They manage! Of course, there are other people who manage as well, but I've picked these two groups for some other reason.

They set up the strategies and tactics! Definitely, but there's something far more unique.

They are the first ones to be kicked out (if something goes awry, there's a suspicion that this is happening, or just "in case"  ;>).

If you're just a little bit into football (or "soccer" - according to Murricans), you know that football manager is a highly unstable job. E.g., in the top-level Polish league ("Ekstraklasa", with 18 clubs), 13 managers were replaced just this season (and it's not over yet ...). Only 4 of the currently employed ones work with their team for longer than 12 months (src: link). Wicked sick, isn't it?

Unfortunately, I don't have such data for top-tier technical leaders (CTOs/VPEngs). Still, based on some empirical data (at least for the startups on a high-growth curve), the average tenure at the position may be similar ...

It may sound counter-intuitive ("Hey, wait! Are they REALLY getting rid of their most experienced engineer SO EASILY?"), but in fact, it makes some sense (from a frustrated CEO's/board's perspective):

  • It'd be expensive (and troublesome) to replace (many) engineers. Maybe the (potential) issue could be solved with a swap at the helm (so-called "new broom effect")? Or in other words: CTO is supposed to be a multiplier for all the engineers - maybe another one would be a higher/better multiplier? It sounds like a bet with a reasonably high expected value.
  • A top leader (by definition) takes accountability for all the mistakes (or even wasted opportunities) of their subordinates - sooner or later (s)he will slip up ...
  • Actually, CTO is relatively easy to "plug into" the organization (once you find a suitable candidate). Top-tier engineering leaders have different cognitive load characteristics than "field-level" engineers. Their experience is an excellent compensation for the lack of particular, narrow technical knowledge and/or familiarity with local implementation intricacies.
  • The engineering department (and hence CTO) is always perceived as a constraint (from an ambitious product organization perspective). Its speed, quality, and agility (regardless of how good they are) have a friction effect on pure business creativity. Unfortunately, in nearly EVERY successful organization, these three qualities tend to decrease (in exchange for the growth: of complexity, scale, dependencies) over time - and there's a certain obvious candidate to be blamed for that ;)
  • Let's face it - many CTOs (simply) fail. The expectations for that position are very high, and we (as humans) at some point meet the ceiling of our competence (according to the Peter principle).
  • Last but (for sure) not least: CTO's responsibilities change very quickly when the company grows; "normal" developers will always have a place to fill, but the company may have very different expectations from the CTO now than a year ago. Not many CTOs are able to keep the growth pace of a blitz-scaling organization.

Of course, there are always exceptions. Both in technical leadership and football (think: Alex Ferguson, Arsène Wenger, or Guy Roux). The most lively CTOs remain in charge (typically) thanks to:

  1. a founder/co-founder status; or strong, trust-based partnership bonds (acquired during prior endeavors) with the rest of the C-suite
  2. unique (highly confidential and/or patent-protected) technical knowledge which is essential to their company's success (sort of Page+Brin casus)
  3. their deliberate effort to make themselves practically immovable: concentration of control/power/knowledge/communication/etc. - this is not always possible, and in many cases, it's a problem per se

Nevertheless, it's pretty naive to bet on exceptions. Hurry to appreciate your CTO - (s)he'll be gone in no time :)