One-on-One Meeting

Whose meeting is a one-on-one?

The report's, and saying that out loud changes the meeting more than any structural change to it. Left unstated, the default takes over: the manager arrives with what they need, the report answers questions, and the time is spent transferring information upward. That is a useful meeting, but it is the manager's meeting, and it will not surface the things the report has not decided to raise yet. Handing over the agenda is a small act with a long tail. The report brings the topics, the manager brings attention and whatever authority the topics require, and the silence when the report has nothing prepared is left alone rather than filled. Managers find that silence uncomfortable and fill it out of reflex, which quietly takes the meeting back.

What does not belong in a one-on-one?

Three things, each of which has a cheaper home. Status belongs in whatever written channel the team already uses, because a verbal walkthrough of a board both people can read consumes the whole meeting and tells neither of them anything new. Evaluation belongs in the appraisal, and importing it turns the recurring conversation into a rolling assessment the report has to manage rather than use. The third is the decision the manager has been avoiding. A one-on-one is a convenient place to hint at a concern without naming it, and hinting feels like progress to the person doing it and like weather to the person receiving it. If a decision is needed, it needs a moment of its own with a clear beginning, not a soft landing inside a standing meeting the report expected to be theirs.

What does a canceled one-on-one communicate?

That it is the least important thing on the manager's calendar, and the report will draw that conclusion whether or not it is true. Every calendar contains a hierarchy of what gets moved first, and the recurring internal meeting with one person is the easiest thing in it to move, because only one person is inconvenienced and that person reports to whoever is moving it. They will not say anything the first several times. What happens instead is that they stop bringing the topics that were only worth raising in a conversation they trusted would happen, and the meeting, when it eventually runs, has nothing much in it, which confirms the manager's sense that it was not urgent. Reliability is doing more work here than content. A short conversation that always happens will hear things a long one that often slips never will.

How do you choose a cadence without copying someone else's?

By reasoning from what the meeting has to carry, not from what a neighboring team does. The longer the gap between conversations, the more each one has to hold, and the more likely something time-sensitive waits for a slot. The shorter the gap, the less has accumulated, and the meeting starts to feel like an obligation with nothing in it, which is how a cadence gets abandoned. Where the right point sits depends on the pair: how new the person is to the work, how much of their week is ambiguous rather than defined. A person deep in familiar work and a person newly into a role they have never done do not need the same rhythm, and giving them the same one because the calendar is tidier is the mistake.

Whatever the interval, the thing to protect is that it holds. A slot that survives a busy week is worth more than a longer one that yields to the first conflict, because the report calibrates what to bring against how confident they are that it will happen. There is a real difference between rescheduling and canceling: moving a conversation says it still matters, while dropping it says the week absorbed it. Where the calendar will not hold, the honest response is to change the standing arrangement rather than keep a commitment that is broken in practice and intact on paper. Managers with many direct reports hit this first, and usually try to solve it by shortening each conversation. That works for a while, then stops, and the constraint being hit is span rather than scheduling.

What has to be written down for the meeting to compound?

Only two things, reliably: what was said that either person needs to remember, and what either of them committed to doing about it. Without that, each conversation starts from nothing, and the pair rediscover the same unresolved topic without noticing they are repeating themselves. A running document, appended to rather than replaced, fixes this at almost no cost. A manager who says they will look into a blocked approval and writes it down where the report can see it has made a different kind of promise than one who says it and moves on. Carrying open items forward is the whole mechanism: when something has appeared unresolved across several conversations, the document says so plainly, and that is usually when it gets escalated or honestly abandoned.

Where the document lives matters more than what it is. A note kept privately is a management aid; a note both people can open is a shared record, and only the second lets the report hold the manager to something. Teams that keep these inside [performance management software](/performance-management-software) get one side effect for free: the trail survives a change of manager, which is when it is most likely to be lost. A report who explained their situation to a departing manager has to explain it again from the beginning, and the second telling is shorter and less candid. The shared record should hold commitments and topics, not the manager's private assessment of the person; a document that becomes an evidence file changes what the report will say in front of it.

What breaks first when a manager has more people than attention?

Span is the constraint that decides whether any of this is available. A manager holding a small number of reports can carry each person's context in their head between conversations, which is what makes the meeting feel continuous rather than episodic. Past some point that stops, and the manager arrives having forgotten the thread, asks a question they asked last time, and the report notices. The usual response is to compress: shorter conversations, more of them back to back, notes taken in a hurry. What actually degrades is not the length but the preparation, because preparation is the part with no meeting in the calendar defending it. A wide span is an organizational decision with a predictable cost, and one of the things it costs is this meeting.

Distance changes what the meeting has to do rather than how it is run. In a shared office a manager picks up a great deal without asking, and the standing conversation only has to cover what that ambient signal missed. Remove the office and the ambient channel disappears, so the meeting becomes the primary place anything is noticed. Asynchronous teams face the sharper version, where the pair can go a long stretch exchanging text without either saying anything they would not put in writing. Written channels are excellent at conveying status and poor at conveying that someone has quietly decided to leave. Keeping the calendar, the reporting line and the record in one place, which is what an [HRMS](/hrms) is for, removes the ambiguity about who is meant to be having the conversation.

What does the meeting look like when the pair did not choose each other?

Most manager and report pairs are assembled rather than chosen, and an inherited report is the ordinary case rather than the exception. The manager arrives holding a written record they did not write, sometimes an appraisal they did not conduct, and an account of the person from a predecessor whose standards they cannot see. The report has to re-establish something they had already built once. The first several conversations are better spent on what the manager does not know than on what they intend to change: what this person has been promised, what they were working toward, which parts of their situation the previous manager was handling informally and are now unowned. An inherited judgment acted on before forming your own is how a reputation outlives everyone who could explain it.

A skip-level conversation runs on different terms and should be described as such before it happens. The senior manager is not the person's manager, cannot resolve most of what is raised without going through the manager in between, and holds an authority that makes candor expensive for the report. Presenting it as just another one-on-one invites the report to say something they will regret when it travels. Said plainly, it works: this is about how the wider organization looks from where you sit, and here is what I will and will not repeat. Comparing whether these conversations are happening across a large organization is the kind of question [performance management tools](/best-performance-management-software) can answer, though the answer is attendance rather than quality, and no system measures whether the thing that mattered was said.