The questions started arriving forty-eight hours before the event. Submit anonymously through the platform. Leadership would answer everything. That was the promise. On the day, 1,200 employees joined. The CEO was on a live stream from the headquarters atrium, flanked by two VPs and a head of HR who smiled at intervals that felt pre-scheduled. A moderator read questions from a queue only the organizers could see. The session ran seventy-three minutes. Fourteen questions were answered. Nobody asked why those fourteen were selected. Nobody asked what happened to the rest.
The post-event survey went out two days later. 68% said they felt "informed." 61% said they felt "heard." The comms team circulated these numbers internally as evidence that the town hall worked. It did not work. It performed working. There is a difference, and most organizations have trained themselves not to notice it.
The town hall, in its current corporate form, is not a communication mechanism. It is a pressure-release valve. It exists to reduce the temperature of a workforce that has been asking questions through unofficial channels, to give leadership visible proof of accessibility, and to create a paper trail showing that dialogue happened. The fact that the dialogue was curated, compressed, and controlled is a detail that rarely makes it into the post-event report. This is not a cynical observation. It is a structural one. The town hall as an institution has specific design constraints that make genuine communication almost impossible: fourteen hundred people and a sixty-minute window, a microphone that only leadership holds, questions filtered by a team whose job is to protect the session from becoming uncomfortable. These are not accidental features. They are the architecture. And the architecture produces exactly the kind of conversation the architecture was designed to produce.
The problem is that organizations keep evaluating the town hall by asking the wrong question. They ask: did employees feel heard? The actual question is: did anything change because of what employees said? The answer to the second question is almost never yes. And if it were, the organization would not need another town hall in six months to cover the same ground.
The question queue is the mechanism that makes the town hall safe for leadership and useless for everyone else. The questions that make it through the filter share a common trait: they are answerable without discomfort. They have a talking point waiting for them. They can be redirected into an announcement that was already planned. The questions that don't make it through are the ones that actually matter. The question about why the restructuring happened the week after the values workshop. The question about whether the VP who left "to pursue other opportunities" was pushed. The question about what the new performance metrics actually mean for the people whose roles are adjacent to the ones being eliminated. These questions don't disappear when they're removed from the queue. They move to the corridor. They move to the WhatsApp group. They move to the exit interview, where someone finally says what they've been carrying for eight months. The filtering problem is not about bad intent. Most moderators genuinely believe they are choosing the most representative questions. What they are actually doing is choosing the most manageable ones. Those are not the same set.
After the town hall, the engagement survey lands. The questions are structured to confirm participation, not to measure impact. Did you attend? Did you feel the leadership team communicated clearly? Do you feel your questions were addressed? Notice what is not there. Did you learn anything you didn't already know? Did anything said today change how you think about the organization? Do you believe the next quarter will look different because of this conversation? The survey is not measuring communication. It is measuring the appearance of communication. And because the questions are structured that way, the scores consistently show that town halls work. Which is why organizations keep running them. Which is why the same workforce that attended the town hall in January is asking the same questions in June. The data creates a loop. The loop creates confidence. The confidence prevents the organization from examining what the town hall is actually failing to do. This is not a measurement problem. It is a design problem that hides behind measurement.
Here is what happens in the ninety minutes before a well-run town hall. The CEO is prepped by the comms team. The key messages are agreed. The three questions most likely to be difficult have been workshopped with legal, HR, and the CFO. The presentation has been reviewed by three people whose job is to make sure nothing in it creates liability or triggers a reaction that can't be managed. Here is what happens in the ninety minutes before the same town hall, for the employee. Nothing. The employee shows up with whatever is on their mind and whatever they've been carrying since the last all-hands. They have no preparation, no advisors, and no guarantee that the question they've been building up the courage to ask will survive the filtering process. One side of the conversation has been engineered for control. The other side has been left to manage itself. And then the organization measures whether the conversation felt balanced. It never will. Because it isn't.
Attrition spikes occur not after bad news, but after bad news delivered badly. The town hall, in its current form, is one of the most reliable engines of the second kind of disengagement.
Organizations calculate the cost of a town hall in logistics terms: the AV setup, the streaming platform, the moderator's time, the preparation hours from the leadership team. They rarely calculate the cost of what the town hall displaces and what it fails to produce. When an employee submits a question that doesn't get answered, they don't simply move on. They draw a conclusion. The conclusion is that the question was too uncomfortable to address publicly, which means leadership either doesn't have an answer or doesn't want to give one. Both of these conclusions are corrosive. The first signals incompetence. The second signals distrust. Neither is what the town hall was intended to communicate. Multiply that conclusion across the percentage of employees whose questions were filtered out. In a 1,200-person town hall with fourteen answered questions, that is a substantial majority of the people who showed up and participated in good faith. Each of them walked away with a version of the same message: this organization performs transparency. It does not practice it.
Communication that changes behavior is built differently than communication that manages perception. The distinction is not about tone or frequency. It is about structure. Genuine organizational communication requires several things the town hall cannot provide. Specificity: the ability to address a question in enough detail that the person asking it has something actionable. Accountability: a mechanism by which the answer can be tested against reality six months later. Two-directionality: a design where the information flow is genuinely bidirectional, not cosmetically so. Continuity: the ability to pick up a conversation that started in January and continue it in March, rather than treating every session as if nothing was said before. None of these requirements are compatible with the all-hands format. They require smaller rooms. Longer cycles. Leaders who are not protected by a filter between themselves and the workforce. They require a tolerance for discomfort that most communication architectures are specifically designed to avoid.
Organizations that run effective town halls are rare. When you find one, the thing that distinguishes it is not the production quality or the charismatic CEO. It is that leadership genuinely does not know what questions are coming. Because the filter doesn't exist. Because the discomfort has been accepted as the cost of real communication. That acceptance is harder to build than any AV setup. It requires leaders who can sit with an answer they don't have yet and say so in front of 1,200 people. It requires an HR function that has stopped protecting leadership from the workforce and started building the architecture for them to meet. It requires metrics that measure change, not satisfaction. SSUNDAR works with organizations that have made this shift: remove the filter, shrink the room, extend the conversation, and measure not whether employees feel heard but whether the organization actually changed. The difference between those two measurements is where the real work starts.
Most organizations will run their next town hall the same way they ran the last one, score it the same way, and report the same results. The questions that mattered most will never leave the queue. The workforce already knows this. The workforce adapted to it years ago. The only people who don't know it are the ones reviewing the 68% favorability score and planning the next session.
Fluency in the language of listening is not the same as having ears.