Making one on ones better

Nvidia’s Jensen Huang may be an uber-talented chief executive, but his avoidance of one on ones is baffling. Maybe a lowly engineering manager isn’t one to critique the workflow of the head of one of the hottest tech companies around. Still, I consider 1:1s a fundamental management tool, one of the best ways to level up the team and build a culture of trust and open feedback within an organization.

Admittedly, Huang’s critique makes a few good points: 1:1s are usually a poor way to align strategy en masse or provide generalized downward communication. Managers should save that messaging for a Slack message or the next group all hands. But I suspect that 1:1s get a bad rap less for their potential, but instead from poor practical experiences.

Focus is a typical problem; many managers reframe 1:1 meetings around their needs. So you get horror stories of thirty minutes of one-way communication from a manager to a report that could have been an email. Or a meeting that devolves into a mini status report, a format for the manager to “deep dive” into the latest tactical project with the report on their back foot. And because it’s their meeting anyway, the manager happily shifts around a 1:1’s timing at the last minute or cancels it outright.

Managers need to reframe their mindset to make 1:1s productive and lasting. 1:1s aren’t about you, the work, or even the team. Instead, they’re all about serving the perspective of the individual report. What will build trust? What will help this engineer grow in their career? What will keep their morale from torpedoing when times get challenging?

1:1s provide no “one size fits all” solution. Details change based on the manager, the report, and the team mission. However, the proper format balances flexibility and rigidity, a concept that may sound contradictory but is by design.

1:1s are, from the report’s perspective, flexible or, at the very least, fungible. Most weeks, the report sets the agenda; it’s their meeting focused on their needs. I like a weekly thirty minute cadence as a starting point, but I often size up or down the meeting frequency on demand. Occasionally, an engineer can be heads down on a project with little on their mind in the long term to discuss; we’ll mutually cancel the meeting and give time back. Alternatively, we’ll touch on a subject for which the conversation will easily blow past thirty minutes. Maybe it’s a promo packet due soon, or a report’s morale has cratered, and we need to get back on track. In these instances, it may make sense to book multiple meetings in a week and dial back the frequency later.

But 1:1s are not a free for all. They work best with a set degree of rigidity and consistency. Because I, as the manager, generally have a packed, inflexible existing meeting schedule, I set the times when my 1:1s happen. I’ll post a signup sheet with a narrow set of open slots. Those openings are purposefully clustered around other existing meetings in my schedule, often back to back on just one or two days so that I can stay in “1:1 mode” for an extended period of my week.

I’m also inflexible around a shared source of truth for 1:1 notes and agenda items. That means a private document format (e.g., Google Doc, Confluence wiki) shared across the report and myself. I list out 1:1 dates in the doc in reverse chronological order, starting from several weeks ahead of the present. Against each date, there’s room for the report or me to write bullet point summaries. For future dates, the space is for proposed topics to discuss on the agenda. After the 1:1 happens, the space becomes a place to recap significant points covered in the 1:1.

The shared doc is purely a tool to facilitate good 1:1 conversations. Some otherwise talented engineers may struggle with topics to discuss, so I encourage them to capture ideas as they pop up in the doc asynchronously throughout the week. I’m also a heavy notetaker, so after a 1:1 wraps, I like to add a few bullet points to summarize our conversation. It’s a trivial effort on my part, and it can give reports a clear paper trail of our chats, hopefully spurning further ideas.

Rigidity can also be helpful around 1:1 topic cadence. In particular, for some reports, especially those earlier in their careers, I like building in a schedule so that once every two to three 1:1s, we explicitly focus on their career growth, usually on a focused, narrow aspect of their job like reviewing PRs efficiently or stakeholder management. Building a consistent schedule can help engineers keep a growth area at the forefront of their minds, even amid an otherwise busy or scattered day-to-day.

My personal 1:1 flow may only work for some. That’s fine! Tools and structure are only beneficial if you see a corresponding gain in the quality of your 1:1s. When one reflects on the latest few conversations, are the reports engaged? Are conversations two-way and productive? Is the total cognitive and time load – prep work, the actual meetings, and follow-ups – manageable? If ‘yes’ is the answer to all these questions, that’s a move in the right direction.

But even at my best, I rarely can give a universal yes to the questions above. On a team of eight or so ICs, I might have one report for which 1:1s remain challenging. The IC may be an otherwise strong developer and quality team member, but in the 1:1, they are unprepared and convey disinterest.

Before I worry too much about exceptional cases, I consider the bigger picture. If the engineer is knocking out the work, is happy, and is growing in their career, there’s no need to stress out. Cut down the frequency, keep check-ins for only the most critical topics, and let them loose. Remember: it’s ultimately their meeting, not yours.

However, if the 1:1s are a drag and the engineer has some clear growth areas, I’ll mix it up with a more direct approach. Without a 1:1 connection, I’ll lose my best venue for building trust and open feedback if I don’t take action.

Usually, I’ll book a straight one-off hour with them, with the focus on exclusively deep, more philosophical questions about why we’re meeting to begin with. I’ll probe on 1:1 experiences with prior managers; there might be a horror story or two to work through that has bred a deeper mistrust in the ritual. I also like to ask what they would want with zero limitations on what we could make happen in a 1:1 manager-report context. Crystal ball wild swings can bring insight into their real goal, to which you can react and size accordingly. If there’s a key growth area, I’ll discuss how I can help. Even the biggest 1:1 cynics want to grow their career, and if they see the meeting format as a way to move up, they may have a change of heart.

And when all else fails, be it with a report resistant to 1:1s or any other 1:1 that misses the mark, I remind myself that the meeting format’s idiosyncratic, humanistic setup ensures not every chat will be a winner. But occasionally, we listen and converse with empathy and curiosity and emerge as better, more productive engineers. Embrace the meeting format’s inherent messiness, and know that a few well timed, impactful conversations will outweigh those that feel unmemorable in the long run.