A new manager joined a fast growing startup and was told she would start interviewing candidates within two weeks. Her training was a 45 minute call with a recruiter and a Notion doc. On her third interview she realized she had been asking the same question three different ways without noticing.
This is normal. The standard onboarding for an interviewer is observation followed by improvisation. You shadow a couple of loops, then you are on your own. The cost of this approach is hidden because nobody measures the quality of a new interviewer's first 10 calls.
The new playbook is shorter and more rigorous at the same time. It works in three steps.
Step one. Before the first interview, the new interviewer reviews three short clips from past loops at the company. Not full interviews. Clips of moments. The moment a senior interviewer redirected a candidate who was monopolizing the conversation. The moment another interviewer surfaced a competency gap with one well placed question. Pattern matching beats theory.
Step two. The first three live interviews run with an AI shadow. The system flags coverage gaps and questioning patterns in real time. The interviewer sees the prompts. No senior person has to sit through the loop.
Step three. After each of the first 10 interviews, a 5 minute review. What worked, what to adjust. By interview 10 the new interviewer is running cleanly.
Mazle was built to compress this curve. New interviewers ramp in days, not quarters. The investment compounds because their first hires set the standard for everyone they will hire after.
The cost of a bad first interviewer is invisible. Stop paying it.
