One-on-Ones Once a Quarter: Why I Skip Weekly Check-ins and What Works Instead
Frequent 1:1s treat the symptom — a missing channel. Safety treats the cause.
The most honest feedback I ever got from people never once happened in a scheduled 1:1. It came on a Tuesday at 11:40, when someone wrote "got a minute?" — because they knew they could.
Some companies run 1:1s every week or two. I see no point in that, and here is why: at that cadence people inevitably drift into checkbox reports. Saying nothing feels like a failure, so they say something anyway. The feedback stops being valid — at best it is useless, at worst harmful, because it creates the illusion that the channel works.
There is a second, deeper problem. The "are you doing okay? — yeah, fine" format leans on self-reporting. And a sizeable share of people — not only developers, and not only in IT — are in weak contact with themselves and their feelings, cannot always separate the subjective from the objective, and cannot always give you facts instead of emotionally colored opinions. If emotional maturity in a team sits at the population average — and there is no reason to hope for more — I would not risk leaning on feedback collected that way.
So why does everyone run weekly 1:1s
To be fair, the orthodoxy did not appear out of nowhere. Frequent 1:1s are a crutch for environments where a calendar invite is the only safe channel. If a person is not confident they can walk in at any moment, you need forced synchronization, otherwise they never come at all. A weekly slot in the calendar solves that problem honestly.
My system does not "run 1:1s less often". It replaces calendar synchronization with permanent access. Those are fundamentally different things, and the second one has a precondition you cannot skip.
The precondition: safety
Quick fixes get expensive over the long run, so you have to start far upstream. First, the rules of the game: write them down, test them, roll them out (how I do it — in the playbook and in the article on team ideology). Then show by action that it is safe here: nobody rolls their eyes at a genuine wish to share something out of turn, nobody gets punished for the personal or for criticism. Every contact between a manager and a person either confirms this or destroys it — there is no third option.
Once that holds, people understand: any "not" — not working, not on time, not to my liking — as well as any "I want to share something good", can be brought up at any moment, and you will be heard immediately, or at least the process will start immediately. The need for frequent formalized meetings disappears on its own.
When I rolled this out, people came often at first — dumping everything that had piled up for years. Then a lot of it got fixed, and they came less often: fewer reasons. Someone came after a long task with indivisible context — asking for a breather. Someone came with a conflict: the team could not get along with a new hire because of his particular communication style, and my work came down to "equalizing terminology" — showing both sides that there is no aggression, just a different style. All of it outside any schedule, as things came up.
The mechanics: two modes
Times are either calm or turbulent, and the feedback mode differs between them.
Calm times: a quarterly 1:1 about process. At the start of the quarter I remind everyone: collect "deviations" in a personal notebook — agreements, screw-ups, observations on tasks. A week before the meeting I ask them to pull up their notes and prepare. The meeting itself is an hour max, on a simple template: first a self-assessment of load (was there overload, did you like the tasks, what to drop or add next quarter), then intra-team interaction (communication, the quality of the work of the people next to you), and at the end cross-team (who blocks you, who is hard to work with, who regularly hands over raw artifacts).
A quarter is not an arbitrary number. It is the minimum span over which a person accumulates facts rather than the mood of one particular week.
The key part is that the loop closes. Complaints about adjacent teams do not die in my notebook: cases pile up and get taken to a retro between leads (anything critical is handled right away). Then comes a plan approved by the department director, owners, acceptance criteria. And always a return to whoever complained: "work is under way, here is what we are doing, watch it." Then a check: did it get better or not. Without that return, quarterly 1:1s would degenerate into the same checkbox, only rarer.
Turbulent times: mental 1:1s once a month. When it storms, I say so directly. People are not stupid — they notice the tremors in top management themselves, and denying the obvious or covering for leadership would be foolish and would cost me trust. So things get called by their names, and the team switches to monthly meetings of a different format: we barely touch work topics and talk about emotional state — is the person anxious about their own future and the company’s, have they started looking elsewhere.
About "looking elsewhere", separately. When the first warning signs appeared, I asked everyone plainly: if you start looking, tell me first. And it works: there was a case when a person came and said honestly — I updated my resume, I am starting a search. We sat down to talk. I use my contacts on the market to give the person an honest picture of what actually awaits them outside, and from there it is only facts: if something specific does not suit you, we discuss it and adjust it; if you are tired, take a rest. I kept that person, they still work with me. Note: this is only possible because they came to talk instead of leaving silently one day. Honesty in both directions is a direct consequence of safety.
Real feedback is the kind nobody writes into a report
The most important consequence of an atmosphere of trust: people relax in the good sense and become themselves. Their energy no longer goes into holding up a mask. And when you see a person at their baseline as they actually are, a manager gets used to that baseline — and any deviation reads instantly. Not because people are being watched, but because a change in a living person is visible if you actually care.
In a remote team — and my teams are distributed — the "baseline" lives in very concrete channels:
- Twice-weekly meetings. Someone who always walked through the details of a task suddenly switches to "eh, fixed a couple of bugs or something" — that is a signal.
- Chat reports. Same principle: dropping out of the usual level of detail.
- Response speed. We have windows when a person is expected to be reachable (they overlap with the activity of adjacent teams). A change in reaction speed inside those windows is noticeable right away.
- Branch dynamics. I pass the idea of balance to the leads: either many small tasks, or one big but decomposed one — so that a person keeps pouring into the branch regularly. One line over a couple of days means something is off.
- Tone. In calls and in writing. The subtlest and the earliest channel.
A real example: a person stopped smiling in meetings and started trying to say their part fast and get out of focus. I reached out — less than a week from the first signal. It turned out immigration was going badly, and then the situation was compounded by a personal tragedy. We did everything we could to help, but in the end we parted ways — on good terms. I bring up this case deliberately: a system of early reading does not guarantee a happy ending. It guarantees you learn about the problem within days rather than months, and that you have time to do everything possible — and that the person leaves (if they leave) without resentment, rather than through burnout and conflict.
Where the approach ends
Honesty requires saying where this works and where it does not.
Scale. With twelve direct reports, holding the "baseline" for each one is realistic. At around 40 people in the loop it goes through the leads: "processing" one lead replaces processing their whole team, and for a lead it is even easier with me — a manager understands a manager’s pain. Some leads need to be talked to more often, but that is within the margin of error.
A new team. If I were building a team from scratch for a new project, I would go back to frequent 1:1s for the first stretch. Until there is a "baseline" for people, there is nothing to read ambient signals against: hand on the pulse first, free swimming later.
New hires. The same logic in miniature: with a new person, communication is denser and more informal at first — human to human, with emotions, stories and anecdotes exchanged, not only statuses. This lines up with the regular ping from my article on onboarding: contact frequency decreases as the person finds their footing — and as I accumulate their "baseline".
In short
- A calm emotional background clears the fog: people do not wear masks, and a deviation in behavior gets noticed in the first days rather than at the quarterly report.
- In turbulent times, artifacts and meetings are the last thing on people’s minds — what matters to them is feeling support and openness from leadership and being heard. Monthly mental 1:1s deliver exactly that (and trust and loyalty grow as a side effect).
- The rest of the time, a quarterly collection of facts about process is enough — with the loop mandatorily closed, so that the person sees: what was said in a 1:1 changes reality instead of settling in a notebook.
Frequent 1:1s treat the symptom — a missing channel. Safety treats the cause. And once the cause is treated, feedback stops being an event in the calendar and becomes a property of the environment.