Most failed chief of staff hires aren't a bad hire. They're a bad setup that a good hire couldn't survive. I've watched this from both sides of the table now, and the same five mistakes show up almost every time.
1. Hiring for the title, not the bottleneck
"We need a chief of staff" is not a scope. Losing meetings, a board deck built at midnight the night before, three strategic projects stalled between check-ins: those are scopes. Hire without naming the actual bottleneck, and the person you hire spends their first month guessing what you meant, which is expensive at any rate.
2. Skipping the decision-rights conversation
Vague authority breaks in both directions. Given too little, a capable CoS sits on their hands waiting for permission that was never going to be denied. Given too much with no agreement, they make a call that was actually yours, and trust takes a hit it didn't need to. Write down, before day one, what they decide outright, what they draft for your sign-off, and what always comes to you first. It's a fifteen-minute conversation that prevents months of friction.
3. Treating them as a senior EA on day one
Calendar and inbox work is real, but it's not the reason the role exists. Hire someone for judgment and strategic follow-through, then spend their first quarter on scheduling, and you've bought a good EA at a chief of staff rate. If you genuinely need both, that's a fair scope: just name it up front, because a CoS hired under a different premise will burn out on it or leave.
4. Hiring on interview performance instead of artifacts
Confidence in a conversation is not evidence of an operating system. Ask to see redacted versions of what they've actually run: a decision log, an initiative tracker, a leadership-meeting agenda. Someone who's really done this can produce them in the interview. Someone who's talked about it in adjacent roles usually can't, and that gap tells you more than another round of questions will.
5. No checkpoint at day 90
Scope that was right in week one is rarely still exactly right by month three: trust has either grown or it hasn't, and the company has usually changed shape a little regardless. Founders who skip the 90-day conversation end up either under-using someone who's earned more scope, or living with friction from a mismatch nobody's revisited. Put the checkpoint on the calendar when you sign the engagement, not after something's already gone wrong.
The fix underneath all five
Every one of these comes back to the same root cause: treating the hire as a title to fill instead of a system to install. Name the bottleneck, write the decision rights down, ask for the artifacts, and put a real checkpoint on the calendar. That's the whole fix, and it's less work than recovering from any one of the five mistakes above.
If you're vetting candidates right now, here's what to actually test for. And if you'd rather skip the search and just scope it against your real week, book a free 30-minute call.