What makes you think productive work is what consulting companies are selling? They're there for laundering accountability. When you bring in consultants to roll out your corporate AI strategy, and it all falls apart in a few years, you can say, "we were following best practices, nobody could have anticipated X," where X is whatever failure mode ultimately tanks the AI strategy.

Do you think that it's possible in principle to have a better or worse corporate AI strategy? I do, and because I do, it seems clear that companies paying top dollar are doing so because they expect a better one. There's no reason to pay KPMG's rates if all you need is a fall guy.

Most criticisms I see of management consulting seem to come from the perspective, which I get the sense you subscribe to, that management strategy is broadly fake so there's no underlying thing for the consultants to do better or worse on. I don't think that's right, but I'm never sure how to bridge the gap. It'd be like someone telling me that software architecture is fake and only code is real.

I'm willing to believe that one can be better or worse at management, and that in principle somebody could coach you on how to get better.

That said, how would we measure if our KPMG engagement worked or not? There's no control group company, so any comparison will have to be statistical or vibes-based. If there is a large enough sample size this can work: I'm sure there is somebody out there that can prove management consulting works for dentist practices in mid-size US cities or whatever, though any well-connected group that discovers this information can probably make more money by just doing a rollup of them. This actually seems to be happening in many industries of this kind. Why consult on how to be a more profitable auto repair business when you can do a leveraged buyout of 30 of them, make them all more profitabl, and pocket that insight yourself? I can understand if you're an poorly-connected individual and short on capital, but the big consulting firms are made up entirely of well-connected people who rub elbows with rich people all day.

Fundamentally, there will never be enough data to prove that IBM engaging McKinsey on AI in 2025 will have made any difference in IBM's bottom line. There's only one IBM and only one 2025!

The fall guy market is very sensitive to credentials. I hired Joey Blows from Juice-My-AI just hasn't that CYA shield of appoval.

Given that "design patterns" as a concept basically doesn't exist outside of Java and a few other languages no one actually uses, I'm apt to believe that "software architecture is fake and only code is real".

Design patterns (as in commonly re-used designs that solve commonly encountered problems) exist in every language used enough to have commonly encountered problems. Gang-of-Four style named design patterns are mostly a Java thing, and repeatedly lead to the terrible outcome of (hopefully junior) developers trying to find a problem to use the design pattern they just learned about on.

you hire consultants so you can cut staff and quality, but the CEOs were already going to do that.

Consulting companies don't sell productive advice. They sell management insurance.

I think this is the kind of logic you wind up with when you start with the assumption that the Big 4 tell the truth about absolutely everything all the time