How to Decide to Fire

TLDR;

When you have a person reporting to you who's performance and behavior is consistently bad, firing them is an easy decision. What's much harder is when you have someone performing inconsistently who shows promise by doing some things well but others poorly. It's even worse and much more ambiguous if that person is managing other people. What do you do then?

Whether you fire or not boils down to how much of their job you have to do. When you find yourself spending significant time and bandwidth helping your person succeed and that investment in assisting them to achieve primary goals doesn't get smaller over time, it's time to cut your losses.

Rocky Road

Once upon a time, I took over an organization that had been ignored by the previous leader. One team had a single member but had a headcount for a manager and five members. The first thing I did was to hire a manager for the team. This manager interviewed well on both the management capability and his engineering capability. He had also been managing successfully for a long time. So we extended the offer and were happy to get him.

His first goal was to build out the team. This is where I saw the early indications of a problem. At that company, the interview process involved several interviews by different interviewers, followed by a final group decision called a debrief. In those debriefs, this new manager was nearly always inclined to hire. Every time the group decided not to hire a candidate, he showed visible discomfort. After those meetings, I would pull him aside to work through his reaction.  Every time, he expressed concerns about certain interviewers having too high a bar. He even went so far as to influence the outcome by putting subtle pressure on other interviewers in the loop. We had several conversations on this topic over months, without seeing any real improvement. The problem went away when we finished filling the seats.

After a month or so, it became apparent that one of the new team members was not delivering. It takes three to six months for a new engineer to get fully productive. I don't expect a lot in the first month, but I do expect them to have gotten something into production. When this new team member got blocked, the manager reassigned that work to one of the productive team members to finish up. When I found out about that, we had a long conversation about performance management and setting expectations. I set goals for this new manager regarding managing the engineer and held the manager accountable for keeping the engineer accountable.

After the engineer failed to deliver again, we started down the path of a formal improvement process. It was arduous. The manager was politely resistant to everything that required him to hold the engineer accountable. In one conversation where the manager said, 'I usually like to balance the team to cover weaknesses, so I never have to have these kinds of hard conversations.' That gave me the insight I needed to understand what was going on finally. This manager wasn't interested in having a team that performed well. He was only interested in having a team. While he understood and could talk coherently about all the facets involved in managing a team, he didn't want to do those things.

I spent a considerable amount of time mentoring this manager, helping him understand my expectations and align with them. In every case where the manager had to hold someone accountable, set a high bar, or have a difficult conversation, he failed to do so independently. Once I had explained what needed to happen and held him accountable, he would do it. However, he never did it on his own.

It took me too long to realize that this manager was failing. He did many things well, but he could not do the things he needed to do to produce a high-performing team on his own. He was able to do that when I gave him explicit direction and held accountable for it. That problem never got better, no matter how much time I spent with this individual. I got trapped in a 'boiling the frog' pattern.  I failed to resolve the situation because the job was getting accomplished. I didn't realize how that time investment affected my ability to focus on my job.

Knowledge vs. Will

There are two causes for someone not performing as they should. One is that the person doesn't understand what they need to do. The other is that they disagree with the goal itself. These two things can look very similar.

Not Understanding the Goal

Not understanding the goal or not understanding how to accomplish that goal is a solvable problem. You solve this through mentorship and support. You intentionally set goals, help that person understand why that goal is essential, and what they need to do to accomplish it.  When they miss something, review the miss with them and help them understand what they could have done better. With this investment, you should see improvement over time.

Disagreeing With the Goal

You need the people that report to you to be pulling at traces with you. That means helping you accomplish your goals by achieving the goals you set for them. At times, people that report to you don't agree with those goals. That's life and leadership. You should try to get them on board. You might be able to shift the person's view by walking them through your experiences and why you set that as a goal by pointing out the impact when the plan succeeds. You can ask them to give you the benefit of the doubt and try it, then point out the positive results when they occur. Turning someone into a believer who didn't start as a believer is often the best way to gain strong supporters for your cause. Sometimes, you need to ask someone to do it because you need it done, regardless of whether they agree.

When you arrive at that point, and they still don't do what you need them to do; it is time to recognize the issue and act on it.

Deciphering the Reason

So how do you tell the difference between someone who doesn't know what they need to do and someone who knows but doesn't care? When does a person have a knowledge gap vs. an alignment gap? The evidence is in improvement. When you are investing time in someone, you should see them improve over time. The amount of time needed and the amount of time you have to give varies by situation. If you are working in a large company, with stable income and your team members are generally good, you may have the ability to give many months and lots of effort for someone to improve. If you are in a startup with a short runway and a strong need to get to market, you may only be able to provide weeks. It all depends on the situation.

In any situation, invest in your people and look for the return on that investment. Where you see a return, continue to invest. Where you do not see a return, and you continue to have to invest, consider that sunk costs and let the person go.

A Word of Warning

The most significant danger here is that you are going to let this situation go on and on. Leaders succumb to a combination of the Sunk Cost Fallacy combined with an aversion to letting someone go. There are a couple of techniques you can use to avoid this trap.

Keep copious notes about everything in some easily cross-referenceable form. I use Roam Research. I also like private TiddlyWikis.  After every interaction, ask yourself, 'Have we talked about this before? Have I seen improvement since the last time we talked about this'? It is especially important to do that as part of your one-on-ones. Schedule a little time after your one-on-ones to clean up your notes and do introspection. Your organization should be doing Continuous Performance Management, that will help with this process.

If you don't see improvement, you should point this out to the person you are managing. Finally, when you realize that the person isn't meeting your expectations and isn't improving, it is time to start a formal process. Reach out to HR. Don't wait. Waiting always makes things worse.

This person may be the only person with some knowledge. They may be well-liked by the team. You may want to take time to find their replacement. Don't do it. Figure out contingency plans and let the person go.