While I agree that poor leadership is common, I strongly beg to differ with the reasoning that leadership is removed from the facts.
Friend - 'Engineering' is not 'the facts', or not usually.
Customers, business opportunities, markets, investors - these are usually the 'facts' in a business, and sometimes it means doing things that don't seem obvious to a coder.
I've never seen something great come to fruition that involved a linear progress from 'need' to 'product'. It's usually messy, and involves all sorts of actors.
For example, hacking together a 'throw away' something to impress a customer or to ensure they believe in the capacity of the company would likely been seen as a 'waste' from an ops perspective, but it's certainly not if the business situation calls for it.
Sometimes the facts are in the code, but this is much less often than one would realize.
That said, Palantir is a little scammy. They're just CGI or Accenture by another name.
Not sure why you're referencing engineering? I didn't say anything about that. Not all startup employees are engineers.
> Customers, business opportunities, markets, investors
Customer support is going to have more facts about customers, investors are going to have more facts about investment terms and opportunities, marketing team is going to have more facts about markets, biz dev is going to have more facts about business relationships.
Management is not going to have more facts in any area than anyone really. Their job isn't to understand what's happening in detail, it's to maintain the structure of communication which allows different parts of the company to make good decisions.
If management starts thinking they know more than their employee on a topic, something has gone terribly wrong. They hired someone who doesn't know what they are doing, or they are preventing information from flowing to where it is needed, or they are just being hubristic.
How much freedom you have within a company, the power structure and the actual composition of the work is a completely different discussion.
That's a rather naive statement. should be: s/startups/organizations/g
Heck, I would wager that >50% of startup employees are not even doing the work they think is most important for the success of the startup. They are doing the work their boss (who by definition is further removed from the facts) thinks is most important.
Heck, I wouldn't be surprised if >50% of startup employees are doing work they know is not important for the success of the startup.
That's what happens when you run a command and control organization. You lose >50% of your output, in exchange for guaranteed alignment of whatever is leftover. As long as you pay people <50% of what they are worth you can still make money.