In the first article in this series, I wrote about psychological safety, the foundation that makes everything else in leadership possible. If there’s one place where that foundation gets tested every single week, it’s the one-on-one.

I’ll be honest: my early one-on-ones were terrible. I’d sit down with a report, open Jira, and essentially run a standup for two people. “How’s the migration going? Any blockers? Cool, see you next week.” I thought I was being efficient. What I was actually doing was wasting the most valuable recurring meeting on my calendar, and probably theirs too.

It took me an embarrassingly long time to realise that if I could get the same information from a Slack message or a board update, the meeting had no reason to exist. The turning point came when a senior engineer I managed told me, quite directly, that our one-on-ones felt pointless. That stung. But it was exactly the feedback I needed.

It’s Their Meeting, Not Yours

The single biggest shift in how I approach one-on-ones came from a simple reframing: this is their meeting, not mine.

That sounds obvious when you say it out loud, but think about how most one-on-ones actually run. The manager sets the agenda. The manager asks the questions. The manager steers the conversation towards whatever’s top of mind for them, usually project status, upcoming deadlines, or that thing that went wrong last Thursday. The report sits there, answers questions, and leaves. It’s an interview, not a conversation.

When I started treating the one-on-one as time that belongs to my report, everything changed. They bring the agenda. They decide what we talk about. If they want to spend thirty minutes working through a career question, we do that. If they want to vent about a frustrating code review process, we do that. If they genuinely have nothing and want to cancel, that’s fine too, though I’ve found that rarely happens once people realise the time is actually theirs.

Vrashabh Irde wrote something that stuck with me on this: engineers should come to one-on-ones prepared to talk about their career goals, how they’re feeling, and what’s on their mind. Not what tickets they’ve closed. The manager’s job is to create the space for that conversation to happen, not to fill it with status updates they could get from the sprint board.

This doesn’t mean I never bring topics. Sometimes I need to deliver feedback, or there’s an organisational change coming that affects them, or I want to check in on something specific. But those are additions to their agenda, not replacements for it.

The Problem With the Status Update Trap

Here’s what I found interesting when I started paying attention to this: the status update trap isn’t just inefficient, it’s actively harmful.

When one-on-ones become project check-ins, you’re implicitly telling your report that their value to you is their output. You’re reducing a complex human being with career aspirations, frustrations, ideas, and anxieties to a progress bar. And people pick up on that, even if you don’t mean it.

Worse, it creates a dynamic where the report feels they need to perform. Every one-on-one becomes a mini-review where they have to demonstrate they’ve been productive. That kills the psychological safety we talked about in the last article. Nobody’s going to tell you they’re struggling, or that they’re bored, or that they’re thinking about leaving, if every conversation starts with “so, what did you ship this week?”

Lara Hogan talks about the “manager Voltron” concept, the idea that different reports need different things from you at different times. Some weeks, someone needs a sounding board. Other weeks, they need you to unblock something political. Sometimes they just need to know someone gives a damn about their career beyond the current sprint. You can’t do any of that if you’re spending the whole time on status.

Finding a Structure That Actually Works

I’m not going to pretend I’ve got some perfect template that works for everyone. I’ve tried several approaches over the years, and what I’ve landed on is more of a loose framework than a rigid structure.

Camille Fournier describes four styles of one-on-one: the to-do list, the catch-up, the feedback session, and the progress report. What clicked for me was realising that a good one-on-one moves between these styles fluidly depending on what the person needs that week. Trying to force every conversation into one format is where things go wrong.

That said, I do find it helpful to have a rough shape to fall back on, especially when someone’s new to the team or isn’t sure what to bring. I’ll usually start with something open, “What’s on your mind?” or “How are things going?”, and let them steer from there. If the conversation naturally covers something meaty, brilliant. If it stalls, I’ll have a few prompts ready:

• “What’s been the most frustrating thing you’ve dealt with recently?”
• “Is there anything you’d like to be spending more time on?”
• “How are you feeling about your growth here?”
• “Is there anything I could be doing differently to support you?”

These aren’t revolutionary questions. But they’re open-ended enough to go somewhere interesting, and they signal that I’m interested in more than just throughput.

Google’s Project Oxygen research reinforced something I’d started to suspect from experience: the best managers are good coaches, and good coaching is mostly about asking the right questions and then actually listening to the answers. Not waiting for your turn to talk. Not mentally composing your response while they’re still speaking. Genuinely listening.

The GROW Model: A Framework Worth Knowing

When I started reading more deliberately about coaching techniques, I kept running into the GROW model, Goal, Reality, Options, Will. Addy Osmani covers it well in his writing on engineering management, and the more I dug into it, the more I realised I’d been doing a clumsy version of it intuitively.

The idea is straightforward. You help someone work through a challenge or goal by moving through four stages:

• Goal: What do you want to achieve? What does good look like?
• Reality: Where are you now? What’s actually happening?
• Options: What could you do? What are the possibilities?
• Will: What will you actually do? What’s the next step?

I don’t sit there mechanically walking through these stages, that would feel robotic and weird. But having the framework in my head helps me notice when a conversation is stuck. Often, someone will come to a one-on-one frustrated about something, and they’ll loop between Reality and Goal without ever getting to Options. Just gently asking “what could you try?” can unstick the whole thing.

What I particularly like about GROW is that it keeps the ownership with the person. You’re not solving their problem for them. You’re helping them think it through. That’s a hard habit to build, especially if you come from an IC background like I do. My instinct is always to jump in and fix things. But the best one-on-ones I’ve had are the ones where I mostly asked questions and the other person walked out having solved their own problem.

The One-on-One Retrospective

One practice I picked up from Marcus Blankenship that I’ve found genuinely useful is treating roughly every fourth one-on-one as a retrospective on the one-on-ones themselves.

It sounds a bit meta, but it works. You’re essentially asking: “Are these meetings useful to you? What would make them better? Is there anything we should talk about more, or less?”

The first time I tried this, I got some uncomfortable feedback. One report told me I had a habit of jumping to solutions too quickly, they wanted to think out loud, and I kept short-circuiting the process. Another said they wished we talked about their career progression more and project stuff less. Both were things I wouldn’t have known without asking.

It also reinforces that the meeting belongs to them. You’re explicitly saying: “I want this time to be valuable for you, and I’m willing to change how I run it based on your feedback.” That’s a powerful signal.

Never Skip, Always Follow Through

Duretti Hirpa wrote something that I think every new manager should have tattooed somewhere visible: never skip one-on-ones, and always follow through on what you say you’ll do.

Skipping one-on-ones sends a message, whether you intend it or not. It says: “Something else is more important than you right now.” Do it once, fine, things happen. Do it regularly, and you’re telling your reports that this time isn’t actually sacred. And once they believe that, they stop bringing the real stuff. They’ll save the meaty topics for someone who actually shows up.

The follow-through piece is equally important. If someone raises a concern in a one-on-one and you say you’ll look into it, you need to actually look into it. And you need to come back to them with what you found, even if the answer is “I tried and couldn’t change it.” Nothing destroys trust faster than a manager who nods sympathetically, writes something down, and then never mentions it again.

I keep a running document for each person I manage. After every one-on-one, I jot down anything I committed to doing, and I review it before the next one. It’s a small habit, but it’s probably done more for my credibility as a manager than any leadership course I’ve taken.

When the Best One-on-Ones Feel Like Nothing Happened

Here’s something that took me years to get comfortable with: the best one-on-ones sometimes feel like they covered nothing at all.

There’s no action item. No decision made. No problem solved. You just… talked. Maybe about how they’re feeling about the team dynamic. Maybe about a technology they’re excited about. Maybe about something completely unrelated to work.

Early in my management career, I’d leave those meetings feeling like I’d wasted the time. I had a mental checklist of things a “productive” one-on-one should cover, and a rambling conversation about someone’s side project didn’t tick any boxes.

What I’ve come to understand is that those conversations are doing something crucial: they’re building the relationship. They’re creating the trust that means when something genuinely difficult comes up, a conflict with a colleague, burnout, a desire to leave, they’ll tell you about it in a one-on-one instead of in a resignation letter.

You can’t manufacture that trust with a template or a framework. It comes from consistently showing up, being genuinely interested in the person sitting across from you, and proving over time that this is a safe space for honest conversation.

Making It Work in Practice

If you’re reading this and thinking your one-on-ones need an overhaul, here’s what I’d suggest based on what’s worked for me:

Start by handing over the agenda. Tell your reports that the one-on-one is their time, and ask them to come with topics. Some people will take to this immediately. Others will need a few weeks to adjust, and that’s fine, have prompts ready.

Drop the status updates. If you need project status, get it from your tools, your standups, or a quick async check-in. Don’t burn one-on-one time on it.

Ask open-ended questions and then shut up. This is harder than it sounds, especially for those of us who came up as ICs and are wired to solve problems. Resist the urge. Let silence do some of the work.

Keep notes and follow through. Write down what you committed to. Review it before the next meeting. Close the loop, every time.

Retro your one-on-ones. Every month or so, ask how the meetings are going. Be prepared to hear things you don’t love.

Never skip. Reschedule if you must, but don’t cancel. The consistency matters more than the content of any individual session.

None of this is complicated. But like most things in management, the gap between knowing what to do and actually doing it consistently is where the real challenge lives. I’m still working on it. Some weeks I nail it, and some weeks I catch myself drifting back into status-update mode. The difference now is that I notice, and I course-correct.

The one-on-one is the smallest, most frequent investment you make in each person on your team. It’s worth getting right.