The alternative I keep coming back to is a small-group feedback session: two or three people at a time, not a full-team meeting and not a private skip-level interrogation. It is intimate enough for texture, but not so intimate that every comment feels like a personal referendum on one employee's relationship with leadership.

The format is not magic. If it is sloppy, it becomes worse than doing nothing. People will read the room, give safe answers, and wait to see whether anything changes. But if it is designed well, it can give a leader a better picture of the system than a long string of isolated 1:1s.

What skip-levels are trying to solve

I think it helps to name the job before choosing the format. Skip-levels usually try to answer a few different questions at once:

  • Are people hearing the company or team direction clearly?
  • Are there problems that are not making it through the normal management chain?
  • Are managers creating clarity, trust, and momentum?
  • Are there small frustrations that have become operating drag?
  • Are people still saying the true thing out loud, or only the approved thing?

Those are not all the same question. A survey can help with trend detection. Office hours can help with access. A town hall can help with direction. Mentorship can help with growth. A small group is strongest when the goal is texture: how people describe the actual working system when they are close enough to compare notes.

The case for two or three people

A full-team session has too much gravity. The most confident voices define the edges of the conversation, and everyone else starts doing political math. A one-on-one can go deeper, but it also puts a lot of pressure on the person speaking. Two or three people creates a useful middle layer.

Format What it gives you Where it breaks
Skip-level 1:1 Depth, privacy, personal nuance Slow to scale, easy to over-index on a single account, socially high-stakes
Full-team feedback session Shared context and visible listening Can become performance, complaint theater, or silence
Anonymous survey Pattern detection and safer negative signal Weak on texture, follow-up, and root cause
Two- or three-person feedback session Enough safety for candor, enough social calibration to avoid one-person distortion Requires careful pairing, facilitation, and follow-through

The small group works because people can build on each other. One person says, "The handoff is unclear," and another can add, "Yes, but only when the priority changes after planning." That second sentence is the gold. It turns a vague complaint into an operating condition.

How I would structure it

I would keep the session short and explicit. Forty-five minutes is usually enough. The goal is not to solve everything in the room. The goal is to understand where the system is generating friction.

  1. Pick one objective for the round: planning clarity, team communication, manager support, cross-team handoffs, execution drag, or morale signals.
  2. Invite two or three people with enough shared context to compare notes, but not so much hierarchy or social pressure that one person dominates.
  3. Send the prompt ahead of time so the meeting does not reward whoever improvises fastest.
  4. Open by naming the rules: this is not a performance review, it is not a complaint booth, and quotes will not be attributed without permission.
  5. Ask a small number of repeated questions across sessions so patterns can be compared.
  6. Capture themes, not gossip.
  7. Close with what will happen next and what will not happen next.

That last line matters. Feedback processes lose credibility when leaders collect input as if every observation creates an action item. Some feedback should lead to change. Some should clarify a tradeoff. Some should simply make leadership more aware. But the participants should not be left guessing whether the conversation disappeared into a notebook.

The questions I would ask

I would avoid starting with "how is everything going?" It is too broad. It gives people permission to answer at the level of mood. Better questions aim at observable work.

  • Where does work most often slow down right now?
  • What decisions are people waiting on too long?
  • What information do you wish you had earlier?
  • Where are expectations clearest?
  • Where are expectations most ambiguous?
  • What is one thing leadership might not realize is costing the team time?
  • What is working better than it used to?

The positive question belongs in the set. Not because every feedback session needs a forced upbeat ending, but because the goal is accuracy. If the process only extracts frustration, it will distort reality too.

What can go wrong

The first failure mode is accidental surveillance. If people feel like the real question is "tell me whether your manager is doing a good job," they will protect themselves. Even when manager feedback is relevant, the better route is usually to ask about operating conditions: clarity, blocked decisions, escalation paths, prioritization, communication, and the places where work gets stuck.

The second failure mode is treating the loudest pattern as the truest one. Small groups are useful, but they are not statistically clean. I would use them as signal, then triangulate with surveys, delivery patterns, attrition risk, manager 1:1 themes, and what I can observe in the work itself.

The third failure mode is no visible loop closure. If people give thoughtful input and nothing visible happens, the organization learns that feedback is ceremony. The follow-up does not have to be dramatic. It can be as simple as: "Here are three themes I heard, here is one thing we are changing, here is one thing we are not changing yet, and here is why."

Where I landed

My current read is that small-group feedback sessions are a good substitute for skip-level 1:1s when the goal is operating signal, not private career coaching. They are especially useful when a leader wants to understand friction across a team without turning every conversation into either a full-team ritual or a sensitive private meeting.

The trick is to make the format modest. Two or three people. A narrow objective. Repeated questions. Careful notes. No attribution theater. Visible follow-up. That is enough structure to make the conversations useful without pretending they are a comprehensive listening system.