Most large firms split the work: the partner who wins it, and the junior team who run the search once you have signed. So you are paying for the judgment of the person you met at the pitch and getting the output of someone you have not.
Ask a technology leader what they built and you tend to hear about the migration, rarely the margin it protected. That distance between the work and what it was worth to the business is where I have watched most senior technology hires go wrong.
In experimental stock markets, prices tracked true value 58 per cent more accurately when the room was ethnically diverse, because homogeneous traders trusted each other too readily and mispriced together. An engineering team that thinks alike makes the same correlated errors, and they get expensive.
A bad brief rarely produces an obvious mis-hire. It produces someone who performs alright for 12 to 18 months, until the gap between the role as written and the business as it runs becomes impossible to ignore, well after a CIO-level search has cost over $500,000.
The exec who closes well in a PE process has already decided before the offer lands. The ones who wobble were often never committed, and catching that too late costs the search four to six weeks it rarely gets back.
The AI models a business runs today won’t be the ones it runs in two years, and whoever picked them won’t be where the lasting value sits. That’s the data layer underneath: architecture, governance and infrastructure that take years to build and decide whether the investment pays back.
