Report · Free · No email required

The 6%

Why AI transformation fails in the org chart, not the model — and how to find out whether yours will, before you finish paying for it.

Ten pages · September 2026 · Lyle Davis, Ed.D. · Free, no email required — and printable to PDF from the page.

9 in 10
Use AI in at least one function
6%
Attribute more than 5% of EBIT to it — unchanged since 2025
75 : 25
Redesigned workflows — leaders vs the rest
1,719
Organizations, 97 countries

Why this report exists

Those four numbers come from McKinsey's State of AI survey, and they are almost always quoted badly. The 6% is not the share of organizations that redesigned anything — it is the share attributing more than five percent of operating profit to AI. The redesign figures are separate and, for anyone selling or buying a transformation, more useful: in the August 2026 survey, nearly three-quarters of the high performers had fundamentally redesigned how work flows, against one quarter of everyone else.

The report reads that survey against the previous one, from November 2025, because the comparison is the argument. Adoption rose over those twelve months. The share of organizations profiting from AI was six percent in 2025 and six percent again in 2026 — McKinsey's own word is unchanged. A single year's snapshot is compatible with "give it time." Two are not.

That is the whole report. Adoption is universal, return is rare, it stayed rare through another full year of better technology, and the separator is an organizational decision rather than a technical one.

The report is free and ungated. There is no form, no drip sequence, and no email required, because at this stage distribution is worth more than a list.


Contents

Ten pages, seven sections

  • 01 — The gapTwo numbers from one survey, stated correctly and sourced exactly, and why the 2024 excuses no longer hold.
  • 02 — It is not the technologyThe 55-versus-20 finding, what "redesigning a workflow" actually means, and why the redesign so rarely happens.
  • 03 — Readiness is a numberPsychological capital as a pre-deployment measure. What it predicts, why it moves, and how it differs from an engagement survey.
  • 04 — The network under the chartThe fifteen-minute organizational network map, the three patterns it always finds, and the highest-leverage move in any rollout.
  • 05 — Where the new way livesAdaptive space, why successful pilots die at quarter-end, and three structural tests you can apply this week.
  • 06 — The thirty-day assessmentAll three diagnostics as a four-week sequence an internal team can run alone, plus a table for reading the result.
  • 07 — Method and sourcesFull citations. Every figure in the document can be checked against its original.

If you would rather have it presented

There is a forty-five minute version of this built for HR and leadership audiences, and it is offered free to Utah professional associations, chapters and conferences. Attendees leave with the same three methods, ready to run.

If you organize programs for a chapter or an association and want it on a calendar, get in touch.