AI investments fail partly because no one owns the thread from spend to outcome. I do.
Twenty years inside the transformation value gap, now the AI value gap. Across waves of optimisation and transformation, approaches evolved, but the underlying challenge remained. Each addressed a symptom, not the gap itself. That breadth now points upstream, at what decides whether AI investment converts.
What the value gap is.
The transformation value gap
The disconnect between invested transformation outcomes and the value realised in operations.
It has existed since organisations first spent capital on technology expecting outcomes that the technology alone could not produce. The gap persisted. Spend grew. Conversion didn't.
The AI value gap
The disconnect between rapid AI adoption and experimentation, and the realised business outcomes.
The latest expression of the same failure. Enterprises are deploying AI at pace. The boards funding it cannot track whether the outcomes remain achievable. The technology is rarely the problem.
The thread
The transformation value gap persists in new form. The AI value gap reflects the same structural break between investment and realised value, amplified by AI scale and speed.
What the data tells us about the value gap.
The value gap is not a new observation. The data has been consistent for over a decade. Two eras, one pattern. The scale of the numbers changed but the underlying failure remained.
Early 2010s · The transformation value gap
Spend was rising. Conversion wasn't.
$3.4T
Digital transformation spend forecast by 2026, up from $1.5T in 2021
IDC, 2022
2013
Gartner's Hype Cycle placed digital business at the peak of inflated expectations, far ahead of demonstrated productive value
Gartner, 2013
2012
Large-scale IT projects routinely ran over budget, over schedule, or below value
McKinsey, 2012
2,800
CIOs surveyed found digital business creating a widening reality gap between ambition and delivered value
Gartner, 2015
30%
of digital transformations succeed. Seven in ten never close the gap between what was funded and what was delivered
BCG, 2020
IDC, Gartner, McKinsey, BCG, 2012–2022
early 2020s · The AI value gap
A decade on. Same signature.
28%
of AI use cases meet ROI expectations. 20% fail outright
Gartner, 2026
<1%
of executives report AI ROI of 20% or greater. 53% report returns of 1 to 5%
Forbes Research, 2026
5%
of enterprise AI pilots produce meaningful revenue acceleration across 300+ deployments
MIT NANDA, 2026
2×
AI projects fail at more than double the rate of non-AI technology projects
Fortune, RAND Research, 2026
42%
of organisations abandoned most of their AI initiatives in 2025, up from 17% the year before
S&P Global, 2026
40%
of agentic AI projects are predicted to be cancelled by end of 2027
Gartner, 2025
Gartner, Forbes Research, MIT NANDA, RAND, S&P Global, 2025–2026
The failure pattern
The transformation value gap has not closed. It reappears as the AI value gap. Same structural break between investment and realised business value, different technology stack.
How the value gap shows up.
Three common investment-to-outcome fracture points.
01
No one is accountable for continuous investment-to-value coordination post-investment
MIT NANDA found 95% of enterprise AI pilots produce no measurable P&L impact. The investment is approved. Delivery owns the build. No one owns conversion. As assumptions change, the link between investment and outcome breaks. The AI initiative progresses. The business case does not.
02
The investment changes form at every domain it crosses, and nothing tracks it across the change
Capital becomes a decision. A decision becomes work. Work becomes an outcome. Each domain hands off a different version to the next, and no framework, governance, delivery, or benefits management, tracks it across the full chain.
03
Organisational complexity destroys outcome alignment at scale
McKinsey puts AI value capture at 20% model, 80% organisational rewiring. As AI scales, it crosses more domains, decisions and funding pools. Informal coordination breaks down. Priorities compete. Objectives misalign. The line between spend and outcome weakens.
"Twenty years working inside the transformation value gap, now the AI value gap: building operating models, applying optimisation approaches, deploying enabling technology, all to extract value. The work has not changed."
Chuks Anochie · Value Governance Architecture · AI Value Orchestration
Two disciplines that bridge the value gap.
The AI value gap is not closed by default. It is contained through the deliberate design and execution of two disciplines. Neither is sufficient on its own.
Value Governance Architecture
The accountability structure that connects every domain an AI investment touches to the outcomes it was funded to produce.
The design discipline of who owns what, who decides what, and who is accountable for what, across every part of the organisation an AI investment touches. Governance failures surface before they become costly.
Aligned to applicable AI governance and data protection obligations across UK, EU, UAE, GCC, and US regulatory environments, including ISO/IEC 42001 and NIST AI RMF.
AI Value Orchestration
The real-time end-to-end coordination of the Value Governance Architecture, from mandate to handover.
The run discipline that tracks feasibility, risks, trade-offs and drift across every domain an AI investment traverses, keeping the board informed on whether value conversion remains achievable. Not compliance governance or investment-decision advisory, a continuous operating discipline governing the path from investment to outcome, not a policy applied once or a recommendation handed elsewhere to execute.
See how an engagement runs on the How I Work page.
Outcome and Value: two concepts that underpin the two disciplines.
Most organisations treat them as the same thing. They are not. An outcome answers one question. Value has four, and nothing sits outside them.
The measurement problem
50%
measure AI value through data quality improvements
Forbes Research, 2025
48%
use employee productivity as their primary metric
Forbes Research, 2025
<1%
of executives report significant ROI from AI investments
Forbes Research, 2025
Productivity and data quality are proxy metrics. They do not appear on the P&L line the investment was funded to move. Measuring them instead of the outcome is what makes AI spend invisible to the CFO.
Value Governance Architecture
Governs financial and regulatory conversion across every domain the initiative touches.
AI Value Orchestration
Drives financial and operational conversion by keeping every domain aligned to the funded outcome.
Value framework informed by: McKinsey State of AI (2025) · BCG Build for the Future (2025) · Gartner Business Value Model · Forbes Research AI Survey (2025) · McKinsey New Economics of Enterprise Technology (2025)
Why you need both disciplines.
Fortune and RAND cite industry estimates that AI initiatives fail at more than double the rate of non-AI initiatives, this makes the sequence critical.
Value Governance Architecture creates the accountability line between AI investment and intended outcomes. AI Value Orchestration keeps every domain aligned through decisions, trade-offs, and dependencies, maintaining oversight of feasibility and value realisation risk to protect value realisation.
An AI initiative without Value Governance Architecture
has no structure for AI Value Orchestration to run against. The initiative moves. The investment is not continuously assessed against its intended outcomes, feasibility, and likelihood of value realisation.
An AI initiative with Value Governance Architecture but no AI Value Orchestration
has a designed structure and no one running the initiative through it in real time. The structure becomes documentation.
An AI initiative without either discipline
can be delivered as planned. Whether the outcome the board funded remains achievable is a separate question, and nothing in the initiative is set up to answer it.
Value Governance Architecture is the foundation. AI Value Orchestration is the live operating layer that runs the AI initiative through that foundation.
51
successful enterprise AI deployments analysed
The differentiator was the organisation, its readiness, its processes, and its governance, not the AI model. Both disciplines sit on that finding.
Stanford University, 2026
What remains broken without both disciplines.
01
Incentives are weighted to activity, not outcome
Executive and delivery incentives track deployment and spend. The board asks. Delivery answers on activity. Finance answers on spend. Neither can answer on outcome.
02
Decision rights aren't named where domains meet
Use cases run as isolated initiatives with local KPIs. At portfolio level, no single seat is accountable across every domain influencing one AI initiative's outcome.
03
What gets built drifts from what was approved, and the baseline doesn't move
Business demand and technical execution evolve past the point of approval. Investment cases stay anchored to the original assumptions. Decisions keep getting made against a baseline that's gone stale.
04
Technical knowledge and governance knowledge sit with different people
AI spend is reported as efficiency and delivery metrics. Revenue, cost, and margin outcomes are tracked on a separate line. Decisions get made at the wrong level, with the wrong half of the picture.
05
Value is not continuously tracked through change
Each trade-off is reasonable on its own. Tracking is episodic, not continuous. By the time the exposure is visible, it's already accumulated across a string of decisions no one document connects.
06
The fastest domain sets facts the slower ones haven't sanctioned
Domains move at different speeds by design. Complexity from scale isn't the exception, it's the default. Traditional coordination isn't built to keep pace with it, so alignment erodes as the portfolio grows.
Who runs both disciplines.
Twenty years helping organisations inside the gap. Every experience helped address symptoms of the gap. None of them closed it. Two disciplines built on that observation.
Learn more about Chuks →$7.4M
portfolio, $30–40M value protected
AI-native governance at a flagship Gulf NOC · Two concurrent agentic AI platforms
20
yrs
Enterprise optimisation across five waves
Lean · ERP · Scaled Agile · DevOps · FinOps · AI-native governance
£30M+
Transformation portfolio managed
Capgemini · Infosys Consulting · Director level
Sector exposure
Markets
UAE · GCCUnited Kingdombridge the gap
The end goal is conversion. The evidence is on the board report. Neither happens by default.
Engagements are fractional, interim, or full mandate.