When Listening Becomes a Substitute for Acting

Every year, Dell asks its employees what they think about working there. One measure in its annual “Tell Dell” survey is the company’s employee net promoter score, which reflects how willing employees are to recommend Dell as a place to work. The score was 63 in 2023 and 48 in 2024. In 2025, the score fell to 32. Employees connected the decline to layoffs, heavier workloads, return-to-office requirements, and the company’s direction. The results were not bad across the board, and Dell acknowledged the dissatisfaction. Still, employee willingness to recommend the company had fallen by nearly half in two years.

During the same period, Dell continued many of the policies employees appeared to dislike. In early 2025, it required many employees living near an office to return five days a week. Dell may have had good reasons for those decisions. An employee survey cannot settle questions about staffing, strategy, or where work should be performed. But Dell did not lack information about how its employees were experiencing the organization. Its own surveys documented the decline quite clearly.

This raises a broader question about what employee surveys are supposed to accomplish. Organizations say they conduct surveys because they want to listen. Leaders want to identify problems, understand employee concerns, and improve the organization. But gathering information is not the same as giving people influence. A survey can tell leaders that morale is low or that employees resent a policy. It can measure how much trust has declined and compare the results with other organizations. None of that means what employees say has any meaningful chance of changing what the organization does.

The distinction is easy to overlook because organizations often treat the act of asking as evidence that employees have been heard. Surveys, town halls, and listening sessions all create opportunities to speak, but they do not necessarily create a path from speaking to influence. When an organization asks employees to be candid, it creates at least some expectation that their answers will enter a real decision process. The promise is not that leaders will agree with them. Employees do not know everything leaders know, and organizations cannot operate by referendum. The more limited promise is that something remains open to influence.

Researchers Xiaowen Huang, Adrian Wilkinson, and Michael Barry describe a related idea as contextual voice efficacy. It refers to whether an organization’s structures allow employee voice to be heard and acted upon. An employee may feel perfectly safe criticizing a decision and still remain silent if experience suggests that speaking will accomplish nothing. The question is not always whether speaking is risky. It may instead be whether anyone with authority is capable of being persuaded.

Albert Hirschman made a similar point in Exit, Voice, and Loyalty. Hirschman described voice as an attempt to repair an organization from within. When people become dissatisfied, they can leave or try to change the organization. Employee surveys appear to make the second option easier. Employees do not have to complain informally or wait until they resign because the organization gives them a regular opportunity to identify problems. But voice remains a sensible response only when people believe it might have an effect. When employees repeatedly speak and see little connection between their input and what happens next, they learn something about the organization.

They also lose one comforting explanation for its behavior. Before a major survey, employees can tell themselves that senior leaders may not understand what is happening. Information may be filtered as it moves upward, or leaders may be insulated from the consequences of their decisions. A comprehensive survey makes that explanation harder to sustain. If employees are encouraged to be candid and the results clearly identify a problem, they know that leaders know. When little appears to change, the inaction communicates something. Leaders may disagree with the diagnosis, place greater weight on other priorities, or lack the ability to respond. The survey does not merely measure how employees feel. The response to the survey gives them new evidence about the organization.

None of this requires leaders to be dishonest. Collecting information is simply easier than acting on it. A survey can be purchased, administered, benchmarked, and presented. Leaders can announce priorities and establish committees. The process consumes money and attention, which can make it feel like progress. The problems identified by the survey are usually harder. Addressing them may require spending money, reconsidering a strategy, reversing a decision that leaders have already defended, or surrendering some degree of control. Diagnosis is not intervention.

Research on what happens after employee surveys finds that follow-up depends heavily on whether leaders support implementation and whether the people expected to make changes have the authority and resources to do so. Without that follow-up, listening can become monitoring. Monitoring tells leaders how employees are reacting to decisions. Listening leaves open the possibility that the organization itself should respond differently. Employee surveys can serve both purposes, but problems arise when organizations claim to be offering voice while mainly measuring reactions to choices that are no longer open.

Employees eventually adjust. Some stop completing the survey. Others answer carefully and expect little. Leaders may describe the decline in participation as survey fatigue, and sometimes that is probably right. In other cases, what looks like fatigue may be a judgment. People do not always stop speaking because they have nothing to say. Sometimes they stop because the organization has taught them what speaking accomplishes.

A more honest process would begin by telling employees what remains open to influence. Afterward, leaders should explain what they learned, what changed, and why some concerns did not lead to action. Sometimes the answer will be that leaders heard employees and chose another priority. That may disappoint people, but it is still more respectful than treating the opportunity to participate as though it were influence.

Dell may have had defensible reasons for the decisions it made. We also cannot know from outside the company how its survey findings affected internal discussions. The public pattern is still revealing. Dell asked employees whether they would recommend the company as a workplace. Their answer became much less favorable for two years in a row. The company knew and continued asking.

Organizations have become very good at collecting information from the people who work within them. The harder task is remaining capable of being changed by what those people say. An organization does not demonstrate that it values employee voice by asking more often. It demonstrates it when employees can reasonably believe that their answers might alter what happens next.