The Friction Ledger: Why Vague Findings Aren’t Findings

Most diagnostic reports end with a findings section that reads like this: “communication needs improvement.” “Change fatigue is present in some teams.” “Leadership alignment could be strengthened.” Every sentence is true. None of them are useful, because none of them can actually be acted on by anyone — they don’t say who, what, or what to do about it Monday morning.

A finding that can’t be assigned to an owner isn’t a finding yet. It’s an observation waiting to become one.

What makes something a finding, not an observation

Four things turn a vague theme into something a real organization can act on:

• It’s specific enough to be wrong. “Communication needs improvement” can’t be disproven — there’s no version of a rollout where that sentence wouldn’t apply. “The logistics team wasn’t told the new system would change their escalation path” can be checked, and either is or isn’t true.
• It names who it affects. Not “some employees” — which team, which location, which shift. A finding that applies to everyone equally usually isn’t a finding, it’s a restatement of the fact that change is hard.
• It has a proposed owner. Not a name — a role. “Ops Lead” or “IT Director,” not a person, since the point is accountability by function, not by individual. A risk with no proposed owner reliably becomes nobody’s problem.
• It has a next action, not a recommendation. “Improve communication” is a recommendation. “Publish the new escalation path and confirm it in the next team stand-up” is an action — someone can actually finish it and report back that it’s done.

Why the vague version is so tempting to write

Generic findings are safer to write and harder to be wrong about. “Communication needs improvement” is true of almost every organization at almost every point in time — which is exactly the problem. A report that could have been written before a single interview happened isn’t adding information; it’s adding pages.

Specific findings carry real risk for whoever writes them: they can be checked, disagreed with, and shown to be incomplete. That risk is also what makes them worth reading. A steering committee that’s seen a dozen “communication needs improvement” slides in past reviews has learned to skim past that sentence. A named risk with a named owner and a specific action gets read differently, because it’s the first version that could actually be wrong — and being checkable is what makes something credible.

Ranked by prevalence, not by how interesting it is to write about

The other discipline a real risk register needs: ordering by how often something actually came up across real conversations, not by severity as judged after the fact or by which finding makes the best story. A friction mentioned by one person in an offhand comment and a friction independently raised by a third of an affected team are not the same weight of evidence, even if the first one sounds more dramatic when quoted. Prevalence is what turns “someone said this once” into “this is a pattern,” and that distinction is most of what separates a risk register worth acting on from a collection of anecdotes.

The test for any finding before it ships

Before a finding goes in front of a sponsor, it should survive one question: could someone read this on a Monday morning and know what to do by Tuesday? If the honest answer is “not really, it’s more of a theme,” it isn’t done yet — it’s still an observation, and it needs to get more specific before it earns a place in the report.