I recently started reading the best-seller book “The Making Of A Manager” by Julie’s Zhuo and I’m still in the early chapters where the role of an engineering manager is being discussed & defined, and the discussion pretty much indirectly touches on a classic question: “should engineering managers write code?”

Julie is making a nice argument on why it makes sense for engineering managers to focus their time on having a multiplier effect, instead of being an individual contributor. The argument & multiplier effect is better illustrated with an example below.

Say you operate a business selling stuff and hired two people. You sell 100 items per day, each of your employees sell 80 items per day. By a simple a calculation you may realize that if you teach your employees your skills, then each of which will be able to sell 100 items. Of course, investing your time in such activity will reduce your output but the results will pay off in the long term. This is an example of how you had a multiplier effect on your business/team.

Another example, is recruiting. If you happen to know a person who is way better than you in selling, then investing your time to recruit them is also causing a multiplier effect, because their output will be higher than yours.

Thus, when engineering managers use their time to have as many multiplier effects as possible, then it’s better invested in such activities than coding.

However, in reality things are more complex and it’s hard to justify each activity simply because it’s hard to quantify the effect. Thus, several companies (especially startups) do require managers to use 30% of their time, especially when the team is small, to code.