Last week, we spoke with an organization that wanted to increase employee autonomy but unexpectedly ran into resistance. What turned out to be the issue? At first, they had held a kind of brainstorming session several times with a small working group. That group had been made up of volunteers from five different departments, and the hope was that it would represent the organization as a whole. But they also realized that 99% of employees had no opportunity to contribute, which created serious risks. They wanted to make agreements about giving employees more say, but they did not know how. They also realized that such a working group usually consists of the same familiar people, the ones who are already eager for more autonomy themselves. They speak from their own needs, assumptions, and perspective. That is understandable, but it is risky, especially for a complex issue like “more autonomy.”
That is why HR and management decided to run an employee survey, also known as a pulse survey. These are often done with Google Forms, Microsoft Forms, or other free survey tools, or with survey platforms already used for other questionnaires, such as Effectory or Enalyzer. We see two common versions of this: one is free, and with some effort and manual work, it is possible to send it out yourself. The other is when people say, “We already have a contract with this survey provider, so let’s just use that.”
In this organization, part of the workforce indicated that autonomy was important to them, and they also gave a score on a 1-to-10 scale. That sounds like useful information. But when they looked more closely at the open-ended responses, it became clear that employees had very different interpretations of autonomy. One person’s score was not comparable to another person’s score. One answer was not comparable to another answer. A response like “more freedom” is not the same as “clear agreements up front.”
So you cannot conclude that there is broad support for “more autonomy.” They used genAI to summarize the open comments. This is exactly where things often go wrong: several employees said something positive about autonomy, and genAI turned that into an overly confident management message. Autonomy, according to the summary, would be “a requirement for working with pleasure.”
That is not only careless, but also risky. If HR or leaders then use that inflated interpretation to shape policy—which they were about to do—it can lead to more uncertainty, more pressure to make choices, and sometimes even less effectiveness. Autonomy matters, but only when it fits the task, the context, and employee needs. And only when there is real support for it. HR and management encountered resistance. A lot of resistance. Later in this case, we return to the mistake of using a working group made up of the usual suspects, and why pulse surveys can make things worse, especially when followed by genAI interpretation or even recommendations to management.
What Autonomy Does
Research shows that autonomy is often positively associated with motivation, engagement, creativity, and job satisfaction. Employees feel more ownership when they have influence over how they do their work. That can contribute to better performance and less stress, especially when workloads are high.
Autonomy can also be valuable from a learning and development perspective. When people have room to make choices, they tend to think more actively, take initiative more quickly, and learn more from experimentation. In that sense, autonomy can be a condition for innovation.
But that does not mean that more is always better. Like many other topics—such as responsibility, education, and development—autonomy is highly context-dependent and depends on the people involved.
When Autonomy Goes Too Far
Autonomy can also go too far. Especially when employees are given many choices without clear boundaries, priorities, or support, freedom can turn into burden. Then uncertainty, decision pressure, and the feeling that you have to figure everything out yourself can emerge.
Research shows that too much autonomy can, under some circumstances, lead to mental fatigue. Team performance can also suffer if autonomy is expanded on multiple fronts at once—for example, when teams are allowed to decide both who they work with and what they work on. In such situations, autonomy becomes more of a burden than a resource.
There is also another important risk: fake autonomy. In that case, it looks like employees have a lot of freedom, but in reality the direction has already been decided. They are given responsibility, but very little real influence. That is often more frustrating than clear direction.
What Happens Next
If HR and managers want to increase autonomy, they need to look beyond the idea of simply “letting go.” Autonomy only works well when the organization also has the right conditions in place. That means clear goals, clear expectations, enough resources, good collaboration, and leaders who do not micromanage but remain available.
Without those conditions, autonomy can become extra individual responsibility without organizational support. Employees then have to prioritize, coordinate, and decide for themselves, while the complexity of the work remains just as high. Instead of creating more ownership, that often leads to more stress—and possibly less of the joy that first drew people to the organization. Engagement and enthusiasm can suffer; autonomy can become counterproductive.
For HR and leaders, the real question is therefore not: how do we give everyone more autonomy? The better question is: where does autonomy add value, and where would more structure help? And this should continue to be monitored and adjusted based on changing circumstances and employee needs.
Hidden Costs
Increasing autonomy also comes at a cost—not just in money, but especially in time and organizational effort. Work processes often need to be redesigned, roles clarified, agreements made, and managers trained in a different way of leading. It also requires monitoring: does the chosen form of autonomy actually work, or does it create more confusion?
Another issue is that autonomy does not mean the same thing to everyone. One person wants freedom in planning, another wants room in execution, and someone else may actually benefit most from clarity and predictability. If organizations ignore those differences, they assume a one-size-fits-all model of human behavior that does not reflect reality.
What Works Better
A smarter approach is to work with autonomy in a differentiated way. Not everyone needs the same level of freedom. It is better to ask three questions:
• Where do employees want and are they able to handle more room?
• Where does more autonomy actually improve collaboration or quality?
• Where is more structure, alignment, or prioritization needed?
That leads to a more realistic approach. For example, give people room in how they do the work, but keep goals and boundaries clear. Let teams help think through solutions, but make decision-making and responsibilities explicit. Build autonomy gradually, through small experiments, and evaluate the effects.
The Core Idea
Autonomy is valuable, but it is not sacred. It is not a universal recipe for better collaboration or learning, and certainly not proof that employees broadly want more freedom. For some people and some tasks, extra autonomy is energizing. For others, and in other contexts, clarity and direction work better.
The key is not to maximize autonomy or simply to let go, but to calibrate it carefully. Good leadership does not mean letting go for the sake of letting go. It means choosing the right dynamic mix of space, support, and structure.
A Dialogic Approach
Our dialogic approach is an innovative method that gets back to what people are truly good at and helps them become more open to the decisions that are eventually made. Topics like autonomy are introduced early, with open questions that get to the heart of the matter. People are given time to think, room to interpret, and privacy before responding.
Management and HR chose to work dialogically. Employees were given thinking time, privacy, and the ability to participate on their own schedule. They then learned from people with different views and were able to look at autonomy in a new way. A network of people turns out to be more intelligent and creative than collecting individual opinions in isolation, as with pulse surveys.
Networks consist of individuals who want to, can, and also need to learn from each other. Their responses are therefore not simply taken at face value, added up, or fed into genAI. Instead, their answers were shared with other colleagues in a way that allowed everyone to think more deeply, in an accessible and appealing format. This way, participants learn themselves, and managers and HR can see whether a weak signal is actually broadly shared.
For example, in the first round of our online dialogue, employees said things like:
• “I want to decide for myself when I plan my tasks and in what order I do them.”
• “I want decision-making authority over how I do my work without having to check every step with my manager, who then gets involved even when I don’t think it is necessary.”
• “I want to feel more trust from the department, from my colleagues, but especially from our department head. I really need that trust.”
In the second round, where participants read one another’s responses and learned from them, it turned out that they most strongly supported the third statement—the one mentioned least often. Neither genAI nor HR nor management could have inferred that on their own, but the network of colleagues could. In the dialogue, participants were also asked to offer concrete suggestions for action. That made them think more deeply about what is possible, what is needed, and what they themselves can own.
As a result, this organization decided to focus—within that department—on building more trust, because that was where support truly existed. They used the suggestions colleagues had made, such as asking fewer controlling questions and coaching more as managers.
This prevented them from pushing for more—and potentially too much—autonomy, along with the related changes, costs, and pressure on employees. The strength of dialogue is that it reveals what might otherwise remain unseen or be interpreted incorrectly.
