A leader inside a large, well-known company recently asked a Big Four firm to evaluate their AI readiness. I am withholding the company and the names, because the pattern matters more than the specifics, and I have seen this exact pattern more than once.
What was the actual problem the AI evaluation was supposed to solve?
On paper, an AI strategy gap. In practice, something much harder: a leadership group that was dysfunctional and siloed, each member protecting their own business unit, publicly cordial and privately hostile to one another. Any AI proposal coming from this particular leader was going to be rejected before it was even heard, because the room had bigger unresolved political problems than AI, even though AI could genuinely have helped address some of them. This is not an unusual pattern: in RAND Corporation’s study of why AI projects fail, 84% of the AI practitioners interviewed cited leadership-driven decisions, not the technology itself, as the primary reason AI initiatives fail.
What did the consulting firm actually recommend?
To their credit, the honest answer: run a parallel change-transformation project alongside the AI project, because if the two are not synchronised, neither one works. That is correct advice. It is also advice from a firm that has already been inside this company, already knows this dynamic, and has had every opportunity to raise it before being asked to evaluate AI readiness specifically.
So why hire the same firm again?
This is the question worth sitting with. Only 12% of CEOs report that AI has delivered both cost savings and revenue benefits, according to PwC’s 29th Global CEO Survey, and most of that 88% are being advised by somebody. If you are going to your big, expensive consulting company for an AI strategy, that consulting company has already failed you. They are already inside the business. They already know it. If a working strategy has not already been delivered, or if the company is already losing ground to competitors on AI, the firm being paid to keep you ahead has already failed at the one job it was hired to do. Why are you paying somebody who has already failed you to fail you more? I go into the structural reasons this keeps happening in Strategic AI Advisor vs Big Four Consulting.
What should replace that relationship?
Not necessarily another large firm carrying the same incentive to protect an existing engagement. Someone whose mandate is defined and finite, with no multi-year relationship to defend, has a different reason to tell you the truth the first time.
If your organisation is still paying the firm that has not fixed this yet, that is exactly the conversation I have with a limited number of senior leaders each quarter. Work with Thomas.
Questions this article answers
Why did an AI evaluation actually fail at this company? Because the real problem was not AI readiness. It was a dysfunctional, siloed leadership group whose political problems were bigger than the AI question, and any AI proposal was rejected before being heard.
Was the consulting firm’s advice wrong? No, the advice itself (synchronise a change-transformation project with the AI project) was correct. The problem is that the firm giving it had already been inside the company long enough to know the dynamic and had not raised it earlier.
Why not just hire the same trusted firm again? Because if a working strategy has not already been delivered, or the company is already losing ground to competitors, the firm being paid to prevent that has already failed at that job.
What should replace an underperforming incumbent relationship? Someone with a defined, finite mandate and no existing engagement to protect, which changes the incentive to tell the truth the first time.
Thomas Anglero is a Strategic AI Advisor (MerkabaPhi AS, Oslo), with 450+ keynotes across 30+ countries. Enquiries: anglero.com.