No one escalated

Think of the person three levels below you who would be first to know. Not your strongest performer. The one whose job puts them in front of a problem before it has a name.

Your escalation log does not contain them. It contains your direct reports, and the week the problem reached a steering meeting. Most escalation paths route through the person whose decision is being questioned, so the path filters the information before the log records it.

Someone already knows. The question is whether it changes anything.

What silence measures

Silence in an organization is a reading rather than a blank. A problem that runs six weeks downstream and arrives upstream in week seven has produced six weeks of data. Those weeks record how long accurate information took to cross a structure. No reporting system has a field for it.

A detector that has stopped going off reports one of two things. The building is not burning, or the detector has been disconnected. The panel reads the same in both cases, which is the same defect as a review line that records a manner.

Why the concern stopped arriving

Most people who go quiet raised something first. They raised it, it was heard, it was agreed with, and the decision stayed where it was. The second time, they ran a calculation about whether raising it was worth the cost. After enough of those calculations, the reporting stops while the perceiving continues.

In my next book, out in April 2027, I put it this way: "The person who stopped speaking up didn't stop perceiving. They stopped reporting." The diagnostic information is still in the building. What the organization lost was access to it.

This is a budget item before it is a values argument. MIT Sloan found toxic culture 10.4 times more powerful than compensation in predicting a company's attrition rate relative to its industry (MIT Sloan Management Review, Toxic Culture Is Driving the Great Resignation, 11 January 2022, from 1.4 million Glassdoor reviews and 34 million employment profiles, April to September 2021). The industry-relative part is the construct, and it is the half that gets dropped. It is a ratio, and ratios survive a budget meeting better than totals.

The objection, and it is the strong one

The standard response is that the person should have communicated better. It is a fair objection. Someone who can price a concern in dollars, name the decision it affects and put it in front of the right person usually does get action. That skill is real.

It works for the person who has it. It does nothing for the next person. An organization whose access to accurate information depends on each employee's ability to argue has built nothing. It handed the routing to individual persuasion and kept no record of what failed to route.

Two dates, and the gap between them

Every expensive problem has two dates. The day someone downstream could have described it, and the day it changed a decision. The gap is the measurement, recoverable from material the company already holds.

Take the last three problems that cost real money. For each one, find the earliest written artifact in which someone described the thing accurately. A ticket, a message, a meeting note. That is date one. Then find the decision that moved in response. The budget changed, the supplier changed, the launch slipped. That is date two.

Three problems, two columns, one afternoon. The median gap is the number to bring to the next operating review, and a median of three is an indication rather than a measurement until there are ten of them.

Then count the escalations that reached you last quarter from below your direct reports. If the count is zero, zero is the reading.

My own accounting, after thirty-seven years running teams and companies, is that the gap never closed on its own. It closed when the information reached someone with no stake in the original call. That is not a published benchmark. It is my accounting. I built explicit systems because my brain could not navigate implicit ones. Written promotion criteria. Structured feedback. Recorded decision rationale. Escalation paths that did not run through the person whose call was in question. Retention stayed near ninety-five percent across the companies I have run. As I put it in The Perception Revolution, psychological safety isn't proclaimed. It's designed.

For a standing instrument, Amy Edmondson's seven-item team psychological safety scale, published in 1999 and tested across thousands of teams since, measures whether people believe speaking up is survivable (Administrative Science Quarterly, Psychological Safety and Learning Behavior in Work Teams, 1999). I use it without modification. Edmondson and colleagues later tracked health-care workers from 2019 through the pandemic and found psychological safety predicted retention years afterward (Bahadurzada, Edmondson and Kerrissey, 2024).

One thing to notice this week

Pick the last thing that surprised you in an operating review. Ask the person who owns it when they first knew, and who they told. Two questions, asked once. The answer is a short interval or a long one, and a long one is the finding.

Where this stops

A reconstructed date is a memory, and memories move toward the teller. Whoever raised a concern remembers raising it earlier and more clearly than the record supports. Whoever received it will remember it arriving later and vaguer.

The written artifact is the only date worth using, and for a lot of problems the first accurate description was spoken and never written down. The gap narrows the question. It does not settle it.

The six-minute Signal Loss Check measures the same thing on a larger sample.

Two dates instead of one number. Almost no organization has either, and the second one is what survives an operating review.

What was the interval on the last problem that cost you real money?

If someone on your leadership team owns the escalation path, this is their number to run.

Sources

• Edmondson, A. (1999). Psychological Safety and Learning Behavior in Work Teams. Administrative Science Quarterly, 44(2), 350–383. Seven-item team psychological safety scale. Administrative Science Quarterly, Psychological Safety and Learning Behavior in Work Teams
• Sull, D., Sull, C., & Zweig, B. (2022). Toxic Culture Is Driving the Great Resignation. MIT Sloan Management Review, 11 January 2022. Culture measured from 1.4 million Glassdoor reviews; attrition from 34 million Revelio Labs employment profiles, April–September 2021. The 10.4 multiple is against compensation, predicting a company's attrition rate relative to its industry. MIT Sloan Management Review, Toxic Culture Is Driving the Great Resignation
• Bahadurzada, H., Edmondson, A. C., & Kerrissey, M. (2024). Psychological safety as an enduring resource amid constraints. Longitudinal study of a health-care workforce, 2019 through the pandemic period; psychological safety predicted later retention, with protective effects strongest for physicians, women and people of color. Cited qualitatively, with no figure attached.
• Mann, P. A. (2026). The Perception Revolution. Fast Company Press. Chapter 2, on the 25/25 Problem and the detector mechanism, and Chapter 10, on the architecture of psychological safety and escalation paths that do not route through the decision under question.
• Mann, P. A. (April 2027, manuscript). The Projection Fallacy. Chapter 4, on the difference between losing a perception and losing access to it. One line quoted in the body and attributed.

Peter Allen Mann is the author of The Perception Revolution (Fast Company Press, 2026) and The Projection Fallacy (April 2027). He spent thirty-seven years in leadership, including four years as a Navy Surface Warfare Officer, director roles at Dell and Tech Data, and two companies founded and run. He writes about late recognition and the profile that gets missed at peterallenmann.com.