Why Taxpayers Pay McKinsey $3M a Year for a Recent College Graduate Contractor
mattstoller.substack.com
mattstoller.substack.com
I feel like I write this response every 3-6 months on HN so I should just save it to Textexpander at this point. I hated working at McKinsey and left in under a year. It was a miserable fit for me and I have no illusions about the firm.
But they are very effective in certain ways. Yes the costs are high but so are the costs of making a mistake or failing to capture an opportunity where impacts are in the billions. Why does McKinsey (sometimes) provide value? One - an organization may be facing something critical which it only faces once every 20 years or once ever. Meanwhile, McKinsey has institutional experience it can draw on dealing with that issue hundreds of times. Second, when the issue is more of a white page strategic question you frankly have more analytical horsepower and broader perspectives than most organizations can field internally.
Are the results always great? Of course not. Bet some of you FAANG and co devs also know some ultra high paid developers and execs at your firms who sometimes strike out. The real world is messy and hard.
People saying McK gets paid millions to have a snot nosed 23 year old tell grizzled veteran operators what to do is either click bait canard or doesn’t understand how this really works.
I am not referring to the younger generation being magically more capable than a veteran technologist. I am referring to tech leadership in the government (say a general or colonel) being so technically inexperienced/naive that a novice technologist (the snot nosed kid) will probably be more capable in directing software projects. You can see a similar/identical sentiment mentioned during a scathing review of the military in the resignation letter of the pentagon’s first chief software officer as seen (bits and pieces at least ) here:
Chaillan announced his resignation in a blistering letter at the start of September, saying military officials were repeatedly put in charge of cyber initiatives for which they lacked experience, decrying Pentagon “laggards” and absence of funding.
“[W]e are setting up critical infrastructure to fail,” he said in his letter, which made only cursory reference to advances by China. “We would not put a pilot in the cockpit without extensive flight training; why would we expect someone with no IT experience to be close to successful? [ . . .]While we wasted time in bureaucracy, our adversaries moved further ahead.”
https://www.ft.com/content/f939db9a-40af-4bd1-b67d-10492535f...
(If you are paywalled try this https://12ft.io/proxy?q=https%3A%2F%2Fwww.ft.com%2Fcontent%2...)I'm a data scientist. I HAVE pitched ideas directly to fortune 500 ceo's in my early 20's. But it's a weird thing to assume this is hubris. Fortune 500 CEO's are not gods. They're just people. And often times they have some pretty important questions that one dude with a python script can answer that they just don't really have the organizational capability to answer otherwise. It would be great if they could do this in house, but many times they can't, and almost always we don't leave until there's someone on their side who's been upskilled to do it going forward. I would recommend looking into it if you're on the ML side and don't care for FAANG.
edit: also fwiw, the marginal cost of a BA / ASC in the provided rate table in this article is $1.7M / year, not $3M.
Obviously not in the taxpayers interest