I am imagining the two Bobs from Office Space, and two more Bobs come to visit them and ask what they would say they do around here.

Office Space needs a sequel something fierce, and something like that scene needs to be part of it. It should be called Open Office Space.

Along with Idiocracy, the smartphones, VR, and self-crashing cars edition.

I don't know what Mike Judge is waiting for...

Mike Judge did Silicon Valley already.

Mike Judge's current work is a soon-to-be reboot of King of the Hill.

King of the Hill is an American treasure

Always some misunderstandings about how consulting firms work whenever McKinsey comes up, so will shed some light as a former MBB consultant

Companies typically have a few "administrative functions" -- HR, Finance, Legal, Marketing -- in addition to Sales and R&D (Product / Eng). What some companies have (but not all) is a "Strategy" organization. The strategy organization, if a company has it, is usually <20 people, with a chief strategy officer reporting to the CEO. At a startup, founders are tasked with understanding the ins and outs of the business and designing a strategy based on that information. At a larger company, the CEO doesn't have the bandwidth to answer questions at an appropriate nuance (e.g., what is the first Latin American country that a food delivery app should launch in given consumer interest, competitive environment, local regulation, etc.). These questions would fall to a strategy team.

The reality is that there is not enough work year-round to employ strategy people all the time, so large companies would rather pay McKinsey $1m every couple of years to help do all the underlying research to answer questions rather than employ a strategy team at a higher cost per year.

Some other things we did for clients: 1. Provide some organizational benchmarks e.g., competitors give their sales people $5M targets per year, and you're only giving your sales people $3M targets. 2. Manage one-time projects e.g., F500 company acquiring another. Lots of (boring) administrative work of determining who reports to who once you merge these two companies 3. Provide some insight into new market segments a company can enter. For a company that sells widget X, what's the next widget they might be able to produce efficiently and sell to existing customers where there is a meaningful market and the company can leverage it's existing capabilities (related to the strategy piece above)

Lots of reasons a company might hire consultants, but generally no one in the organization wants to / has those responsibilities in their job description, so makes sense to do a one-time engagement.

I also found that even lower-level employees confide in consultants and consultants provide a fast-track to surface those issues to the CEO. Eg we were working for a consumer electronics company where the HW and SW orgs could not agree on anything and accused eachother of blocking their roadmaps.. Surfaced that to the CEO, and he made fixing that his top priority

Ah yes, organisational benchmarking. Is that the one where you tell your investment bank client that it’s not doing enough in mortgage-backed securities, just before the credit crisis hits?

Not that I’m bitter or anything (honestly), but for advisor and client alike it is surely the laziest form of strategy consulting

Meanwhile: https://www.politico.eu/article/france-relyies-on-consultant...

You win an award for having the only Hacker News comment about management consulting that is based in facts/reality and not entirely driven by ad hominem attacks in bad faith, congratulations

Who advised McK for this? Bain or BCG?

Or maybe they've started, what's the saying... "believing in thei-" no... "drinking their own champagne."

Dog-fooding!

I think "urine" was the noun you were looking for.

Consultants all the way down

McK is truly an evil company.

I have a friend suffering from depression and burnout and they asked him to wait till Friday to have his sick leave. Horrendous work culture and abuse.

My little glimpse into them revealed notably-bad management of internal projects and management culture of caring very much about appearing busy while not really doing much (IDK about "engagements", I assume they basically do what they need to for those, which is why they keep getting paid). Which is pretty funny considering what they sell. But what I saw may not have been a representative look at them.

That's my impression of it as well however it still leads to crazy hours and a shitton of work. My friend ended up going to sleep 5am for many consecutive days.

Right, didn't mean to imply your pal's slacking, meant to add on to the "horrendous work culture" bit. Hell, even for the ones in the position of not having much to do but needing to keep a full-to-bursting calendar for what I assume are defensive-against-severe-organizational-dysfunction reasons, that's not exactly a happy-go-lucky stress-free time.

I didn't take it that way, just meant as complement to your point :)

As a forward economic indicator, when the people who make their money effectively managing institutions start hedging against growth, that's pretty grim. If these people aren't winning, who is?

This is the absolute typical layoff story we're seeing at the moment - 2018- 28k employees, 2023 - 47k, layoffs? 1,400. That's crazy! Either structurally we've seen a massive movement of employment into these sectors, or these companies have grown way less efficient, and given the ZIRP I'm very worried it's primarily the latter. What changed in the last decade that meant McKinsey needed to double in size? Not much would be my guess.

It seems like GPT style tools would allow McKinsey et al to operate with much fewer employees. You only need the public facing folks who interact with clients, the other stuff could be done/accelerated by LLMs mostly.

They'll wait until on-prem LLM's can be done relatively cheaply/easily. The potential ramifications of a data breach would be staggering for a consulting firm.

Although it would be interesting to see what would happen if every client sued the firm at once.

I doubt they will have the competency to do on-prem competitive LLM anytime soon. They can use the Azure OpenAI service which I think offers better enterprise peace of mind.

They'll have the competency to try. Unless their tech org has changed a lot, in fact, I can just about guarantee they'll burn a lot of money trying (and failing, by having approached the whole thing all wrong from the outset).

Seems plausible that interacting with clients could be done by GPT too, with no reduction in value delivered.

The LLM is quite far away from even being able to shake client's hands though. I doubt any consultancy work involving large money can be 100% virtual.

Ok but you could have GPT on the client-side too.

This is the way

“ You only need the public facing folks who interact with clients, the other stuff could be done/accelerated by LLMs mostly.”

Maybe the unemployed people they always find on the street to do the actual project work are cheaper than LLM?

The LLM will have no overheads though, and will be much faster. Many advantages over and above just cost.

If it was the case, and it isn't, people would use the LLMs themselves and not pay top dollar for an intern to copy their question in chat gbt

Right, a lot of roles in consulting are political, the LLM isn't going to play politics. Companies aren't going to replace consultants with LLMs for their slide decks and advice, because the service the consultants are offering is persuasion.

I mean, the LLM is happy to give you the answer you told it you wanted in the first place, with lots of flowery language and justifications. We're really in core LLM competencies here.

But, sure, the customer can't _know_ it's just an LLM or it's not going to work -- what are they even paying for then?

But really, secretly replacing all consultants with an LLM seems just so perfect for these folks, whose "consulting" was already at the level of quality of an LLM regurgitating what everyone else is saying but making it sound pretty and customized to you. An LLM is _perfect_ for what this industry delivers.

Happily these are consultants who appreciate the value of mass layoffs, so I'm sure they'll appreciate the opportunity to improve their executive's bonuses.