Two hires failed before I blamed the job description

The first VP of Customer Success failed at month 9. Smart, polished, ten years at a name-brand company. He could not get the team to ship. His replacement came from the same résumé profile: smart, polished, similar tenure at a similar brand. He failed at month 7. The third hire was someone the first panel would have screened out. She has run the function for eighteen months, expansion revenue has climbed, and it's the strongest the team has been.

Two failures on the same role profile is not variance. Two failures is signal. The signal is that the role you defined doesn't match the role the company needs filled. Most growth-stage teams iterate the candidate and leave the spec alone. The third search runs against the same job description as the first two, maybe with a bullet point moved.

Here is how it goes wrong. The first JD gets written from what the role looked like at a bigger company. The candidates who match it have done that job at a bigger company. The bigger-company version rewards different instincts: more delegation, less hands-on work, more internal politics, less time with customers. Those instincts are the wrong ones for a company this size. The hire behaves the way the last place paid them to behave. The hire fails. The team swaps the candidate and keeps the spec.

The correction is to read the failure as data about the role, not the person. "He could not get the team to ship" looks like a hire problem. Underneath, it's a spec problem: the job needed someone who would still do part of the work themselves, in a company too small to close the gap by delegation alone. That is a different person than the JD was hunting for. The same trap sinks a promotion that quietly breaks a function: right title, wrong demands.

Most teams skip this read because it stings. Admitting the spec was wrong implicates the manager who wrote it. Admitting the candidate was wrong implicates only the funnel. The second is easier to say out loud, so it's the one that gets said.

The third hire works when the spec gets rewritten. The rewrite usually looks less impressive on paper: smaller names in the background, more functional depth, less management pedigree. A profile the first round would have passed on, because the first JD was tuned for the impressive answer instead of the right one. If two hires have failed the same way, stop interviewing and go rewrite the job. The candidate was never the variable. A team that has not shipped in six months is usually pointing at a definition, not a person.