Managers vs Engineers

Defmacro, 2020-10-05

Managers vs Engineers

  • Tech companies advertise two tracks of career progression: become an even better engineer (i.e. a Specialist) or an engineering manager (i.e. a Manager / Generalist).

  • Theoretically the pay grades for managers are very similar to engineers and advance at the similar speed.

  • But managers have an upper hand over engineers: they (managers) have the powers over the engineers that are not reciprocal: the reward power (running the engineers’ performance reviews + determining the salary), the information power (having access to the company’s plans and the comp structure of engineers), and the coercive power (being able to assign the engineer to a losing project and then terminate them).

  • While engineers are the ones that are operating the money printing press, the press itself was designed not to be dependent on a particular operator, with a fair degree of duplication.

  • US: engineers’ turnover rate is 13.2% (embedded engineers: 21.7%). But this is not the sign of their power to leave, it’s the indication that companies are built to withstand such turnover. The engineers are not able to stop the machines.

  • Pay grades. Google’s engineers: $188k - $1.35m p/a. Looks cool when you’re on the upper side of the pay grade, but there’s a compartmentalization problem. In short, this is a way to keep people in their narrow field for as long as possible, thus reducing the marketability of their skills and exerting downward pressure on the salary levels as a result.

  • This is a known problem to both sides (no one would call engineers inept, too). Applied games theory tells us the solution would be to reduce compartmentalization. Applied management theory tells us the solution would be to hire intermediaries AKA engineering managers.

  • The engineering managers are no longer the engineers who can fix the machines [MK: I stopped writing software 15 years ago and while I’ll moderately pick up the trade in 3 months or so, I won’t be able to catch up with the new concepts for another 3 years or so.] Their source of power is not an expert power as the engineers’.

  • The powers of the engineering managers listed above are not absolute and have to be shared with the engineers for a very obvious reason (no amount of authority can fix a broken machine).

What Could Possibly Go Wrong?

  • Engineering managers’ salaries are roughly on par with the engineers’ salaries, but … why not less? The official answer is the “servant leadership” [MK: a very popular concept I personally believe in to a large extent.] The idea is simple: insulate the working population (engineers) from the bureaucracy and provide timely updates on the state of the Party thinking.

  • Ahem… But why bundle someone’s performance review and removing of the roadblocks in a manager and not the engineer?

  • As it turns out, bundling control (over salary and advancement) and handling communication is not a natural law.

  • Dual tracks always evolve into overlapping status hierarchies.

  • Informal Prestige Hierarchy: there are always people in firms [MK: and I’m glad we have fantastic developers like this in our company] who are the go-to people when something goes wrong, and they’re deeply respected by everyone from the CEO down the chain.

  • The managers catapulted from above don’t have this sort of respect and at best help secure some resources and resolve personal conflicts.

  • The core issue is that this parallel track approach creates two markedly different organizations: the management organization with its reporting, growth prospects, language and issues, and the engineering organization that has the same except for the lack of loyalty towards them.

  • An engineering manager’s career progression has very different control points. The first one is the ability of not doing anything stupid on the new job, and then gradually the person gets exposed to political games where anything but the engineering skills can be useful.

  • Loyalty to fellow engineers becomes a liability, because of very different goals. And the survival technique for a manager becomes pretending to be one the “tech boys”, but acting and dreaming like a “biz boy” (or girl, no problem whatsoever). And here’s the circle of the cultural life (lasting roughly 25 years).