Every feedback system is built to find patterns and to surface what enough people are willing to say. Michael N. Weinberger, Executive Human Resources (HR) leader, points to the limitation buried in that design. The most important comment, however, is often the one nobody else made.
When a single employee describes a culture the company does not recognize, he treats it as more urgent than a theme raised by 50. “Even if it is one person saying it, you’ll have multiple people not saying it,” he says. Volume measures comfort as much as truth, and a system that acts only on consensus will reliably miss the gap between what an organization believes about itself and what its people experience.
Trust Gets Decided on Small Things
Before any of that information arrives, employees make a quieter judgment about whether the exercise is real. Weinberger describes an employee survey at a company he worked at years ago, where several people mentioned that a recently relocated copier had disrupted their work. The move had been made for sound reasons, but the company moved it back.
“When people saw that we actually responded to what they were saying, they were then able to give more feedback,” he says. The copier was never the issue. Employees were testing whether anyone was reading, and the answer determined what they were willing to raise later. Credibility in a feedback system is established on items too small to matter, and everything consequential depends on it.
The same attention applies to people who do not speak in groups. Anonymous surveys make follow-up impossible, and calling on an introvert, for example, in a meeting achieves nothing. Weinberger watches how people react during the conversation and approaches them privately afterwards, referencing what he noticed.
Acting on Feedback Means Not Acting on Most of It
Gathering feedback creates an obligation. An organization cannot change everything, and attempting to do so ensures nothing changes. “If everything’s a priority, nothing’s a priority,” Weinberger says. He caps the year at two or three initiatives and builds accountability around them. For example, management-level ownership, communication back to employees, and pulse surveys measuring progress against those specific commitments. That discipline separates a company that collected feedback from one that used it.
Managers Were Trained to Give Feedback, Not Take It
Weinberger cites Gallup’s 2026 report on the manager’s role in engagement and names the structural obstacle behind it. Managers understand feedback as something they deliver downward and receive from above, rarely as something their teams should be giving them. Changing that means making upward feedback normal and consequence-free. He advocates radical candor supported by skip-level interviews and coaching conversations that ask managers whether they have checked with their teams on a new initiative or on their own style.
He acknowledges the difficulty for people trained to provide direction, and puts the question to them plainly: how would you know whether you are effective without hearing it from the people you manage? Anonymous channels help at first, though the goal is regular face-to-face exchange. The question he asked his own teams gets there fastest. “How can I get out of your way?”
Real-Time Sentiment Still Requires Choosing
AI now reads employee sentiment continuously, and Weinberger cautions against treating that stream as a mandate to fix everything. Dashboards are useful for locating pinch points, the friction that appears during a rollout or a reorganization, and leaders should respond to what they find.
What matters is closing the loop, telling people what the data shows and what is being done about it. That is the copier example, only performed at scale. Whatever the instrument, the organization is answering the same question employees asked the first time they filled out a survey: “Is anyone on the other end actually listening?” To learn more, connect with Michael N. Weinberger on LinkedIn.
