Management tool tip: Retention Risk Matrix

Dear Readers,

This is the first in a series of articles about tools that I have found useful in my career as an engineering leader. Some of these may be specific to software engineering (pertaining to the Software Development Life Cycle, for example,) but most will apply to any team. 

This article addresses a challenge that has come up a lot in my conversations with managers across industries, and I hope you find it helpful. I call it the “Retention Risk Matrix”, and it’s an action-oriented approach to measuring and reducing organizational risk. 

Trial by fire

Unfortunately, many of us have experienced layoffs, especially in the last year. Layoffs are tough for everyone, but unarguably most traumatic for those directly impacted. Before and during a layoff, it’s our responsibility as leaders to make the process as thoughtful and humane as possible for our departing team members. (A boss and mentor of mine, Wendy Sheppard, once taught me that I had to “plan like a robot and execute like Mother Teresa”)

That said, once the dust settles, there is a long road ahead for everyone remaining. Rebuilding a team after saying goodbye to your friends is tough, especially once the reality sinks in that someone (most likely everyone) is going to have to pick up their responsibilities.

As a leader, not only is it essential for you to drive this process, it is also essential that you keep a pulse on the morale of the folks who are in the middle of it, making sure that they are supported and heard.

After experiencing a layoff at our then employer (Amino Health) last fall, my friend and colleague Anne Taylor and I developed what I now call a “Retention Risk Matrix”. The matrix tracks the risk of losing each of your remaining employees, and even more importantly, what you plan to do about these risks. This is a pattern that I was familiar with from my work with operational risks in SaaS platforms (most notably at New Relic).

Here is an example (hypothetical) snippet from a retention risk matrix.

Building the framework

Anne (who was my HR business partner) started the process by conducting a “listening tour” of our staff, having informal conversations with each employee separately to both gauge their morale and gather feedback for how we can improve our practices. This was an ongoing practice for Anne (and one of many ways that we channeled feedback anonymously to the leadership team,) but she conducted a focused campaign soon after the layoff to make sure we had an accurate and up-to-date assessment. In parallel, we also asked all managers to make their own assessments during their 1:1s (which were weekly or fortnightly). Not shown in this example, in the case where managers rolled up to a functional leader (most notably in engineering) we also included that functional leader’s assessment. 

Just as with interviewing candidates for an open position, having multiple perspectives was essential. The most interesting data points (and the ones that required the most attention) arose in cases where the observers were not aligned. As you can see in our Jane Doe example above, misalignment can often indicate a blind spot for the manager, and/or a growth area for the employee.

Getting to the heart of the matter

As I mentioned earlier, the most valuable output of this process is the plan. Managers were responsible for developing a plan to retain (or in some cases thoughtfully let go of) each and every employee. And just as important, we reviewed and updated the plans on a regular basis. As with any project review, there was an expectation that each plan would either have progressed or updated to accommodate for new learnings.

The ideal would be to achieve a “low risk” rating for each employee. But as with any risk matrix, we can use the product of the impact and risk to prioritize. In the example above the “Business Continuity Risk” column represents the impact. In this case, both Jane Doe and Joe Schmo are high risk cases, but for very different reasons. Jane is a highly productive employee who we are at risk of losing, whereas Joe is a disruptive employee who risks dragging down the morale of his entire team. Given that we also consider both of these employees high risk, the focus should be on immediate interventions for them.

This prioritization should also be reflected in the nature of the plans. In both of our high priority cases we see that leadership is going to intervene with the management to prod and support more immediate action. In the lower priority cases we are taking a more strategic and gradualist approach. 

The Mother of Invention

While this tool was born out of an urgent need, I find that it is a helpful part of a steady-state management practice. If you’re new to an organization, building this up is a good forcing function for you to get a pulse on your entire team and collaborate with HR. If you’re not new, but feel like you would like to have a better handle on your team member’s morale, this could be the tool for you as well. I hope you find it useful, and welcome any questions or comments.