MARKHAM
Research
White paper · MKM-R-2026-005

AI adoption as an operating-model problem

AI does not fail on model quality. It fails on ownership.

Why AI initiatives that report to IT underperform those that report to the operating committee.

Length18 pages
Samplen = 63 initiatives
Period2023–2026
AuthorsMarkham Institute
ReferenceMKM-R-2026-005
Version1.0 · Current
Download the excerpt (PDF)Request the full 18-page edition
The summaryA nine-minute read

AI does not fail on model quality. It fails on ownership.

We tracked 63 AI initiatives across the transformation cohort from approval to month 18. The single strongest predictor of value delivery was not vendor choice, data readiness or model quality. It was the reporting line. Initiatives owned by the operating committee and priced per use case delivered a median 68% of their value case; initiatives owned by IT delivered 23%. The technology was frequently identical.

The mechanism is the same one that governs every transformation in our data: AI creates value only when the operating model changes around it — when a forecast changes an ordering decision, when an automated answer closes a case without a queue. Those are operating decisions, and IT does not own them. This paper sets out the ownership test, the per-use-case pricing discipline, and the four failure patterns that account for most of the write-offs in the sample.

Key findings
68%

Median value delivery at month 18 for AI initiatives owned by the operating committee — against 23% for those owned by IT.

11

Use cases in production in the strongest programme in the sample, each priced and verified separately.

4×

Higher write-off rate for initiatives approved as platforms rather than as priced use cases.

54%

Share of sampled initiatives that could not state, at approval, which operating decision the model would change.

Inside the report5 chapters · 18 pages
01
The reporting-line result68% against 23%, on frequently identical technology. The evidence in full.
4 pages · 5 min
02
Why the operating committeeAI value is realised at the operating decision. Whoever owns that decision owns the outcome.
4 pages · 5 min
03
Pricing per use caseThe approval discipline that separates the top of the sample from the write-offs.
4 pages · 5 min
04
The four failure patternsThe platform trap, the pilot loop, the orphaned model, and the unowned decision.
4 pages · 5 min
05
MethodSample construction, verification standard and attribution rules.
2 pages · 3 min
If you only act on four things

The findings, as Monday-morning decisions.

a

Move AI ownership to the operating committee. If the initiative cannot name the operating decision it changes, it is not ready to be funded.

b

Approve use cases, not platforms. Each use case carries its own value case, baseline and kill test.

c

Verify per use case at a fixed cadence — the same standard applied to any other transformation outcome.

d

Treat data and platform work as enabling spend inside a priced use case, never as a value case of its own.

Methodology & governance
Sample63 AI initiatives across 41 organisations in the transformation cohort, approved 2023–2025, tracked to month 18.
ClassificationOwnership is classified by where decisions about scope, funding and adoption were actually made — read from minutes, not org charts.
VerificationValue delivery is measured against the initiative’s approval-time value case under the Verification Standard (MKM-F-003).
ExclusionsInitiatives without a written value case at approval are excluded from delivery statistics and reported separately.
Citation

Markham Institute, AI adoption as an operating-model problem, MKM-R-2026-005, v1.0 (March 2026). Citation permitted with attribution.

Revision history
v1.0 · Mar 2026First publication. 63 initiatives tracked to month 18.