I’m an engineer. I’ve built a 45 person engineering team that shipped dozens of releases of award winning software without ever missing a schedule or compromising quality. There is nothing but analytical tasks for the engineering manager.
1. How do you maximize developer productivity? Motivation? Job satisfaction? How do you help them make technical decisions and plans?
2. What process best allows your specific teams to ship quality releases on time? How much planning overhead is necessary to do so, and how much isn’t benefiting the process or product?
3. How do you make your quality assurance staff most productive at finding and reporting problems? How can they best organize their testing to ensure the most useful/necessary test coverage, and how do they communicate their progress clearly?
4. How well do all these things meet the needs of your first level customer, the product owner?
5. How do you communicate your groups capacities to executives? How do you tell them what can be done, what can’t be done, and what shouldn’t be done?
Your tools, technologies and customer requirements are changing over time. Analysis never ends, you can always get better.
All the things you think are important, aren’t
Now I understand that there exist some people still named engineers that are only preoccupied about how to make a mechanical item solid and light enough to satisfy a requirement, or how to design a really cool and fast hash table. This is still completely in the engineering field, but only constitute a (tiny) part of it. And even then it is dubious to consider they are either not at all preoccupied about how to work efficiently with their peers, or not even subject to that kind of consideration.
Designing a system for humans is NOT an analytical task. It can be supported via analytics, but it's a political task (that requires others to be convinced that it is a good solution. For example, you can raise income levels, but if people get pissed off at inequality, you can't argue your way with numbers. They may or may not take it. And that was my point in the article).