AI Business Value Framework

Your AI portfolio is spending.
Is it delivering?

Map every AI investment to the metrics that matter. Diagnose governance gaps. Hold up the mirror.

📊

Executive View

Board-ready dashboards. Role-based KPIs. See what your AI portfolio is really delivering.

CEO · CFO · CDO · CIO · CHRO · CRO

⚙️

Practitioner View

Full assessment toolkit. Build your portfolio, run the agent, diagnose governance gaps.

AI Leads · Transformation · Advisory

88% Use AI in at least one function McKinsey 2025

39% Report enterprise EBIT impact McKinsey 2025

50%+ GenAI projects abandoned after PoC Gartner 2026

How It Works →

🔒 No data leaves your browser

📊 Model assumptions shown with every result

🏛 EU AI Act & DORA aligned

⚡ Setup in under 5 minutes

← Back to Welcome

Quick Start Guide

How the BVF Works

The AI Business Value Framework maps your AI investments to the financial outcomes and governance standards your board expects. Here is how to get the most from it in under 10 minutes.

Set Your Company Profile

Enter your revenue, cost base, EBITDA and industry. This scales the planning ranges and makes every assumption visible, then replace those assumptions with organisation evidence.

Define Strategic Priorities

Add 3 to 5 board-level priorities your organisation is measuring AI against. These drive the alignment scoring across your portfolio and the Boardroom Sentence on the Executive Dashboard.

Import or Build Your Portfolio

Bring in the file you already have, then enrich what is missing: owner, budget, change allocation, AI tier, function, risk and priority linkage.

Read the Executive Dashboard

Switch between CEO, CFO, CDO, CIO, CHRO and CRO views. Each role sees the KPIs and language that matter to them, drawn from the same underlying data. Print the Executive Readout for board packs.

Explore the Alignment Map

See your full AI value potential mapped by business function, investment vs return, and strategic priority alignment. This is where the commercial story comes together.

What Each Tab Does

◈

Dashboard

Role-based C-suite views with live KPIs, strategic priorities and the Boardroom Sentence

⬡

Portfolio

Add and manage AI initiatives. Budget, CM allocation, tier classification and value scoring

⚡

Agent

AI-powered classification engine. Paste an initiative description and get instant BVF scoring

⚖

Governance

EU AI Act, DORA, GDPR mapping. Regulatory exposure and compliance posture across your portfolio

◬

Alignment Map

Full-picture view mapping value potential by function, investment vs return and ROI ranges

📊

Benchmarks

Published findings kept separate from the AI BVF planning assumptions used in each calculation

Ready to Start?

Choose your entry point and the BVF will guide you from there.

Portfolio position

Strategic priorities and board objectives

Define the board outcomes every AI initiative must serve. Strong priorities have an owner, timeframe and measurable target.

0 of 5 defined

Define 3-5 strategic priorities your board is measuring AI against.

e.g. "Cut operational costs by 35%", "Drive €50M new revenue via AI", "Achieve regulatory compliance by Q3"

Executive view by role

CEO · Strategic Portfolio Alignment

Are my AI investments moving strategic priorities — or just consuming budget?

The BVF maps every AI initiative against your board-level priorities, quantifies the governance gap, and delivers a Decision Confidence Score that tells you which investments to accelerate, fix, or stop.

BVF Answer

Complete the Company Profile and build your Portfolio to see live strategic alignment metrics.

—

AI Initiatives

Total portfolio tracked

—

Total Investment

Combined AI budget

—

Modelled drag exposure

AI BVF planning model

—

CM Health

against the AI BVF 10% review point

CFO · Financial Governance & ROI

What is the true ROI of our AI portfolio — and where is the financial drag?

A planning range that includes change, training and readiness assumptions, with each assumption exposed for review. EBITDA impact per initiative.

BVF Answer

Complete the Company Profile and build your Portfolio to see financial governance metrics.

—

EBITDA Impact

Projected AI contribution

—

Modelled drag exposure

AI BVF planning assumption

—

ROI Range

Investment return potential

—

Change allocation

against the AI BVF review point

CDO · Data Maturity & Scaling

Which AI initiatives are ready to scale — and which are governance liabilities?

Gen 1/2/3 AI maturity assessment. Agent Skills Directory. Regulatory mapping across the portfolio.

BVF Answer

Complete the Company Profile and build your Portfolio to see AI maturity and scaling metrics.

—

Gen Maturity

Portfolio AI generation mix

—

Initiatives

Total portfolio tracked

—

Alignment Rate

Mapped to strategic priorities

—

Agentic Ready

Gen 3 portfolio share

CIO · Technical Governance & Risk

How do we govern agentic AI deployment without becoming the bottleneck?

Cross-functional governance framework. Integration complexity scoring. Technical risk posture across the AI portfolio.

BVF Answer

Complete the Company Profile and build your Portfolio to see technical governance and deployment readiness.

—

High Risk

Initiatives flagged high-risk

—

Gen 3 Count

Agentic initiatives to govern

—

Reg. Frameworks

Active compliance obligations

—

Pilot Trap Risk

Initiatives failing readiness

CHRO · Change Management & Adoption

Is our workforce ready for AI — and are we measuring adoption or just deployment?

Change readiness assessment, work architecture evidence and reskilling investment calculation. The people side of every AI initiative.

BVF Answer

Complete the Company Profile and build your Portfolio to see change management and workforce readiness metrics.

—

Change investment

against the AI BVF 10% review point

—

CM Health

RAG status across portfolio

—

Adoption Risk

Low CM initiatives

—

CM Budget

Total people-side investment

CRO · Regulatory Risk & Compliance

What is our regulatory exposure across the AI portfolio — and can we prove compliance?

EU AI Act classification. DORA compliance posture. GDPR and AML exposure mapping across every initiative.

BVF Answer

Complete the Company Profile and build your Portfolio to see regulatory exposure and compliance posture.

—

Active Frameworks

Regulatory obligations tracked

—

Compliance Cost

Estimated regulatory investment

—

High Risk Inits

Requiring governance attention

—

Agentic Exposure

Gen 3 regulatory risk

Holding Up The Mirror

"What percentage of the ROI for this AI programme relies on people actually adopting and implementing the change?"

The silence that follows is diagnostic. — Craig Horton

AI Initiative Portfolio

A complete view of your AI investments — existing and planned

Why map your AI portfolio?

Most organisations already have AI initiatives in flight — some delivering value, others stuck in pilot. This isn't about starting over. It's about getting a complete picture of what you've invested, what's working, and where the gaps are. Add your existing programmes first, then layer in new opportunities. The BVF will show you how each initiative connects to board-level outcomes.

Start by mapping what you already have

The best AI strategies build on existing investments. Add your current programmes first — even if they're early-stage or underperforming. Then add planned initiatives to see the full picture.

📋

Add Existing Initiative

A programme already in flight — deployed, in pilot, or approved

✨

Add New Opportunity

A proposed programme to evaluate for investment and board alignment

🤖

Use AI Agent

Paste a strategy doc and let AI extract and map initiatives automatically

AI Portfolio
Risk Intelligence

Paste your client's SaaS inventory. The BVF Agent applies disclosed classification rules and produces review prompts, which must be checked against current product and contract evidence before a replacement decision.

Anthropic API key

🛡 Your key stays in your browser — never sent to any server except Anthropic's API. Nothing is logged or stored.

Remember key for this session Facilitator mode — saves to browser session only, cleared on close

Get a key at console.anthropic.com · Keys use Anthropic's standard pay-per-use pricing

Client name

SaaS & AI initiatives inventory

0 tools entered

Live classification uses your Anthropic key. Demo mode shows the board-ready output instantly.

⚙️ Agentic Control Panel

Add context about your client's environment. The BVF agent evaluates every Agent-Replaceable and Agent-Augmented tool and recommends the optimal LLM, deployment model, and implementation approach — scored against platform fit, process type, risk profile, and business case.

Industry / sector

Risk tolerance

Data sovereignty requirement

Technical maturity

Primary cloud platform

Annual AI budget range

Change management health check

AI BVF uses 10% as a funding review point and 15% as an upper sensitivity band, then asks whether the change work is named and costed.

Add initiatives in the Portfolio tab to generate your CM Health Check.

Holding up the mirror

"What percentage of the ROI for this AI programme relies on people actually adopting and implementing the change?"

The answer is almost always: most of it. And the CM budget is almost always the first thing cut. — Craig Horton

Regulatory readiness, 2025 to 2027

Select the regulations that apply, then review the obligations and estimated cost against your revenue.

Holding up the mirror

"What percentage of the ROI for this AI programme relies on people actually adopting and implementing the change?"

Regulatory readiness is not a legal question. It is a board-level risk and cost question. — Craig Horton

v2.0 Diagnostics

Three diagnostic layers that sit behind the Decision Confidence Score. Change Enablement measures your people-side readiness using the EY/Oxford Six Drivers. Pace Layer detects structural misalignment across transformation horizons and puts a financial cost on it. Healthcare Readiness applies sector-specific overlays when the industry warrants it. Together, they answer the question the DCS alone cannot: why an initiative will succeed or fail.

EY and Oxford Saïd studied 935 leaders and 1,127 workforce members across 23 countries. This diagnostic uses their six drivers to test the organisation's capacity to absorb change and feeds the answers into the Decision Confidence Score.

What to do: Score each of the six drivers below from 1 (weak) to 5 (strong). The composite converts to a CE percentage that weights into the DCS. If CM budget is below 10% of initiative budget, CE is capped at 40%.

The Six Drivers

The study found that above-average practice across all six drivers increased the predicted likelihood of transformation success from 28% to 73%, a 2.6 times difference.

EY and Oxford Saïd Business School, 2022

Score Each Driver (1-5)

CE COMPOSITE (1-5)

—

Average of all 6 drivers

CE PERCENTAGE (0-100%)

—

((Composite - 1) / 4) x 100

CM BUDGET ADEQUACY CHECK

If CM Budget < 10% of Initiative Budget, CE is capped at 40%

AI Alignment Map — Board Summary

One-page board-ready view. Complete the Company Profile and build your Portfolio to activate all sections.

AI Alignment Map · Craig Horton Advisory · 2026

[Company Name]

Universal / Other

Holding Up The Mirror

"What percentage of the ROI for this AI programme relies on people actually adopting and implementing the change?"

This is the board view. Every number flows from your inputs. The gap between value potential and CM allocation is the conversation that needs to happen. — Craig Horton

Evidence and planning assumptions

Every number in the BVF is derived from published research, not opinion. This page explains what we use, who published it, when, and how we apply it.

✓

Evidence reviewed 23 August 2026

Next scheduled review: Q3 2026 · Sources are validated against latest publications annually

AI value potential by function

Revenue uplift & cost reduction ranges per business function

Core model

Primary sources

McKinsey, The State of AI 2025, 88% report regular AI use in at least one function, while 39% report enterprise-level EBIT impact.

Deloitte, AI ROI 2025, most respondents expect satisfactory returns in two to four years, while 6% expect them in under one year.

Gartner, GenAI project failure 2026, at least half of GenAI projects were abandoned after proof of concept by the end of 2025.

How the calculation uses them

AI BVF applies disclosed function ranges, industry multipliers and readiness adjustments as planning assumptions.

The cited research frames the adoption and value problem, it does not validate the model's individual percentages.

Replace every estimate with a measured baseline, addressable volume, unit economics and an agreed capture rate before funding.

Organisational drag cost model

Adoption friction that erodes AI ROI based on organisational readiness

Drag model

Primary sources

Prosci, budgeting for change management, 10% is the most common allocation in its survey data, with an explicit instruction to tailor the budget to the organisation and work.

McKinsey, The State of AI 2025, workflow redesign is one of the practices associated with enterprise EBIT impact.

Model boundary, AI BVF readiness tiers, drag rates and 5%, 10% and 15% bands are internal planning assumptions, none of the sources endorses them.

How the calculation uses them

Four organisational readiness levels with base drag rates:

                Bureaucratic   40% drag

                Traditional    25% drag

                Mixed          15% drag

                Agile           5% drag

Drag amplifiers increase the base rate for high-risk initiatives (+20%) or low CM allocation (<5% adds +15%, 5–10% adds +5%).

The 10% and 15% bands adjust the internal model by three and five points, this is an AI BVF sensitivity assumption and not a Prosci outcome formula.

Change management allocation review

AI BVF planning bands for change investment as a percentage of programme budget

CM MODEL

Primary sources

Prosci, How to budget for change management, 10% is the most common allocation in its survey data and the article says the budget must be tailored to the organisation and work.

AI BVF boundary, the 5%, 10% and 15% bands are disclosed planning assumptions, not compliance thresholds or promised returns.

How the calculation uses them

RAG (Red-Amber-Green) thresholds for CM as % of initiative budget:

Low allocation     <5%, review whether named change work is funded
Below reference  5–10%, below the AI BVF review point
At reference  ≥10%, at the AI BVF review point

These thresholds apply to each initiative individually. The BVF uses portfolio-level CM percentage to assess overall readiness for the CEO and CFO executive views.

Pilot trap diagnostic

5-question readiness check to assess whether an initiative can scale beyond pilot

DIAGNOSTIC

Primary sources

Gartner, GenAI project failure 2026, at least half of GenAI projects had been abandoned after proof of concept because of data quality, risk controls, cost or unclear value.

Model boundary, Gartner supports the need to test value and controls, it does not validate the five AI BVF questions or their score.

How the calculation uses them

Five binary questions (Yes/No), each worth 1 point:

1. Does this initiative have a named executive sponsor?
2. Is there a dedicated product owner?
3. Is there a plan to move beyond pilot?
4. Are KPIs defined and measurable?
5. Is CM specifically funded in the budget?

A score below 2/5 flags the initiative as at risk of the "pilot trap" — technically functional but unable to scale to deliver business value.

AI generation tier classification

Gen 1 / Gen 2 / Gen 3 framework for categorising AI maturity

TAXONOMY

Primary sources

AI BVF taxonomy, Gen 1, Gen 2 and Gen 3 are internal categories used to compare automation, generative AI and agentic systems.

Model boundary, the tier multipliers and payback values are planning assumptions, not analyst findings.

Craig Horton Advisory — Practitioner experience across 50+ enterprise AI engagements. The three-tier naming convention (Gen 1/2/3) is our simplification for board-level communication.

How the calculation uses them

Gen 1 — Automation
RPA, rules-based, ML classification, predictive models

Gen 2 — Generative AI
LLM-powered assistants, content generation, semantic search

Gen 3 — Agentic AI
Autonomous agents, multi-step reasoning, tool use, decision-making

Regulatory compliance frameworks

EU AI Act and DORA compliance mapping

REGULATION

EU AI Act — Regulation (EU) 2024/1689. Entered into force 1 August 2024. Prohibitions apply from Feb 2025; high-risk obligations from Aug 2026. The BVF maps initiative risk classifications against the Act's four-tier risk system (Unacceptable → High → Limited → Minimal).

DORA — Digital Operational Resilience Act (EU) 2022/2554. Effective 17 January 2025. Applies to financial entities. The BVF flags AI initiatives in financial services that require DORA compliance consideration for ICT risk management, incident reporting, and third-party oversight.

NIST AI RMF — AI Risk Management Framework 1.0 (Jan 2023). Used as supplementary guidance for risk categorisation methodology. Non-binding but adopted as best practice by multinational organisations.

EY/Oxford six drivers of change enablement

People-side capability assessment powering the CE dimension of the DCS

v2.0 DIAGNOSTIC

Primary sources

EY and Oxford Saïd Business School, the 2022 study surveyed 935 leaders and 1,127 workforce members. Above-average practice across all six drivers increased the predicted likelihood of success from 28% to 73%.

Model boundary, the study supports the six questions, it does not validate the AI BVF weighting or score thresholds.

Prosci, its budgeting guidance supplies a practical review point, while the AI BVF cap remains an internal policy choice.

How the calculation uses them

Six drivers, each scored 1-5:

                Lead         Executive sponsorship & vision clarity

                Inspire      Purpose alignment & motivation

                Care         Wellbeing & psychological safety

                Empower      Autonomy & decision rights

                Build        Capability development & upskilling

                Collaborate  Cross-functional working & trust

CE Composite = average of all 6 scores. CE Percentage = ((Composite - 1) / 4) x 100. This feeds the Change Enablement weight (20%) in the DCS formula.

If change funding is below 10% of initiative budget, AI BVF caps change enablement at 40%. This is a disclosed internal rule and should be tested against the work and evidence for the initiative.

Healthcare readiness factor

Sector-specific multiplier for healthcare and MedTech AI initiatives

v2.0 VERTICAL

Primary sources

FDA — AI/ML-Based Software as a Medical Device (SaMD) Action Plan (2021, updated 2024). Defines regulatory pathway requirements that determine go-to-market timelines for clinical AI.

WHO — "Ethics and Governance of Artificial Intelligence for Health" (2021). Establishes six guiding principles for health AI deployment. Informs our clinical validation overlay scoring.

Healthcare overlay, the interoperability and clinical-workflow questions are AI BVF policy choices and should be replaced with the organisation's controls and evidence.

AMA — "Augmented Intelligence in Health Care" policy (2023). Emphasises clinician time reclaimed as the primary value metric, not cost reduction. Informs our third overlay.

How the calculation uses them

Four overlays, each scored 1-5:

                Regulatory Readiness     FDA/CE pathway status

                Clinical Validation      Evidence base & trial maturity

                Clinician Time Reclaimed  Workflow integration depth

                EHR Interoperability     Systems integration readiness

Healthcare Readiness Factor = Average of 4 scores / 5, with a floor of 0.5. This multiplies into the DCS calculation, ensuring healthcare-specific risks are quantified rather than qualitatively noted.

A factor below 0.7 flags material risk. This tab appears only when the company profile industry is set to Healthcare/MedTech, keeping the assessment focused for non-healthcare organisations.

Pace layer diagnostic and organisational drag cost

Organisation-level misalignment detector across three transformation horizons

v2.0 DIAGNOSTIC

Primary sources

Stewart Brand — "The Clock of the Long Now" (1999). Introduces pace layering as a model for how civilisations and organisations operate at different speeds. We adapt this for AI portfolio alignment.

Jeroen Tas & Rob Theunissen — "Changing Fast and Slow." Applies pace layering directly to enterprise transformation, demonstrating how organisations that fail to synchronise fast-moving digital initiatives with slower-moving cultural and structural change create systemic drag. Core influence on the BVF Drag Coefficient model.

AI BVF pace model, the three horizons and drag rates are internal planning assumptions used to expose mismatched delivery and governance speeds.

MIT CISR — Digital Transformation research (2024). Identifies pace misalignment as a leading cause of portfolio-level failure, distinct from initiative-level failure.

Model boundary, published research informs the questions, it does not validate the drag rates or a euro loss.

How the calculation uses them

Three pace layers with diagnostic questions:

                Operational   0-12 months   Defend core operations

                Strategic     12-36 months  Extend capabilities

                Existential   36+ months   Create new business models

Drag Coefficient = Fastest layer score / Slowest layer score (range 1.0 to 3.0). Failure rate is interpolated linearly: 27% at 1.0 rising to 90% at 3.0, derived from cross-referencing MIT CISR and S&P Global data.

The model applies a disclosed readiness scenario to portfolio value. Treat the euro output as a sensitivity range and replace it with measured delay, rework and adoption data before presenting it to a board.

Update & Review Protocol

The evidence register records each source, its date, the claim it supports and what it does not support. AI BVF ranges, multipliers and weights are reviewed separately and remain planning assumptions until organisation evidence replaces them. Regulatory mappings are checked when official guidance or enforcement dates change.

A Note on Methodology

"Every consultant can produce a number. The question is whether you can stand behind it in a boardroom. We use published research — not internal estimates — because when the CFO asks 'where does this come from?', you need an answer that doesn't start with 'we think'."

— Craig Horton, Craig Horton Advisory

Facilitator Mode

Prepare for your workshop, then switch to Live Mode to guide the session step by step.

Pre-Session Readiness Check

Workshop Session Plan

Recommended: 90–120 minutes

TIME

SECTION

DURATION

TAB