There are a lot of people who repeat things in engineering that sound fair and true and good.
And some of those things are completely wrong.
Have you ever heard someone say a variant of the classic saying, “Everyone knows it takes a woman nine months to have a baby. But you Americans think if you get nine women pregnant, you can have a baby in a month.”
I mean, Elon Musk sure seems to be trying to test this hypothesis.
The problem with this expression is that it is absolute psychobabble.
Software isn’t babies, and if King Solomon was trying to figure out how to identify who owned a software project by splitting it into a bunch of pieces, everyone would look relieved. It would be the worst “You ARE the father” episode of Maury anyone has ever seen.
If you have no idea what I am talking about, that is fine. Soon, you will be old enough to purchase your own alcohol.
This ridiculous expression is meant to give engineering teams a “get out of deadline free” card. And it works because most people find out they are about to miss their deadline when it is too late to do anything about it.
The actual problem with this problem is that software parts are generally interchangeable, so for some value of pregnant moms, you can have a baby in some number of months other than nine.
The real challenge is in the timing.
If you are four months away from a deadline, and you have twenty person-months worth of work, there is a universe where you can actually ship on time.
If you have three people on your team, you can get more than halfway there on your own. A little sixth-grade arithmetic also lets you evaluate the value of adding two more engineers to the mix. The challenge is that they need time to ramp into the project, and you have to give them at least a month.
So you will lose about a week of productivity from one engineer who is evaluating all of their hello world level ramp tasks, and you will also invest one month into each engineer for them to be successful. That still gives you a net add of around six months. Sixteen and six is twenty-two person-months, which gives you a safer path to a successful launch.
The reason we don’t have that super fast nine moms grunting out a one-month baby is that most people do not want to do the elementary school level math to size the impact of extra resources early enough in the project.
By the time you tell your boss the project is going to be late, and he wants you to add someone to the project, the ramp time window alone is going to eat any net gains of a new team member.
I have explained this very slowly to people, using very small words.
Repeatedly.
The average C-level executive will stop day drinking long enough for you to go and make your pitch for extra team members to finish a product on time at least twice, and the first time will be too far away in the future.
Random leaders generally don’t think tactically that far into the future, unless they are doing the math of acquiring themselves a schwaggy new Bentley ride. They work hard daydrinking until 4:59 PM; They deserve those sick wheels.
Hiring two more engineers will bloat the costs of the team, and if it is that far away from the deadline, then bossman will be at risk of losing their own bonus. Suddenly, that five-year-old Jetta is blocking the view of the Bentley. While it is good for the team and good for the product, you are likely going to get vetoed that far away from the end of the project because the average suit would love to roll around town in a Bentley instead of a washed-up Jetta.
There is a point to all of this.
You are likely to get shot down the first time you ask. You have to try. It is okay to be shot down with this kind of request. Once. We all get one turn inside the barrel. The important thing is to capture this data so the next time it happens, you can point out the consequence. When someone finds out that the responsible party for the last failure is looking out at them from a mirror, you can get people approved and onboarded in time to make a difference.
I will also add that you probably don’t want to try this with eighteen person-months of work still remaining and trying to hire 9 people to each do one part of the project. Eating nine months of ramp time is hard. You will likely be able to get away with two or possibly three headcount to solve this problem.
So go ahead and get three people pregnant. In four months, you will have a wonderful baby. Maybe you can give it one of those stupid Elon Musk names.
I gotta go now, HR is on their way after reading this post.
See you next week.
