The main weakness in my management approach

Every manager has a different approach that, ideally, aligns with their priorities (although alarmingly often does not – different topic though). There is no perfect approach; there is only a set of tradeoffs that you make relative to your team's strengths and weaknesses, company goals, and your personal style and preferences. Eliminating weaknesses completely is a fool's errand and pretending you don't have weaknesses is worse. If you understand your weaknesses, you can mitigate them or simply choose to accept them based on your capacity and other priorities.

One of my guiding principles is to always trust my team and give them plenty of leeway in terms of both technical approach and the amount of time their assigned tasks take. This has two major benefits when things are going well: I have more time to devote to strategic tasks and whatever tactical interventions are needed, and my team feels more empowered and autonomous. And let's be real, my team should be more knowledgeable than me about many of the technical details anyway. If I need to be in there telling them what packages to use or what format to save their data in, we are straight up not having a good time.

This approach works fantastically well when my team is stocked with self-motivating high performers, which has usually been the case throughout my career (see my previous post on what makes a good Data Scientist), whether I hired or inherited them. Of course, we're not always that lucky, and no interview process is going to screen out every potential underperformer – we make our best guess based on a few hours of getting to know someone, while it usually takes at least a few months for underperformance to become apparent.

The main drawback of this approach is that it's going to take me longer to notice a certain type of underperforming team member and take corrective action compared to a more hierarchical approach. There are plenty of types of underperformance I'll still catch quickly, but someone who more or less gets things done but takes too long to do the work can fly under my radar for a bit, whether they're coasting or just work too slowly.

I am willing to accept this drawback for a few reasons. First, I think this approach allows high performers to flourish and develop and the productivity gains from the high performers outweigh the productivity losses from any underperformers. Second, the time it would take me to assess whether every team member is making appropriate technical progress is better spent on tasks that produce value rather than attempting to find a problem that may not exist. Third, how long a task should take is really hard to assess anyway, so you'll only get an assessment over several projects and many sprints.

By the way, the biggest negative to having this kind of underperformer on the team is not the reduced productivity, it's the effect on team morale. Most high performers are intrinsically motivated to perform well regardless of what's going on around them, but people talk and you need to catch and deal with this issue before you land in a situation where team members no longer trust each other to get stuff done efficiently. Sometimes a team is willing to accept a member who works more slowly than the rest because they bring other benefits, including being good to work with, and they're generally more tolerant of temporary slowdowns they can attribute to some external cause. But once they've decided someone on the team isn't pulling their weight, bad things happen to your culture and morale.

Sometimes, a team member will perceive underperformance as unfair if they think someone is "getting away with something" or doing less work at the same title and pay (or higher). I don't actually care about the justice aspect of this issue outside of its effect on morale – my job absolutely isn't to police my team and make sure nobody is slacking off. My job is to make the team as productive as it can be. Monitoring individual output too closely can lead to neglecting overall team output, and people who produce are going to have better careers and salaries over time than people who don't produce.

If your team members are feeling like someone isn't pulling their weight, they might tell you if you've earned enough of their trust for them to feel safe doing so. This is especially likely to come up during review and promotion seasons. As a bonus, if you know there's a temporary reason for underperformance (personal issues, hidden complexity, hardware problems, etc.) they will likely accept this rather than assuming their teammate is unreliable.

If I haven't been alerted to the issue by someone on my team, I'm going to have to spot it myself, which usually occurs during one-on-ones. I always check in on longer-running tasks to understand what's going on. Often, there's an unexpected roadblock, a competing priority, or the task is more complex than we anticipated. Sometimes that person is simply less familiar with the system or technology than other team members, which will improve over time, or they're having a hard time working with someone on the project. On top of that, estimating how long something will take before you do it is an inexact science and sometimes we get it wrong. These are all things I may be able to help with.

If you're seeing a pattern with the same person's projects always taking longer than expected, you have to be a good empiricist and rule out all of the acceptable reasons before concluding that you have an underperformance problem. That means eliminating as many external decelerators as you can control, being crystal clear about your expectations, and following up regularly with your team member.

The majority of times someone underperformed on a task, or had been labeled as an underperformer by a previous manager, talking to them about why they were struggling and addressing what they said resulted in them performing better. Often none of their previous managers had even told them they were underperforming, and they had no idea. They also might not have realized that you could help with some of the external issues they're facing, especially if they'd already been written off.

If they tell you about difficulties in their personal life, I suggest you show them some grace while they get through the tough times and see if their performance doesn't naturally improve afterwards. I know I didn't do as well at work when I was going through my mom's health issues and my divorce, and the lack of understanding and empathy from my employer led directly to me quitting when I was back to full capacity. Your job is to make the team the best it can be over a long period time, not at any one specific moment, and improvement today is a bad tradeoff for burnout or losing a team member later.

If none of that works and they still struggle after you've unconfounded as many external factors as possible, you have an underperformance problem.

This is where I'd love to drop in a list of steps you can take to fix an underperformance problem but the reality is that it's hard and it's not always going to work. You'll need to be crystal clear about your expectations and where they're falling short, and either work more closely with them or assign another team member to help keep them on track. Even if they improve, you'll need to keep an eye out to make sure their performance stays where it should be after you stop working closely with them.

In addition to being clear, be kind throughout this process because hearing that you're not meeting expectations can be very difficult. Give your team member their best shot at improving or leaving with dignity (by choice or not). Sometimes a person just isn't a good match for a role, a team, or a company, and there's no shame in that. Most of the time, finding a better fit for that person will unlock their ability to perform, even if it's outside your team or company. I have found that it's rare for someone to be an underperformer everywhere they go, no matter how bad their reputation in a previous role.

For me, part of my role is to help people find that better fit even if it's not on my team – if they want my help. I don't think this is required to be a good manager, but it aligns with my approach and values. I think it's possible to handle underperformance without making any of it personal, but I also understand when people take it personally, making me the bad guy in their narrative. That's their right as well, and that's part of my job to accept. If you're hung up on wanting people to view you positively, you probably aren't supporting your team as effectively as you could be. Prioritize being helpful over being likable.

As an aside, early in my career I brought an analysis to my boss showing that I was the lowest-paid person with my title in my department and I highlighted one person that I was positive I was much better at the job than (NB: don't give data people access to salary data and expect them not to use it to make sure they're being paid fairly.) His advice, while pushing to get me a raise, was to be careful about tying my performance to any other specific person's performance. If I am indeed better, then in a few years I should leapfrog that person, at which point I won't want anyone to anchor their understanding of my value on that person's value. And he was right, within a few years I had gotten a promotion and had long since passed that person in terms of salary. I've also learned that you can always cherry pick someone a level higher than you that you're better at the job than, so it's not that impressive.

I'd love to hear about the weaknesses you feel in your own approach, or how you see the weaknesses in my approach – let's get better at this together.