Organizational Diagnosis
Seventeen recurring patterns sit behind most execution problems: who can decide, what information reaches them, and what the organization remembers.
This page explains the framework in plain terms. If you would rather start with your own organization, the assessment takes about five minutes and runs in your browser.
Use this to
The Problem
When an organization struggles, the explanation offered is usually “culture.” The leadership team isn’t aligned. Communication is poor. People aren’t accountable.
Those describe what the situation feels like. They rarely say what is broken, so the fix becomes a values refresh or an offsite, and the same problems come back with new names.
Look underneath and the conditions are more specific. A team is accountable for an outcome but needs approval for the decisions that drive it. The people who decide are working from information that stopped being accurate a year ago. The organization has been reorganized so many times that the reason behind a standing rule left with the person who wrote it.
Conditions like these can be examined, and they repeat from one organization to the next. That is what makes them worth naming.
The Framework
Every failure mode maps to one of three structural dimensions — the minimum conditions for an organization to function.
Can the organization see itself clearly? Does the information that reaches decision-makers reflect what’s actually happening?
Do the people making decisions have the mandate, context, and standing to make them? Does accountability follow authority?
Does the organization remember why it made the decisions it made? Does context survive transitions and reorgs?
When all three hold, the organization is coherent: it can decide, act, and remember. When one breaks, people make up the difference by hand. When two break at once, the patterns start feeding each other, and from inside they look like personalities rather than structure.
The Taxonomy
The taxonomy names 17 failure modes: conditions that repeat across organizations, sectors, and sizes, and that feed themselves until something structural changes. They came out of twenty years inside large service operations, frontline to vice president, and were then tested against public evidence on 31 organizations.
One example. Responsibility Compression: responsibility is pushed down to the people doing the work while the authority and context needed to act stay upstream. Everyone is accountable for the outcome and nobody can change the conditions that produce it. Most people who have worked in a large organization can name a time they lived inside it.
The Research
The assessment on this site asks you about an organization you know. The research behind the taxonomy works the other way: it reads an organization from the outside, using only public evidence. Filings, earnings calls, press, employee reviews, litigation, and the record of organizational changes.
A pipeline of language models extracts candidate findings from that evidence and maps them to the seventeen patterns. Each finding then gets a second pass from a separate model prompted to break it. Findings that fail are dropped. That is a check on the instrument’s consistency, not proof that a finding is true of the company, and the Coherence Record publishes the instrument’s limitations alongside its results.
135 runs across 31 organizations in six sectors, with the methods, failures, and corrections written up as they happened.
If you want to see what the output looks like, there is a sample scan on the Labs site, and the Record on this one.
For Leaders
Treat the seventeen patterns as vocabulary. Most of them describe things your team already talks around: the initiative with five owners, the approval that takes a month, the dashboard nobody on the floor believes. Having a name for the pattern makes it discussable as structure instead of as somebody’s failing.
A practical way in: pick the two patterns that sound most like your situation, read their pages, and take the question at the end of each into your next staff meeting. The answers you get are the first piece of evidence.
Read the patterns at your own pace, or start with an organization you know.
Pick whichever matches what you came for.
Learning the framework
Explore the seventeen failure modes and the three structural dimensions behind them. Reading, at your pace, no inputs required.
Explore the Taxonomy →Naming a live problem
Sixteen scenarios about an organization you know. About five minutes, in your browser, and you leave with the patterns your answers point to and a question to ask for each.
Try the assessment →DRI Practice is the work of Justin R. Greenbaum, run as a practice at Greenbaum Labs, an independent advisory and applied-AI practice. The seventeen failure modes came out of running large customer operations and have been tested against public evidence on 31 organizations, with the methods and their limits published in the Coherence Record.