Who is in the room, and why
A loop is not four copies of the same interview. Each seat is assessing something different, and often only one of them decides.
By the final stage you are usually meeting three to five people in sequence, and they have divided the assessment between them deliberately. Knowing roughly who does what lets you spend each conversation on the thing that conversation is for, instead of giving the same answer four times.
The hiring manager
Owns the decision and lives with the consequence. They are asking one question underneath all the others: can this person do the job, and will managing them be work? Be concrete about scope, delivery, and how you operate. This is the conversation in which to ask what success at month six looks like, and to listen carefully to how confidently they answer.
The future peer
Usually assessing whether you would be good to work beside — which in practice means whether you explain things clearly, admit what you do not know, and take input. They often have a veto in the debrief even when they have no formal say. This is also the most honest source of information you will get about the team; ask them what surprised them in their first three months.
The cross-functional interviewer
Someone from a team you would work with but not in — design meeting an engineer, finance meeting an operations lead. They are checking whether you can work across a boundary: do you understand their constraints, or do you treat their function as an obstacle? Speak about their work as a set of trade-offs rather than a service you consume.
The skip-level or founder
Often the shortest conversation and the least structured. They are sampling for judgement, ambition, and whether you think about the business rather than just the task. Have a view — about the market, the product, where the role goes — and be able to say it in two sentences.
Practical things that matter more than they should
- Ask at the start how long each session is. Pacing a 30-minute conversation like a 60-minute one is the most common avoidable loss in a loop.
- Take thirty seconds between sessions to write down what you were asked. By the fourth conversation you will not remember, and the debrief compares your answers.
- Eat and drink water when offered. Four hours of talking is physically harder than candidates expect.
- Assume everyone talks afterwards, because they do. Consistency across the loop is itself a signal.
The panel is not four opinions. It is one decision assembled from four samples, and they are taking different samples on purpose.