[sevensor]

This is a surprising take, considering the consultants I’ve worked with never took the trouble to learn anything about the business despite sitting shoulder to shoulder with us for weeks, and then proceeded to give us spectacularly terrible technical advice. “You need machine learning for this. Trust us, we’ve done this before.” They hadn’t, and it became clear they’d wasted our time and hadn’t listened to a word we said.

[whatever1]

There is nobody left, old enough, to remember that the last project with that consulting firm failed.

[jumping_frog]

Are we living in middle ages that electronic records don't exist of what happened in the past engagement. Maybe one can locate the last McKinsey dossier on OneDrive? The problem is and always was of incentives of people involved.

[whatever1]

Would you trust a random soggy doc for which you don't know the person who wrote it and their competence, or the flashy deck from McKinsey? The answer is 99% of the times: McKinsey.

Only time that answer is different is when a person in power is still around and they pretty much veto the engagement with a consulting company.

[ljm]

Don’t even need that trust when your org probably has some intensely bureaucratic procurement setup where past performance has zero influence on a new tender, so long as they can undercut the competition.

Their reputation doesn’t come from the quality of their work but their size: they can just muscle their way in.

[Scalestein]

Generally there Firms prize "first principles" thinking and don't put a ton of effort into leveraging previous engagements. Think of it like how engineers will throw out a solution and rebuild it rather than understand and build upon it.

It is also quite difficult to surface useful information amongst all the noise in the giant archive of documents

[throw4950sh06]

Oh no, that's not true at all. They care about data, and they will actually tell you themselves that they already did it for you - right after convincing you that you need to do it again.