Two scenarios, both from candidates who did everything right. First: weeks of a life handed over to a marathon of rounds, ending with the news that an internal candidate got the role — the process, in hindsight, was a formality wrapped around a decision that may have been made before it started. Second: a process scoped as a handful of rounds that quietly grew mid-flight, each new stage announced as though it had been there all along.
If you've been through either version, you know the specific flavour of that anger. It's not rejection anger. Rejection is a risk you signed up for. This is scope anger — the deal changed after you'd paid in, and everyone on the other side acted like that was normal. It is normal. That's the problem. So let's take it apart: why rounds multiply, how to spot the theatre version early, and how to cap what a process can take from you.
Rounds are deferred decisions
A round gets added when someone with power couldn't decide. That's it. That's the whole physics of it.
Hiring decisions are scary — a bad hire is visible, attributable, embarrassing — and in most orgs, nobody was ever criticised for adding a step. Adding a step looks diligent. So when a debrief ends in a shrug, the path of least resistance isn't “decide anyway.” It's “let's have them meet Priya's team too.”
Notice what's missing from that mechanism: any accounting of your time. The candidate's hours appear on nobody's budget. Every additional round costs the company a few calendar slots; it costs you an evening of prep, a day of nerves, and another week of your life in limbo — and that asymmetry is why processes bloat in one direction only. Nobody ever removes a round mid-flight. Growth is free; the bill goes to you.
This reframe matters practically, because candidates read added rounds as progress — “they keep wanting to see me, must be close.” Read it cold instead: an added round means the existing rounds didn't produce a decision. That's not momentum. That's a stall, dressed as due diligence. It can still end in an offer. But treat it as the yellow light it is, not the green light it pretends to be.
The theatre process
Now the harder case, the first scenario above: the process that was never real, run because policy demands external candidates be interviewed while an internal favourite waits at the finish line. It happens, it's miserable, and you can't reliably detect it — but there are tells worth reading in combination:
- The suspiciously specific role description — asking for an oddly precise mix of systems and history that reads like someone's current resume, because it is.
- Interviewers who don't sell — in a live process, good interviewers pitch the team almost involuntarily; in theatre, they run their script and stop, because there's nothing to win.
- Timelines that fog under direct questions — “we're still aligning internally” — since the real answer, we're waiting to see if the internal person accepts, can't be said aloud.
- Rounds that feel like laps — each conversation covering the same ground at the same depth, because the process needs volume for the file, not information for a decision.
Any one of these alone means nothing. Stacked together, start protecting your downside: keep other processes running at full speed, spend less prep on each additional lap, and treat the process as practice with a lottery ticket attached. The candidates most burned by theatre are the ones who paused everything else out of hope. Never pause everything else. Hope is not a scheduling principle.
Every real round can name its question
So what separates a legitimate long process from bloat? Not the count — the purpose per round. A defensible process is a set of distinct questions: can this person do the work (skills check), can they think beyond the work (manager conversation), can the people who'd sit near them stand them (team or panel), does someone senior need to bless it (exec chat). Each stage answers something the previous stages couldn't.
Which gives you a beautifully simple audit, usable from inside the process: for each upcoming round, can anyone tell you what question it exists to answer? Ask exactly that, politely, when a new stage appears: “Happy to — so I can prepare properly, what's this round evaluating that hasn't been covered?” A real process answers instantly: “the panel's about collaboration; the exec round is a formality for sign-off.” Fine. Prep accordingly. But if the answer is a vague “just another chance to get to know you,” you've learned the round has no question — and a round with no question is not evaluation. It's indecision, wearing a calendar invite.
The position, stated outright since the whole industry tiptoes around it: process length is not rigour. A company that can't decide after seeing your work, your thinking and your face several times isn't being thorough — it's outsourcing its fear of deciding onto your unpaid time, and the resulting hire is no better for it. The best-run teams decide quickly because their rounds have questions — speed and rigour aren't opposites; they're what each other looks like when the debrief works. Sprawl isn't a high bar. Sprawl is a broken debrief.
Capping your exposure
Practical protocol, start to finish. At the start, scope it with the walk-me-through-every-stage question from reading the interview process — and write the answer down; that written answer is your contract-in-spirit. Mid-process, when a new round appears, invoke the scope, warmly: “Of course — though I'll note we've now gone past the process as described. Is this the final conversation? I'm planning around other timelines.” You're not threatening. You're billing. Companies that respect candidates respond to that sentence with dates; companies that don't respond have told you the price of joining them.
And before any of it — decide your cap in advance, while you're calm. Sunk cost is a fog that thickens with every round; the version of you several stages deep will always find a reason to do one more. The version of you at the start, deciding “past this many conversations without a decision, I walk” — that version is the one whose judgment you should trust. Walking away from a stalled marathon isn't quitting the job. There was never a job on the table. There was a process, eating.
One last note on where the bloat is heading: companies are increasingly bolting automated rounds onto the front of these processes — AI screens, recorded assessments — while the back stays as indecisive as ever. Watch what the automation is used for. Tech that compresses the deciding would be worth celebrating; tech that just makes the queue longer and cheaper to run is the same stall at industrial scale. Same yellow light. Bigger intersection.
Back to all articles