Culture Shifts - Technical vs People Managers

This is part 2 of a post on how culture changed within engineering at Medallia.

When we initially split into teams, each team had a product manager, a unique tech lead, and a single shared people manager.

What does that manager do?

In my opinion, an engineering manager is primarily responsible for resource allocation:

• which projects/teams are funded
• who is staffed on projects/teams
• how salaries/stock options/bonuses are distributed to individuals
• setting hiring goals for existing/new teams
• prioritizing which tasks the team should focus on

However, there are two distinct styles of enginering management:

• people managers
• technical managers

People managers tend to play a more supportive role; enabling individuals by finding opportunities for them (e.g. technical projects, mentorship opportunities, leadership initiatives). They ensure a team has the resources to accomplish tasks; whether it’s bringing in an architect from another team, an external consultant, or bringing in other roles to assist (SRE, product support). They rely on technical leads and product leads for technical and product-related decisions.

Technical managers tend to play a more direct role; from setting a technical vision, and defining/tracking metrics towards goals, to choosing projects/libraries/frameworks to use, to actually writing and reviewing code.

As our engineering team grew, we clearly needed more managers. Since we already had strong tech leads, we originally focused on acquiring more people managers.

Why did we change this?

In short, it’s very rare to find a manager candidate with a solid technical foundation (necessary to work with a strong tech lead), but wants to focus on people management instead. Managers tended to be promoted from the ranks of “good engineers” and these candidates liked to stay in touch with the latest technologies. Managers that didn’t come from pure software engineering roles still tended to be career-focused. Progressing upwards towards higher management positions requires showing an ability to handle increased responsibilities. Demonstrating such proficiency is done by making the “right” calls for important decisions. It’s much easier to show that a technical decision was the right call vs. a people management decision, e.g. a “20% increase in sales due to critical feature X which I drove” is a more objective resume builder than “20% increase in sales due to my tech lead’s decisions which I supported”. In the end, high-performing eng mgr candidates tended to desire technical impact in their role and so we altered the role to fit the demand.

How did this affect the culture?

A tech lead is an individual contributor responsible for coding the best solutions for the team, product, company, and customers. Debates about libraries/frameworks/maintainability/performance are (or should be) followed up with investigations and data points. There tends to be a high level of transparency to these discussions; even when they become philosophical or religious (e.g. emacs vs vim), at least everybody knows who’s on which side and why. It’s much easier to trust a tech lead with decisions because you can always ask why.

A people manager is a manager responsible for allocating resources towards projects. These decisions are much more subjective and touchy involving who is better at X or who doesn’t like to work with Y. These tend to have less transparency due to the sensitive nature of pride and avoiding hurting people’s feelings. In a transparent world, people wouldn’t feel bad just because somebody else is better at X, and Y and Z would resolve their differences via open feedback and compromise. However, this isn’t the state of the typical American engineering work culture. In the absence of such transparency, we have to trust our people managers to make the right decisions for the team.

When these responsibilities are split, it’s clear that if the tech lead’s proposal was not taken, then there were other non-technical factors at play. When all of these responsibilities fall on one person, it’s a higher burden to explain why a particular decision was made. When the decision maker can not be transparent about why a decision is made, it’s very easy for him/her to lose the trust of the team.

I’ve seen managers that are able to obtain/retain that trust but it’s very difficult for every manager in a growing company to hold it. Lack of trust in a manager quickly leads to all sorts of unhealthy culture issues; e.g. people avoiding confrontation (leading to bad technical decisions).

I believe that “trust” is the most correct measure of a manager’s effectiveness. A manager that does not have the trust of his/her superiors is unable to make decisions that are objectively better for the team. A manager that does not have the trust of the team can not effectively execute on those decisions.