Kind managers address underperformance early and accurately.
Underperformance is when a person or a team is not bearing their share of the organization’s load. Their colleagues are either relying on them and getting let down, or they’ve learned not to rely on them at all. There are two fully unrelated causes of underperformance: Refusal to Align and Failure to Execute.
Refusal to align
Every person I’ve fired, both ICs and managers, refused to align their goals with the company’s. They were well-intentioned and often highly capable but pursued their own direction.
This might sound like a small problem but it’s fundamentally a question of values. Does this person want to be a functional part of this team? Or is this team an excuse to work on some cool technology or their pet project? It’s impossible to coach someone into different values — all we can do is set boundaries and let the person decide if they’d rather align or leave.
I was an underperformer in this way in 2018 when I led Infrastructure at Gusto. I thought I’d been hired to improve both Platform and Infrastructure but my CTO wanted me to stay below the Infrastructure Gravity line.
I worked late nights to improve the platform because I was sure it was important — even after my CTO encouraged me to stop. I had to relearn that it’s better to fix things slowly with trust than quickly without.
I share this to illustrate that competent people can underperform in this way. I’ve even come to expect it when someone moves from a big company to a small company for the first time — they’ll often do just what I did and try to solve problems they alone believe are critical.
Addressing Refusal to Align
The first time I managed someone who refused to align I put them on a 60-day PIP1. They did better and I declared the PIP a success. One week later complaints about them resumed as the individual returned to their old ways. I fired them shortly after.
The second time I did a 30-day PIP.
With each subsequent instance I’ve acted faster. I now believe that, if I understand the dynamics at play, a leader only really needs a week to work with the person and see if success is possible.
A one-week timeline
Here’s what I’ll do: I’ll meet with them on Monday and explain precisely what I need. I’ll ask them to convince me — and their colleagues — by that Friday that they understand the problem and are changing direction. These are the kind of hard, short conversations where I have to show up as a leader, not a friend; their job and their colleagues’ morale are on the line.
If they’re willing to realign it’s pretty easy for them to prove it. Success here is a message to the broader team saying “I know I’ve been focusing on implementing my new datastore while you’ve all been solving actual problems. I’m sorry about that and I’m going to shelve it until we all agree it’s a priority.”
If they don’t want to do it they can run out the clock in many imaginitive ways. I recommend not letting it drag out.
Managerial failure modes:
- Treating this as a skill gap as if training will help
- Waiting until the report agrees with you about the problem
- Assuming the cost of doing nothing is merely one person’s lost wages
Failure to Execute
The other kind of underperformance is the kind we normally think about: Missed deadlines, low-quality work, lack of throughput, no progress without handholding, disengagement, and unresponsiveness.
There are many complex reasons someone might not execute and it’s important to not jump straight to “they’re bad at the job.” Any one of us will struggle with execution when we’ve been handed rapidly-changing or poorly-defined goals, so in many cases the fix has to happen at the managerial layer.
I therefore look for the cause in roughly this order:
Do they know what’s expected of them?
Clear goal setting is very hard. When in doubt, have someone repeat back to you what they believe is expected of them and by when.
Remediation: In your 1:1 doc with this person create SMART2 goals and regularly check on progress.
Have they learned helplessness?
Looking at this person’s past projects, consider whether the projects were impossible. They may have been underfunded, undercut by cross-functional surprises3, poorly scoped, improperly staffed, etc.
In the extreme form this becomes burnout. My favorite definition of burnout is “being unable to make progress on something meaningful that’s in line with our values.” If you were to apply current to an electric motor but hold the rotor still it would catch fire. People burnout in similar ways – if they can’t move forward they’ll just heat up.
Remediation: Identify a small upcoming win and tightly coach this person into and through that success. Once they’ve won at something you’ll uncover if this is the blocker.
Do they have the skill for their current project?
They might be an amazing platform engineer but perhaps they’re on the Growth Eng team and they’re bad at building UIs. There are many specialties within engineering and not all skills transfer.
Remediation: If the report lacks the skill for their current project then either move them off the project or assign them a dedicated mentor for the duration.
Have they received literally any feedback that they’re off track?
If the report has been operating in an information vacuum then meet weekly (or more) with them for 15 minutes to each share what you believe is happening and how it’s going. Synchronizing their view of reality with yours will either solve the problem or reveal a deeper one.
Crises in their personal life?
We all go through hard times and that’s okay. As long as a person’s colleagues know what they can and can’t expect then lightening someone’s load has no negative effect on team morale (in my experience quite the opposite).
Remediation: Put this person on important but not time sensitive projects and tightly watch that they’re making some progress. If not, switch them to a simpler project.
Are they subject to a harmful managerial relationship?
We all hope the managers who report to us are kind and respectful at all times but they’re just people. Nothing cuts through a person’s ability to execute like the belief that their manager doesn’t have their back. If you have a skip-level report who doesn’t trust their manager you may need to enlist HR but there are some simpler options to try first.
Remediation: Meet with this person weekly for four weeks to see if you gain more insight. Consider simply moving them to another team. Coach the manager as needed.
Getting back to winning
In practice underperformance comes from a mix of the above reasons. The one remediation that I find most useful it getting the report (and their team, and their manager) to experience victory. The sensation of winning solves a lot of problems.
My primary tool for doing this with underperformers is a silly little trick derived from the Theory of Change4: I meet with the person who’s underperforming and make sure the goals are clear. Then I ask them to make a prediction about what will be true at a specific moment in the future.
This trick is so simple and has saved me and my teams so much grief. I have the person tell me what they believe they’ll have accomplished in one week, one month, etc.. I prod for as much detail as they can give me and I write it all down in our private 1:1 doc. Then I create a calendar invite for us at that moment in the future and on that day we re-read their prediction together. It’s usually hilariously far off from what happened. At this point we try to make new, more accurate predictions and suddenly it’s not me versus them it’s the two of us shoulder to shoulder against a problem.
When a report doesn’t take this process seriously I’ve learned to associate that more with a refusal to align. If someone doesn’t actually want to deliver for their team that’ll be apparent here.
Seeing this at the team level
Underperforming teams look much like underperforming individuals.
Failure to execute
Failure to execute shows up as missed deadlines, low-quality work, disengagement, etc.
Remediating a team is harder than an individual because the manager is individually underperforming (or else you wouldn’t have to get involved at all) and, additionally, individual members of the team may require personalized help.
The first step is see if the manager can even perceive the problem. If so, develop a plan and figure out why they hadn’t raised it earlier. If not, you’ll need to take over for this manager temporarily and directly lead the team until they’re on track.
If the manager works alongside you (without ego) on a fix then you’re about to get a hugely upleveled manager. If not, this person might be more functional in an IC role.
Refusal to align
Refusal to align at the team level looks much like when an individual refuses to align: The team works on pet projects, has an us-versus-them mindset, or goes dark periodically. Teams get misaligned when their manager is misaligned so it may be that the manager is using the team for some purpose other than what the company needs.
As of this writing I’ve witnessed DataEng teams at two companies where the reports wanted to work on what’s best for the business but their manager pushed them to roll out an interesting (and unnecessary) technology. In both cases removing the manager let the team quickly realign with their colleagues.
-
A “Performance Improvement Plan” which actually does sometimes work, but most of the time it’s a formality
-
Specific, Measurable, Achievable, Relevant, and Time-Bound
-
If you think an executive team changing major goals every month is rare I am so happy for you.