It
Continuous cost and value governance — consumption and outcomes tested against intent on a fixed rhythm, not on an ad hoc basis. The review that only happens when someone remembers is the review that never catches the expensive mistake.
What Changes
Verification is scheduled and independent, not assumed. Material errors surface while they're cheap to fix, not after they've propagated.
UNDERSTANDING THE APPROACH
Why sector-specific outcomes matter.
Enterprise performance problems are universal: unclear cost allocation, fragmented governance, transformation failures, and compliance uncertainty. But the consequences — and the solutions — are sector-shaped.
A PE portfolio company needs speed and cash clarity. A regulated financial services firm needs defensible governance and regulatory confidence. A real estate operator needs transparent cost-to-serve and investor reporting accuracy. The EPF system adapts to these realities.
Each use case shows how EPF moves organisations from uncertain symptoms to credible decisions, with the evidence, economics and governance to make those decisions hold.
The Consistent Pattern
- →Diagnostic clarity on root causes, not symptoms
- →Defensible evidence for every recommendation
- →Named accountability for every decision
- →Measurable outcomes that survive scrutiny
PORTFOLIO VALUE CREATION
Private Equity — Portfolio Value Creation
A PE fund acquires a mid-market services business. The investment thesis assumes 15–20% EBITDA improvement through technology cost reduction, operating model simplification, and commercial repricing. Twelve months in, the portfolio company cannot evidence where the value sits, who owns the decisions, or which benefits are cash-releasing versus reallocation.
1. Review — Establish what exists
Build the baseline the investment thesis assumed but the portfolio company never had. Map total technology cost from the GL through cost pools, resource towers, and solutions to consuming business lines. Identify every managed service contract, its P&L, margin structure, and renewal timeline. Establish governance structure — who holds decision rights over technology spend, vendor relationships, and capital allocation. Classify evidence quality: what is documented, what is inferred from interviews, what is assumption.
Methods & Tools
- • TBM Allocation Engine
- • DOAR Builder
Key Outputs
- • DOAR baseline
- • TBM cost model (four views)
- • Managed service contract inventory
- • Governance structure map
2. Assess — Find the root causes
The DOAR diagnostic isolates the structural issues: uncontested shared-cost allocations inflating business-line P&Ls; duplicate application estates from a pre-acquisition bolt-on that was never rationalised; managed service contracts with margin leakage driven by scope creep and volume-insensitive pricing; governance decisions held by committees with no named accountable individual — which is a hard stop under EPF discipline. The Managed Service Contract Analyser builds the contract P&L for each outsourced relationship, exposing unit economics, demand volatility, and renewal leverage. Every finding is logged in the Evidence Ledger with its evidence class.
Methods & Tools
- • DOAR diagnostic
- • Managed Service Analyser
- • Evidence Ledger
Key Outputs
- • Root cause analysis
- • Managed service contract P&Ls with leakage
- • Governance weakness register
3. Design — Model the value and build the case
The Economic Scenario Model quantifies Conservative, Base, and Stretch outcomes for each lever — application rationalisation, managed service renegotiation, demand controls, shared-cost reallocation. Every benefit is classified: cash-releasing (genuine cost-out, margin recovery) enters the model; reallocation (moving cost between business lines) is flagged but never booked as a saving. The Business Case Builder assembles the Five Case for the sponsor board, with the Governance Mapper resolving every decision right to a named individual and the Conflict Checker clearing cross-mode consistency before the recommendation is issued.
Methods & Tools
- • Economic Scenario Model
- • Business Case Builder
- • Governance Mapper
- • Conflict Checker
Key Outputs
- • Three-scenario economic model
- • EPF Five Case (sponsor board-ready)
- • Governance gate status
4. Engage — Align sponsor and management
Adapt the decision case for two audiences: the PE sponsor board (investment thesis validation, EBITDA impact, exit readiness timeline) and the portfolio company management team (operational accountability, execution ownership, governance cadence). The sponsor sees the Conservative case as the committable floor — the only scenario backed entirely by Evidenced-grade claims. Management receives the execution plan with named owners, decision rights, and the governance cadence that will track delivery.
Deliverables
- • Sponsor board pack
- • Management execution plan
- • Named RACI per initiative
5. Integrate — Track benefits to exit
The Benefits Realisation Tracker (Business Case Builder sheets 8–10) runs on a quarterly cadence aligned to sponsor reporting. Period actuals are measured against the committed Conservative case. Each benefit carries a named owner and a realisation gate: ON TRACK, ESCALATE, or WATCH. The TBM model is refreshed each period so the cost base remains current, not a stale baseline from month one. At exit, the value creation narrative is evidenced — not asserted — with a full audit trail from original DOAR finding through to realised cash benefit.
Operating Model
- • Benefits Realisation Tracker
- • Quarterly reporting cadence
- • Refreshed TBM model
Typical Outcomes
12–20%
Cash-releasing cost reduction evidenced and tracked to realisation
100%
Decision rights resolved to named individuals, not committees
Exit-ready
Value creation narrative backed by classified evidence, not assertions
REGULATED TRANSFORMATION
Banking & Insurance — Regulated Transformation
A UK retail bank or specialty insurer faces rising technology run costs, an application estate with significant duplication, and a regulator expecting evidence that cost allocation supports fair value under Consumer Duty. The CIO knows the cost base is too high. The CFO cannot explain cost-to-serve at a product level. The compliance function cannot map SM&CR accountability to the technology decisions that affect customer outcomes.
1. Review — Cost baseline with regulatory overlay
Build the TBM cost model with allocation disclosure — basis, confidence, unallocated percentage, reconciliation — at the standard required for regulatory defence. Activate RCM-02 (Financial Services): cost-to-serve must support regulatory cost-of-service and fair-value evidence; allocation basis must be defensible to a regulator. Map the application estate, identify duplication, and classify technology cost as run, change, or mandatory (regulatory/compliance). Establish the current governance structure, mapping SM&CR Senior Manager functions to the technology and cost decisions they are accountable for.
Methods & Tools
- • TBM Allocation Engine
- • DOAR Builder
- • RCM-02 Financial Services module
Key Outputs
- • Regulator-grade TBM cost model
- • Application portfolio baseline
- • SM&CR accountability map
2. Assess — Root causes under regulatory constraint
The DOAR identifies structural cost drivers: applications maintained for regulatory reporting that duplicate data held elsewhere; managed service contracts where exit and substitutability have never been evidenced (a PRA operational resilience gap); consumption behaviours that inflate cost-to-serve but are invisible because showback does not exist. The Governance Mapper resolves SM&CR accountability — where a technology decision affecting customer outcomes has no named Senior Manager as Accountable, that is a hard stop. Every finding that rests on a regulatory interpretation is classified as Assumption-grade until validated by the compliance function.
Methods & Tools
- • DOAR diagnostic
- • Governance Mapper
- • Evidence Ledger
Key Outputs
- • DOAR with regulatory overlay
- • SM&CR governance gap analysis
- • Third-party risk assessment
3. Design — Optimisation within regulatory bounds
The Economic Scenario Model values the addressable cost — but regulatory compliance cost is classified as mandatory and excluded from optimisation targets. Benefits resting on regulatory interpretation remain at Assumption-grade and cap confidence accordingly. The Business Case explicitly separates discretionary optimisation (application rationalisation, demand management, managed service renegotiation) from mandatory spend (regulatory reporting, operational resilience remediation). Consumer Duty fair-value assessment is embedded wherever pricing or cost-to-serve decisions affect customer outcomes. The Conflict Checker runs across all active modes before any recommendation is issued.
Methods & Tools
- • Economic Scenario Model
- • Business Case Builder
- • Conflict Checker
Key Outputs
- • Three-scenario economic model
- • EPF Five Case
- • Consumer Duty fair-value assessment
4. Engage — Board, regulator, and operational alignment
Three audiences, one evidence base. The board sees the decision case with confidence constrained by gate status — if the governance gate is BLOCKED because SM&CR accountability is unresolved, the case does not proceed to Approve. The compliance function validates every Assumption-grade regulatory claim before confidence is released. Operations receives the showback model and governance cadence that will sustain cost discipline beyond the initial review. Where operational resilience remediation is required, the execution plan incorporates NIS2/DORA constraints as hard inputs.
Deliverables
- • Board decision pack (gate-constrained)
- • Compliance validation schedule
- • Showback operating model
5. Integrate — Sustained cost intelligence and regulatory defence
EPF-Cost IQ delivers the recurring operating model: TBM cost model refreshed monthly, showback reports to business-line owners, consumption governance dashboards, and an executive decision cadence that connects cost movement to accountability. The Benefits Realisation Tracker measures actual savings against the committed case. The regulatory constraint register is maintained as a live artefact — when the regulator asks how cost allocation supports fair value, the answer is structured, evidenced, and current.
Operating Model
- • Monthly TBM refresh
- • Showback and chargeback model
- • Benefits tracking
Typical Outcomes
15–30%
Technology cost reduction with mandatory spend ring-fenced
Regulator-grade
Cost allocation defensible under Consumer Duty fair-value scrutiny
SM&CR mapped
Every technology decision resolved to a named Senior Manager
OPERATIONAL PERFORMANCE
Commercial Real Estate — Operational Performance
A commercial property firm manages a £2bn+ portfolio across office, retail, and logistics assets. Investor reporting takes three weeks every quarter because data sits in disconnected systems — lease administration, property management, asset management, and finance all operate from different versions of the truth. Technology cost-to-serve per property is unknown. Managed service providers are on auto-renew with no commercial challenge.
1. Review — Map cost, systems, and service boundaries
Build the TBM cost model specific to commercial real estate: cost pools (labour, software licences, outsourced services, data and infrastructure) allocated through to property-level consumption. Establish unit economics — cost per property under management, cost per lease event, cost per investor report cycle. Map the system landscape: which platforms hold lease data, which hold financial data, where the overlaps create duplication and reconciliation effort. Document every managed service contract and its commercial structure.
Methods & Tools
- • TBM Allocation Engine
- • DOAR Builder
Key Outputs
- • TBM cost model with property-level unit economics
- • System and data landscape map
- • Managed service contract inventory
2. Assess — Quantify duplication, leakage, and governance gaps
The DOAR diagnostic exposes the structural problems: lease data maintained in two systems with manual reconciliation consuming 400+ hours per quarter; investor reporting assembled manually from five data sources with no single version of the truth; managed service providers charging volume-insensitive fees while portfolio composition has shifted materially since contract inception. The Managed Service Contract Analyser builds each provider's P&L, isolating margin leakage from scope creep, demand-insensitive pricing, and services that duplicate internal capability. Governance assessment reveals that property-level cost decisions have no named owner — the investment committee approves capital expenditure but nobody owns the operating cost per asset.
Methods & Tools
- • DOAR diagnostic
- • Managed Service Analyser
- • Governance Mapper
Key Outputs
- • Data duplication analysis
- • Managed service P&Ls with leakage
- • Governance weakness register
3. Design — Target operating model and commercial recovery
Design the future-state operating model: a single lease-to-report data flow eliminating the reconciliation burden; a renegotiated managed service structure with volume-sensitive pricing, defined service catalogues, and exit provisions; a showback model that allocates operating cost to each fund or portfolio so investors see cost-to-serve alongside performance. The Economic Scenario Model values each lever — data consolidation, managed service renegotiation, demand controls, reporting automation — with the cash-releasing/reallocation split enforced. The Business Case assembles the Five Case for the board with the Governance Mapper ensuring every property-level cost decision now has a named accountable individual.
Methods & Tools
- • Economic Scenario Model
- • Business Case Builder
- • Governance Mapper
Key Outputs
- • Target operating model
- • Managed service renewal positions
- • Three-scenario economic model
4. Engage — Align board, fund managers, and service providers
The board receives the investment case with confidence constrained by evidence quality — managed service savings are Evidenced where contract terms are known, but demand reduction benefits from reporting automation remain at Inferred grade until a pilot confirms adoption. Fund managers receive the showback model showing cost-to-serve per fund for the first time. Service providers receive the commercial challenge — renewal positions built from the Managed Service Analyser with unit economics, benchmark comparisons, and a clear alternative-provider assessment.
Deliverables
- • Board investment case
- • Fund-level showback model
- • Managed service renewal packs
5. Integrate — Investor-grade operating discipline
EPF-Cost IQ embeds the recurring operating model: monthly TBM refresh with property-level unit economics, quarterly investor reporting from a single data source, managed service performance tracked against renegotiated SLAs with commercial consequences. The Benefits Realisation Tracker measures actual cost reduction against the committed case — reporting cycle time, reconciliation effort, managed service cost per property. The governance cadence ensures cost-to-serve is a standing board agenda item, not a one-off diagnostic that decays after three months.
Operating Model
- • Monthly cost-to-serve reporting
- • Quarterly investor dashboard
- • Benefits tracking
Typical Outcomes
£1.5–3.6m
Annual value through reporting rework, data deduplication, and service optimisation
3 weeks → 3 days
Investor reporting cycle compressed through single-source data
Unit economics
Cost per property, per lease event, per report — visible for the first time