MARKHAM
Research
The Transformation Ledger · 2nd annual edition · June 2026

The State of Mid-Market Transformation 2026

Funding is up. Delivery is down. The gap is structural.

Annual outcome study. 214 programmes, four years of data, one uncomfortable pattern.

Length48 pages
Samplen = 214
Period2022–2026
AuthorsMarkham Institute
ReferenceMKM-R-2026-014
Version2.1 · Current
Download the excerpt (PDF)Request the full 48-page edition
The summaryA nine-minute read

Funding is up. Delivery is down. The gap is structural.

Mid-market transformation is not failing for lack of ambition or investment. Programmes in our cohort were funded at a median 2.3% of revenue and sponsored at board level. They failed — when they failed — on load: too many concurrent priorities, too little verified sequencing, and governance rhythms that reviewed activity rather than outcomes.

The 2026 edition adds 61 programmes to the cohort and, for the first time, isolates the effect of implementation oversight. Programmes with independent delivery assurance retained 84% of projected value at month 24, against 47% for those without. The report sets out the full cohort data, the verification method, and the three structural decisions that most reliably separate the top quartile.

Key findings
71%

Median value delivery at month 24 for programmes holding three or fewer concurrent priorities.

22%

Median value delivery for programmes running six or more priorities — at 40% higher external spend.

84%

Value retained under independent implementation oversight, against 47% without it.

9 wks

Median decision latency saved per year in organisations that moved to a fixed operating cadence.

Exhibit 04 — Value delivered vs concurrent prioritiesn = 214 · month 24
78%
71%
54%
39%
28%
22%
1–234567+

Share of projected financial value verified as delivered, by number of concurrent programme priorities. Full method in section 6 of the report.

Inside the report6 chapters · 48 pages
01
The load problemWhy the priority count predicts failure better than budget, sponsorship or sector.
12 pages · 14 min
02
Sequencing beats shrinkingThe cohorts that phased their ambition kept it. The cohorts that trimmed it lost both.
8 pages · 10 min
03
The oversight effectNew this year: 84% against 47% value retention. What independent assurance actually changes.
9 pages · 11 min
04
Cadence and latencyNine weeks a year sit inside slow decisions. Where the time goes and how a fixed rhythm recovers it.
7 pages · 9 min
05
The top quartileThree structural decisions separate the programmes that held. None of them is a technology choice.
8 pages · 10 min
06
Method and cohortDefinitions, cohort construction and the verification standard, in full.
4 pages · 5 min
If you only act on four things

The findings, as Monday-morning decisions.

a

Count your concurrent priorities. If the number is above three, the data says you are already paying for it.

b

Sequence the ambition instead of trimming it. Phase the portfolio; keep the full programme alive on paper.

c

Separate delivery assurance from delivery. The 37-point retention gap belongs to whoever owns that split.

d

Review outcomes weekly against the month-0 baseline, not activity against the plan.

Methodology & governance
CohortProgrammes qualify at $50M–$500M revenue with a written value case and a month-0 baseline we can audit. Self-reported outcomes are excluded.
VerificationEvery outcome is tested against the Markham Verification Standard (MKM-F-003) before it enters the dataset.
Review cycleCohort re-audited and the edition revised annually, each June. Errata are published, not overwritten.
Data governanceClient identities are held under NDA; published figures are aggregated so no single programme is identifiable.
Citation

Markham Institute, The State of Mid-Market Transformation 2026, MKM-R-2026-014, v2.1 (July 2026). Citation permitted with attribution.

Revision history
v2.1 · Jul 2026Exhibit 04 re-based after the annual cohort audit.
v2.0 · Jun 20262026 edition published; 61 programmes added; oversight effect isolated for the first time.
v1.0 · Jun 2025First edition, n = 153.
Previous edition2025 · n = 153Retained for cohort comparison · MKM-R-2025-012
This edition2026 · n = 214 · v2.1Oversight effect isolated for the first time
Next editionJune 2027Be notified when it publishes