How often should 1:1s happen?

Weekly for new reports and anyone in a changing role, fortnightly once things are stable. Monthly is too infrequent to catch a problem while it is still small, which is most of the value a 1:1 provides.

How long should a 1:1 be?

Thirty minutes weekly works for most pairs. Sixty minutes fortnightly is worse than thirty weekly, because the value comes from the frequency of the check-in rather than the total minutes.

Who should own the 1:1 agenda?

Your report. If you own the agenda it becomes a status update, and you will hear about work rather than about obstacles. A shared document they add to during the week works better than either person preparing at the last minute.

Should a 1:1 be cancelled when things are busy?

Reschedule rather than cancel, and never cancel two in a row. A repeatedly cancelled 1:1 tells your report where they rank, and they will stop bringing you problems early, which is exactly when problems are cheap to fix.

What is the difference between a 1:1 and a performance review?

A 1:1 is about the current work and the obstacles in it. A performance review is about a period of work as a whole and about the person's trajectory. Nothing in a review should be a surprise if the 1:1s have been happening.

Do 1:1s work with remote teams?

They matter more. In an office, some of this happens incidentally at a desk. Remotely, if it is not scheduled it does not happen, and the first sign of a problem is often the resignation.

How to Run 1:1s That Produce Decisions, Not Small Talk

Why the 1:1 carries so much weight

The 1:1 is the only recurring meeting where somebody can tell you something is wrong before it becomes visible in the work. That is its whole function. Everything else it gets used for, status, planning, project updates, has a better home.

70%Of the variance in team engagement is accounted for by the manager, according to Gallup's manager research.Gallup, State of the American Manager

That number gets quoted a lot without much attached to it. The practical reading is narrow but useful: the recurring, private conversation between a manager and a report is where most of a manager's influence is actually delivered. If that conversation is a status readout, the influence is not being delivered.

A structure that survives a busy week

The structure below takes thirty minutes and works whether or not either person prepared.

Their agenda first, and give it real time
Fifteen of the thirty minutes. Whatever they want to raise: a blocker, a decision they need, something that is annoying them, a question about direction.
Going first matters. If you lead, they will spend their half responding to your points, and the things they were unsure about raising will not come up at all.
15 minutesThey talk more than you do
Your agenda second
Ten minutes. Feedback, context they are missing, priorities that have shifted, anything you need from them.
Bring at most two things. A manager arriving with a list of six turns the meeting back into a status update by volume alone.
10 minutesTwo items maximum
Write down what was decided
Five minutes, in the meeting, not afterwards. Every decision, who owns it, and by when. Two or three lines is usually enough.
This is the step that separates a 1:1 from a chat. It also means the next 1:1 starts with a check on those lines rather than with both of you trying to remember what happened.
5 minutesIn the meeting, not after

The written record does more work than it looks like it should. It makes the meeting accountable in both directions, it gives you a trail when a performance conversation eventually happens, and it stops the same unresolved issue being raised for the fourth time with neither of you noticing.

Questions that surface real problems

"How is everything going" reliably produces "yeah, good". These do better, because they are specific enough to be answerable honestly.

Question	What it surfaces
What took longer than it should have this week?	Process friction and tooling problems
What are you unsure about that you have not asked?	Missing context, unclear ownership
Is anything blocked and waiting on someone?	Dependencies you can clear today
What would you do differently if it were your call?	Disagreement that is otherwise unvoiced
What is the least useful thing you did this week?	Work that can be stopped
Where do you want to be better in six months?	Development, and retention risk

Two rules make these work. Ask one, not six, or it becomes an interview. And when the answer is uncomfortable, resist explaining why the situation is the way it is. Explaining teaches people that raising things costs them a lecture, and they will stop.

Cadence, length, and what to do when it slips

Weekly for anyone new, anyone whose role is changing, and anyone struggling. Fortnightly for stable, experienced people who prefer it. Thirty minutes in both cases.

Put them at a fixed time and treat that slot as genuinely booked. The most common way 1:1s degrade is not cancellation but drift: moved twice, shortened to fifteen minutes, then merged into a project call.

When you have to move one, move it rather than skipping it, and never miss two consecutively with the same person. If you are managing enough people that this is difficult, that is worth noticing in itself. Beyond seven or eight direct reports, weekly 1:1s at any real quality stop being possible.

Five ways 1:1s go wrong

It becomes a status update. The most common failure. If you can get the same information from a board, a digest, or a project tool, it should not be consuming this slot. Ask about obstacles, not progress.

The manager talks most of the time. A rough target is that they speak for two thirds of it. If you finish the meeting having done most of the talking, you learned nothing you did not already know.

Nothing is ever written down. Then decisions get re-litigated, commitments quietly evaporate, and neither person can remember what was agreed three weeks ago.

Feedback is saved up for the review. Feedback delivered six weeks after the event is about the past. Delivered in the next 1:1, it is about the work.

It is only ever about work. Not every session needs to be personal, but a manager who knows nothing about what their report wants next is guessing about the thing that most affects whether they stay.

The check on all of this is simple. Look back at your last four 1:1s with someone. If you cannot point to a single decision that came out of them, the meeting is not doing its job, and the fix is structural rather than a matter of trying harder.

Frequently asked questions

How often should 1:1s happen?: Weekly for new reports and anyone in a changing role, fortnightly once things are stable. Monthly is too infrequent to catch a problem while it is still small, which is most of the value a 1:1 provides.
How long should a 1:1 be?: Thirty minutes weekly works for most pairs. Sixty minutes fortnightly is worse than thirty weekly, because the value comes from the frequency of the check-in rather than the total minutes.
Who should own the 1:1 agenda?: Your report. If you own the agenda it becomes a status update, and you will hear about work rather than about obstacles. A shared document they add to during the week works better than either person preparing at the last minute.
Should a 1:1 be cancelled when things are busy?: Reschedule rather than cancel, and never cancel two in a row. A repeatedly cancelled 1:1 tells your report where they rank, and they will stop bringing you problems early, which is exactly when problems are cheap to fix.
What is the difference between a 1:1 and a performance review?: A 1:1 is about the current work and the obstacles in it. A performance review is about a period of work as a whole and about the person's trajectory. Nothing in a review should be a surprise if the 1:1s have been happening.
Do 1:1s work with remote teams?: They matter more. In an office, some of this happens incidentally at a desk. Remotely, if it is not scheduled it does not happen, and the first sign of a problem is often the resignation.