Playbooks·02 Recruitment Operations
02 Recruitment Operations

Interview
Process Design

Better interviews start before the first question.

K

Kristina Golovko

MindDesign

7 min read

The evaluation system

Signals

What matters for this role

Stages

Who evaluates what

Ownership

Who leads each conversation

Calibration

Shared language for good

Interview quality is a design outcome, not a talent outcome. Structured processes surface better signal from every candidate.

Why this matters

Most interview processes are designed for comfort, not signal.

Interviewers ask what feels natural. What they enjoyed being asked. What they find interesting. The result is a conversation that feels good but collects almost no useful information about whether this person can do the job.

Signal-poor interviews create three problems: inconsistent evaluations across interviewers, bias-driven decisions that can't be explained, and candidate experiences that reflect poorly on the company.

Great interview design starts with the question: what does this person need to demonstrate to earn the offer? Everything else — the stages, the interviewers, the questions — flows from that.

Founder reality

Before designing any interview process, answer these:

01

What are the three most critical capabilities this role requires?

02

Which capabilities can be assessed in a conversation versus a practical exercise?

03

Which interviewer is best placed to assess each dimension?

04

What does 'strong signal' look like for each stage?

05

How will disagreements between interviewers be resolved?

If interviewers can't answer these before the process starts, they'll improvise answers during it — and the decisions will reflect that.

The design system

Four elements of a designed interview process

Apply to every role above individual contributor. Revisit after each hire.

01

Signals — define what you're evaluating before you meet anyone

For each role, identify 4–6 evaluative dimensions. Examples: technical depth, communication clarity, ambiguity tolerance, leadership instinct. Every interviewer evaluates a defined subset — not everything.

02

Stages — design a progression, not a repetition

Each stage should add new information. A phone screen should differ from a hiring manager call, which should differ from a panel. If two stages ask the same things, one is redundant.

03

Ownership — assign a lead interviewer to each stage

Every stage has one lead who is responsible for the experience, the evaluation, and the feedback. Co-interviewers support — they don't run the room. Clarity here prevents candidates experiencing incoherent conversations.

04

Calibration — align on what good looks like before the pipeline starts

Run a 20-minute calibration session with all interviewers before the first candidate enters. Review the role brief. Discuss what strong versus weak signal looks like for each dimension. Record the outcome.

Common mistakes

01

Too many interviewers

More interviewers create more noise, not more signal. Five people with no coordination produce worse decisions than two with clear roles.

02

No structured feedback format

Impressionistic feedback ('felt good', 'not sure') cannot be compared across candidates. Feedback must map to the dimensions defined in stage design.

03

Using the same process for every role

A senior engineer role and a sales role require entirely different evaluation logic. Copy-paste processes produce inconsistent signal.

04

Calibrating after the pipeline is full

Calibration done after candidates are already in the process anchors on the candidates you've already seen. It must happen before the first profile arrives.

Example scenario

An engineering-led AI company, 35 people, hiring their first VP of Product. Eight candidates were interviewed over six weeks. No decision was made.

What happened

Four different people ran first-round interviews with no shared brief.

Feedback used words like 'strategic thinker' and 'strong communicator' without examples.

Two candidates were eliminated for reasons not discussed before the process started.

The hiring team disagreed on what the role was actually for.

The redesign

Role brief rewritten. Three evaluation dimensions defined. Two interviewers per stage, with clear ownership. Calibration session run. New pipeline opened.

The outcome

First qualified candidate moved from screen to offer in 3 weeks. Decision unanimous. Onboarding started on time.

The interview is not a conversation — it's an evaluation.

Treating it like a conversation produces comfortable interactions and inconsistent decisions. Designing it as an evaluation produces better signal, faster alignment, and more confident hiring.