How to Run a One-to-One That Is Worth the Hour

TL;DR

• The core decision: whose meeting it is, because that single choice determines what the conversation can contain.
• When doing nothing is right: when your reports raise difficult things unprompted and you regularly leave the meeting with something you didn't know.
• What has to be true: the status information has somewhere else to live, or it will colonise the hour by default.
• How the options split: by who owns the agenda and whether the format makes room for things that take a while to say.
• Decision rule: if the meeting could be replaced by a written update with no loss, it's currently a written update with worse economics.
• Outcome to expect: a few uncomfortable conversations you'd otherwise have had much later, or not at all.

The Meeting Both People Would Cancel

A manager and one of their reports meet every week. It's been in the calendar for a year. The report gives an update on what they're working on, the manager asks a few clarifying questions, they cover anything outstanding, and it wraps up early. Both people would describe it as fine.

Both would also, quietly, be relieved to see it cancelled. The manager has the information already, mostly from the project channel. The report is repeating things they've written down elsewhere. Neither would say the meeting is useless, because it isn't unpleasant, and because cancelling it would look like something. So it continues, in a form that costs an hour a week and produces almost nothing.

The pattern is close to universal. One-to-ones are the most widely held management ritual there is, and most of them have drifted into the one function they're worst at: transmitting status. Status wants to be written. It's factual, it benefits from precision, it needs to reach more than one person, and it doesn't require anybody's tone of voice. Putting it in a meeting means paying a conversational price for something that would have been better as text.

Recommended Tools

Best tools for Leadership & Culture

C Culture Amp Culture measurement L Lattice Leadership development L Leapsome People enablement platform

View all recommended tools

The real problem isn't that the meeting is badly run. It's that the meeting has filled up with the easy thing. There's a conversation that has no other home, about how the work is actually going, what somebody is worried about, whether they want to still be doing this in a year, and it only happens if the status stuff is out of the way. The whole design question is how to keep the easy thing from eating the hour.

When You Genuinely Do Not Need to Act Yet

Your current setup is genuinely fine. Your reports bring things you didn't know, including things they're not proud of, and you regularly leave with something to think about. If that's happening, the format is working and you should be suspicious of any advice to restructure it. This is one of the areas where a bad change is easy to make and hard to detect.

Friction is starting to show. The meeting is getting moved, or ending early, or you've noticed the conversation is the same every week. Any of those is drift rather than crisis, and the cheapest diagnostic is to ask what was in the last three of these that wouldn't have been in a written update. If the honest answer is nothing, you've found it.

It has become a real cost. You're learning about problems from somewhere else, or a resignation surprised you, or somebody has been unhappy for a quarter and you didn't know. These meetings are the designated place for that information to surface, so when it surfaces elsewhere the hour has been spent on something else. At this point changing the format is worth a slightly awkward conversation about why.

The edge case that forces it. Somebody new joined, or somebody moved into a role where they'll struggle for a while, or the team went remote. All three change what the meeting needs to do: a new person needs more frequent and more structured contact, and a distributed team has lost all the incidental conversation that used to carry the things nobody schedules.

Five Questions This Reader Asks at 11pm

How often should these happen? Weekly for most people, fortnightly for experienced people in a steady period, and more often for somebody new or somebody struggling. The cost of too frequent is a meeting with nothing in it, which is recoverable. The cost of too infrequent is that anything difficult waits, and by the time the slot arrives the person has either solved it alone or stopped caring about it.

Who sets the agenda? The other person, with you adding to it rather than owning it. This is the single most useful change available in this subject and it's nearly free. A manager-owned agenda produces the conversation the manager already knew they wanted to have, which is by definition not the conversation that would have told them something new.

What if they have nothing to raise? Common, particularly at first, and it usually means either that the format has trained them out of it or that they don't believe the answer will change anything. Sitting in the silence is uncomfortable and it's the right move. The wrong move is filling it with your own agenda, which teaches them that not preparing is fine because you'll take over.

Should I take notes? Yes, and visibly, because a person watching you write something down draws a reasonable conclusion about whether it's going anywhere. The thing to record is what you committed to, not what they said. A manager who returns to the last item of the previous meeting at the start of the next one has done more for the quality of these conversations than any question list would.

Can I cancel when there's nothing urgent? You can, and the pattern of cancellations is more informative than the meetings. If the ones you cancel are always with the same people, you've built a system where the quiet, self-sufficient person gets the least of you, and that's usually the person whose departure is most surprising. Cancelling occasionally is fine. Cancelling predictably is a decision you're making without noticing.

Three Honest Categories the Approaches Split Into

Manager-led, with a set agenda. You bring the topics, run the conversation, and cover what you need to cover. It's right when there's genuinely something to work through, when the person is new and doesn't yet know what these are for, and in any situation where a specific message needs delivering. It fails as a default because it produces the conversation you already anticipated, and the entire value of the meeting is in the part you didn't anticipate. It also fails quietly: people learn to prepare nothing, since preparing is pointless when the hour is already allocated.

Report-led, with an open agenda. The other person decides what's discussed and you respond. It's right in most steady-state relationships and it's the only format that reliably surfaces things you didn't know to ask about. It also shifts the ownership of the relationship, which is worth something in itself. It fails when the person doesn't yet know what's appropriate to raise, which is common early and with anybody who's been managed badly before, and it fails when the manager treats an open agenda as a licence to arrive unprepared, which is visible immediately.

Structured around a fixed question set. The same handful of questions every time, working through the same ground. It's right when a manager has many reports and needs consistency, when the manager is new to managing and needs a scaffold, and where comparability matters. It fails on depth, because questions asked identically every week get answered identically, and both people learn the ritual. It also crowds out the unscheduled thing: a fixed set has no slot for the thing somebody has been working up to saying for a fortnight.

Five Diagnostic Questions You Can Self-Assess Against

What was in the last three that a written update wouldn't have covered? Be strict. If the answer is nothing across all three, the meeting has become a spoken document and both of you know it. This is the fastest and least comfortable test in this article, and it's the one worth running first.

Who cancels, and with whom? Look at the last two months. Cancellations cluster around the people who cause the least friction, which means your calmest and most self-sufficient reports systematically get less of your attention. That's the opposite of a deliberate allocation, and nobody ever decided it.

Do they arrive with anything? If the other person shows up with no agenda, week after week, that's information about the format rather than about them. Either they've learned you'll fill the time, or they've learned that what they raise doesn't lead anywhere. Both are fixable and neither is fixed by asking them to prepare more.

What have you changed because of one of these conversations? Something should have. If you can't name an instance from the last few months, the meeting is functioning as a listening exercise with no output, and the other person has almost certainly noticed before you did.

When did they last tell you something difficult? Not a problem with the work, something difficult about the work for them: that they're bored, that they're struggling, that they disagreed with a decision. If that's never happened, this meeting isn't doing the job that justifies its existence, whatever else it's producing.

Six One-to-One Formats, Reviewed

The status update

The report walks through what they're working on and the manager asks questions. It earns its place in genuinely fast-moving situations where written channels lag, and with a very new person where the manager needs frequent visibility until trust is established. It's also easy, which is not nothing: it requires no preparation from either side and never feels awkward.

Where it falls short is that it consumes the whole hour with the one thing that could have been text. Status is precise, it benefits from being recorded, and it usually needs to reach more people than the two in the room, so a conversation is a poor container for it. The deeper cost is opportunity: this is the only guaranteed private hour between the two of you, and a status format spends it on the most transferable content available.

If your meeting has become this, the fix isn't a better agenda. It's putting the status somewhere written and then sitting with the silence that follows.

One thing worth saying in defence of the format, because the case against it is easy to overstate. Talking through the work does produce incidental discoveries: a manager asks a clarifying question, the answer reveals an assumption nobody had examined, and something useful comes out of it. That's real. It's just an expensive way to buy it, and the same discoveries arrive more reliably from reading a written update carefully and bringing the two questions it raised.

The open agenda owned by the report

The other person brings what they want to discuss and you follow. It earns its place as the default for most established relationships, because it's the only format that produces the things you didn't know to ask about. It also communicates something real about whose career this is, which changes how people use the time over months.

It falls short at the start, and with anybody whose previous manager ran a different regime. Asked to own an agenda with no guidance, most people bring status, because status is what's obviously legitimate. It also fails when the manager reads open as unprepared: turning up with nothing while asking the other person to turn up with something is a visible asymmetry.

The version that works has a short explanation attached. Say what the meeting is for, give examples of what's welcome, and be explicit that it's not a status check. Then prepare your own contribution anyway, and add it after theirs.

The standing question set

Three or four consistent questions asked every time: what's going well, what's blocking you, what do you need from me, and something forward-looking. It earns its place for a manager with many reports who needs some consistency, and for anybody new to managing who'd otherwise be improvising under pressure. It also guarantees a floor: the important ground gets covered even in a bad week.

It falls short because repetition dulls it. The same question in the same order every week becomes a ritual both people perform, and the answers get shorter over time rather than deeper. It's also poorly suited to the conversation that matters most, which is the one somebody has been building up to, since a structured round gives that no natural opening.

Rotate the questions, or keep two fixed and vary the rest. Better still, treat the set as a fallback for weeks where nothing has been brought rather than as the format itself.

There's one situation where a fixed set genuinely outperforms an open agenda, and it's worth knowing. When a manager has a large number of reports and limited time, consistency protects the people who ask for least. An open agenda rewards whoever is most comfortable raising things, and a structured round gives the quieter person the same prompts as everybody else. If your team is large, the fairness argument may outweigh the depth argument.

The walk with no notes

You have the conversation while walking, or over coffee away from the desk, with nothing written down. It earns its place because removing the notebook and the screen changes what people say. Difficult things are easier to raise side by side than across a table, and the absence of a record is precisely what makes some conversations possible.

It falls short on follow-through and on fairness. Nothing is recorded, so commitments made in the conversation are more likely to evaporate, and the format suits some people considerably more than others. It's also a poor default in a distributed team, and treating it as a treat for the reports you enjoy talking to creates a visible difference in access.

Use it occasionally and deliberately, particularly when you can tell something is unsaid. Write down what you committed to afterwards, even if you wrote nothing at the time.

The fortnightly long session

Instead of thirty minutes weekly, an hour every two weeks. It earns its place with experienced people in steady periods, where the weekly slot has become a check-in nobody needs. The extra length matters more than it sounds, because the things people find hard to say tend to arrive late in a conversation, after the easy material has been cleared.

It falls short when things move quickly, when somebody is struggling, or when the relationship is new, since a fortnight is a long time for a problem to sit. It also raises the stakes on each meeting: if one is cancelled, a month passes, and the person who had something to raise has had to hold it.

It's a reasonable setting for the right person at the right time, and worth revisiting whenever the situation changes rather than treating as permanent.

Cancelling them in favour of working together

You drop the standing meeting and rely on working side by side on something real, plus conversation as it arises. It earns its place in a small, closely-working team where you're genuinely in contact daily, and it avoids the specific failure of a ritual that's stopped meaning anything.

The limitation is that it depends entirely on the things that are easiest to defer. Ad hoc conversation covers the urgent and skips the awkward, and the topics that need a scheduled slot are exactly the ones nobody raises spontaneously: dissatisfaction, ambition, a slow-building problem. It also concentrates access on whoever is most comfortable interrupting you, which is not a neutral filter.

If you go this way, keep something periodic and protected, even if it's monthly. The scheduling is the point, not the frequency.

It's also worth being honest about why this option appeals. Cancelling a meeting that has stopped meaning anything is a reasonable response to a real problem, and it's frequently chosen by managers who find the conversation itself uncomfortable. Those are different motivations with the same outcome, and only one of them is a decision. If the prospect of half an hour with nothing scheduled to discuss is the part you'd most like to avoid, that's a reason to keep the slot rather than to drop it.

The Decision Table

The second row is where most one-to-ones actually sit, and the fix has an uncomfortable middle step. Once the status has moved to a written update, the first two or three meetings feel empty, and the natural response is to conclude the meeting isn't needed. That emptiness is the space the useful conversation eventually fills, and abandoning the change at that point is the most common way this improvement fails.

What Belongs in the One-to-One and What Does Not

Most of the difficulty here is a placement problem rather than a conversation problem. Some topics belong in this meeting, several don't, and the ones that don't are the ones that fill it.

The fourth row causes more avoidable damage than any other. Feedback held back and delivered in the one-to-one because that's where feedback goes arrives stale, and a collection of small stale items feels to the recipient like a case being built. Say it when it happens, and use this meeting for the pattern rather than the instance.

The last row is the reason the meeting needs to be private and predictable even in weeks where nothing seems to be happening. A person dealing with something difficult outside work will not create a meeting to tell you. They'll mention it in one that already exists, usually near the end, usually after saying that everything's fine.

When the Conversation Stays on the Surface

Some one-to-ones never get past the pleasant and the factual, week after week. It's the most common complaint managers have about this meeting, and the usual explanation, that the person isn't very forthcoming, is usually wrong.

Consider first whether anything has ever come of these conversations. People calibrate quickly: if two or three things were raised and nothing visible followed, the sensible conclusion is that this meeting is for reassurance rather than for change. The repair isn't a better question, it's closing the loop on something old, out loud, including the ones where the answer was no.

Then consider what you do with discomfort. The moment somebody says something mildly critical or admits a struggle, there's a reflex to reassure, to explain the context, or to move to solving. All three end the disclosure. Sitting with it for a few seconds and asking one more question about it is the single most useful habit in this meeting, and it feels unnatural for a long time.

Check the setting, too. Some people will not say a difficult thing across a desk in a glass room with the next meeting visible in the calendar. Changing where and how the conversation happens sometimes changes what's in it, and it costs nothing to test.

Then ask directly, and accept a delay. A question along the lines of what you'd be reluctant to tell me, asked plainly and followed by silence, occasionally produces an answer immediately. More often it produces one three weeks later, because it establishes that the question can be asked. The manager who asks it once and concludes it doesn't work has misjudged the timescale.

And allow for the possibility that things are genuinely fine. Not every quiet report is suppressing something. The tell is whether they've ever raised anything difficult, at any point: somebody who has and currently hasn't is probably fine, and somebody who never has, across a year, probably isn't.

What to Put in Writing

Almost nothing about this meeting should be formalised, and the little that should is about continuity rather than record-keeping.

The second row is the one that carries the meeting. Opening with what you said you'd do last time, and saying plainly when you didn't do it, changes the character of the conversation more than any question you could ask.

Questions to Ask Before You Commit

On content. What was in the last three that a written update wouldn't have covered? A bad answer is that the conversation is valuable in itself.

On ownership. Whose agenda is it, and how would they answer that question? A bad answer is that it's shared.

On cancellations. Who do you cancel on most? A bad answer is that it's random.

On output. What changed because of one of these? A bad answer names something already in motion.

On difficulty. When did they last tell you something hard? A bad answer is that they would if there were something.

On placement. Is there feedback you're saving for the next one? A bad answer is yes.

What Getting This Wrong Costs

The first cost is the obvious one and it's the least important: an hour a week per person, spent on something that could have been a document. That's real, particularly for a manager with a large team, but it's the cost people reach for precisely because it's easy to quantify, and it isn't why this matters.

The second cost is that you lose your only reliable channel for information that doesn't travel any other way. Project status will reach you regardless, through channels, reports and escalation. Whether somebody is bored, or has quietly decided to leave, or thinks a decision you made was wrong but doesn't want to say so in a meeting, reaches you only here or not at all. A one-to-one that's become a status update hasn't just wasted an hour, it's closed the channel, and you'll find out what was in it from a resignation letter.

The third cost accumulates in the relationship. These meetings are where somebody either learns that raising things with you produces change, or learns that it doesn't. Every instance where something was mentioned and nothing followed is a small piece of evidence, and they compound into a settled belief that's much harder to shift than the original neutrality. By the time a manager notices that their reports have stopped bringing things, the pattern usually has a year behind it.

So before you change the format, work out which problem you have. A content problem means the hour is full of the wrong material, and moving status to writing fixes it. An ownership problem means you're running the conversation you expected rather than the one available, and handing over the agenda fixes it. A credibility problem means things have been raised and nothing happened, and no format change addresses that at all, because what's needed is to close an old loop and be seen doing it.

When You Are Ready to Go Further

Nothing here needs a tool or a template. It needs status moved somewhere written, the agenda handed to the other person, and a note of what you committed to so you can open with it next time.

The step beyond your own practice is consistency. When each manager runs these differently, two people doing the same job in different teams have completely different access to their manager's attention, and neither knows that the other's experience differs. That's not a meeting-design question any more, it's a question about what the organisation expects of a manager and how it checks.

HROpsLab publishes independent comparison work across HR tooling, applicant tracking and payroll. We sell nothing, we take no vendor money, and we publish no paid placements. If the next step is looking at how consistently your managers do this, our comparison work is one place to start.

Frequently Asked Questions

If the meeting could be replaced by a document with no loss, it already has been.