Meh, past a certain base of technical understanding (enough to talk through solving a programming problem on the phone), it's all connections and soft skills. At Microsoft if you get an interview for an internship there's like a 50% chance you get the offer, but like 1% of the people who apply actually get the interview. And even though these companies have at-will employment, in reality it's INCREDIBLY difficult to get rid of people if they don't want to leave, and with big tech they often print enough money that it's either not worth the hassle of firing slight underperformers or it's not worth getting rid of proven engineers who have nothing to work on right now and having to find actually good new senior engineers to hire if and when the need for them arises. There's no industry that's really immune to nepotism/cronyism IMO.

Since Satya Microsoft has gotten much mellower in terms of performance ratings maybe to a fault; 80% of interns get full time offers and I personally know of a team of five senior engineers making collectively over a million in base pay alone whose sole collective responsibility is managing the UI buttons for a fairly minor product. The issue in an industry that actually produces stuff (as opposed to creating like financial instruments from pre-existing assets and selling those) is that it's IC-based, and sometimes it's justifiable resource-wise to just keep on senior engineers that you know for a fact will be useful in the future even if you have nothing for them to do right now rather than go through the hoops to find actually good new ones at that level if and when the need arises. Balmer didn't understand this and tried to run it like a financial firm instead of a tech company, which I'm honestly shocked didn't kill it altogether. I think what Google is doing isn't what Enron was trying to do; what Enron execs wanted was to light a fire under their employees asses, what Google is doing is judging long-term growth and whether it's worth it to keep on engineers that aren't really necessary to whatever product they're working on right now. Not excusing them, just saying big tech is often pretty bloated and often prints enough money that it's not an issue for them to keep people on rather than hiring new, but things seem to be getting bad enough to the point that they feel the need to take such measures.

At Google there's a culture of promotions and accolades being handed out for starting innovative shit, but not much for properly maintaining those initiatives. It's why they have all these products that die after two or three years, and why some of their executives made the jump from IC to management without knowing how the hell to sustainably manage. A good engineer can almost always be a good manager, but there are definitely some who simply can't or who don't get a proper crucible before being placed in a management role.

This is what people don't understand about cherry-picking data; with many things that don't have a direct and comparable representation (ie you get to choose what measurements make up the final results) there is almost always a combination of empirical measurements directly suited to the results you want to achieve.