Structured Interviews: The Complete Employer Guide

Two interviewers meet the same candidate for forty-five minutes and reach opposite conclusions. Both are experienced. Both are confident. Neither can point to the evidence that produced their view, because they asked different questions, weighted different answers, and formed an impression in the first few minutes that the rest of the conversation quietly served. This is the ordinary condition of unstructured interviewing, and decades of selection research have found it to be a weak predictor of job performance.

Structured interviewing fixes this without turning the conversation into an interrogation. The same competencies, the same questions, the same rating scale, scored independently before anyone talks. It is more work upfront and considerably less work later. As a recruitment agency in Mumbai that has run assessment for over 500 client companies since 2001, we find the employers with the lowest early attrition are almost always the ones with the dullest, most disciplined interview kits. This guide sets out how to build one.

Why unstructured interviews fail

They fail for reasons that have nothing to do with the interviewer being unskilled.

• First-impression anchoring. A judgement forms in the opening minutes and the rest of the time is spent finding support for it. Interviewers experience this as insight.
• Similarity bias. We rate people who resemble us more highly — same institution, same city, same speech patterns. In Indian hiring this often shows up as institutional pedigree standing in for demonstrated capability.
• Non-comparability. If each candidate is asked different questions, the panel ends up comparing feelings.
• Articulacy over competence. Unstructured interviews reward people who describe work well. Some of them also do work well; the interview cannot tell you which.
• Question drift. By the fifth interview of the week, the questions have wandered and the standard has moved with them.

Structured interviews improve accuracy for a simple reason: they gather comparable evidence about defined requirements, and force it to be recorded before discussion contaminates it.

Step one: build a competency framework for the role

Start from the work, not the CV. Ask the hiring manager to describe what the person will actually do in their first year, and what separates someone who succeeds from someone who struggles. Then reduce that to four to six competencies. Fewer than four and you are not assessing enough; more than six and nothing gets probed properly.

A workable set for a mid-level delivery manager might be: technical judgement, stakeholder management, ownership and follow-through, people development, and structured problem-solving. Define each one in a sentence describing observable behaviour, not an adjective. “Communication” is not a competency; “explains technical trade-offs clearly to non-technical stakeholders and secures agreement” is.

Distinguish three categories while you are at it: must-haves at entry, teachable within ninety days, and nice-to-haves. Most inflated requirement lists collapse once a hiring manager is asked which items they would genuinely reject a strong candidate for. This exercise also produces a far better job description as a by-product.

Step two: write behavioural questions properly

Behavioural questions ask what a candidate actually did, because past behaviour in comparable situations is the most useful signal available. Hypotheticals ask what they would do, which measures how well they can imagine a good answer.

Use STAR to structure the probe: Situation, Task, Action, Result. Most candidates volunteer the situation at length and the action barely at all. The interviewer’s job is to push into the action — what did you specifically do, what did you decide, what did you say, what would you do differently. Prepare two or three follow-up probes per question so the interviewer is not improvising.

Step three: design a scorecard with anchored ratings

An unanchored 1-to-5 scale collects opinions. An anchored one collects assessments, because everyone is measuring against the same described behaviour. Write the anchors for 1, 3 and 5 for each competency; 2 and 4 sit between.

Two further rules. Require a written evidence note beside each score — the specific example the candidate gave — so that scores are traceable. And require a hire recommendation on a four-point scale with no neutral middle option, forcing interviewers off the fence.

Step four: allocate competencies across the panel

The most common waste in a five-stage process is five people asking about the same two things. Assign each interviewer two competencies to own and probe in depth. Everyone else leaves those alone.

Give every interviewer their allocation, their questions and their anchors in writing before the interview. An interviewer who reads the CV in the lift is not conducting an assessment. Our guide to the recruitment process step by step covers how to compress this calendar without dropping stages.

Step five: the debrief that avoids groupthink

The debrief is where a good process is most often destroyed. One senior voice speaks first, everyone recalibrates, and the independent evidence you spent a week collecting evaporates.

1. All scores submitted before the debrief. No exceptions, no editing afterwards. Late submissions are excluded rather than accepted.
2. No corridor conversations. “How did you find her?” between interviews contaminates the sample. Say so explicitly when briefing panels.
3. Junior speaks first. Then peers, then hiring manager, then the most senior person in the room.
4. Discuss divergence, not consensus. Where two interviewers scored the same competency 2 and 5, that gap is the most informative thing in the room. Go to the evidence each of them recorded.
5. Decide against the bar, not the shortlist. “Best of the three we saw” is how weak hires get made in slow markets. If nobody clears the bar, nobody clears the bar.
6. Record the decision and the reservations. Any residual concern becomes an onboarding objective for the first ninety days.

Interviewer training and calibration

Most Indian employers train interviewers once, informally, years ago. Two modest investments change outcomes materially. First, a mandatory two-hour session before anyone joins a panel, covering the scorecard, STAR probing, note-taking and the questions that must not be asked. Second, quarterly calibration: the panel independently scores the same recorded or role-played interview, then compares. Divergence surfaces the interviewer who marks everyone a 4 and the one who has never awarded a 5. Where you can, track score distributions and how each interviewer’s recommendations correlate with hire performance twelve months on.

Work samples and take-homes, done fairly

Work samples predict performance well because they measure the work rather than the description of it. They also generate the most candidate resentment when handled carelessly.

• Time-box explicitly and design for that box. If you say two hours, mark what two hours produces; candidates who spend twelve should not be rewarded for it.
• Pay for substantial work — anything beyond roughly three hours, or anything you could plausibly use.
• Keep it representative. Assess the work of the job, not a puzzle that filters for people with free evenings.
• Never use candidate output in your actual business. It is indefensible and travels quickly on AmbitionBox and LinkedIn.
• Mark blind against a rubric, ideally by two reviewers, with names removed, and give everyone who submits at least two sentences of feedback.
• Offer a live alternative. A ninety-minute paid working session suits candidates with caring responsibilities or demanding current jobs better than a weekend take-home.

Fairness and legal guardrails in India

Some questions are irrelevant to capability, expose you to discrimination and harassment claims, and damage your employer brand. As general guidance rather than legal advice, avoid asking about:

• Marital status, pregnancy, or plans to have children. Still routinely asked of women candidates in India, and indefensible.
• Caste, religion, region, or mother tongue. Language proficiency required for the job can be tested directly; identity cannot be an entry criterion.
• Age or date of birth beyond verifying eligibility to work.
• Health, disability or medical history, other than a neutral question about reasonable adjustments needed for the process or the role.
• Family background, parental occupation or household income. Irrelevant, and a proxy for social class.
• Political or union affiliation.

Handle current salary carefully as well. Anchoring an offer on a candidate’s existing pay perpetuates historic under-payment; asking for expectations against a stated band is fairer and works better. Our guide to CTC versus in-hand salary covers how to run that conversation without ambiguity.

Frequently asked questions

Turning your interview process into an advantage

A structured process costs a few hours of design per role and repays it in fewer failed hires, faster debriefs, better candidate experience and defensible decisions. It also does something less obvious: strong candidates recognise a well-run process and read it as evidence of a well-run organisation. In a market where good people have options, that matters.

Ace Corporate Services has completed 5,000+ placements for over 500 client companies since 2001, building competency frameworks, interview kits and scorecards with clients across 15+ industries and running assessment on their behalf through talent sourcing and search mandates. Call +91-22-67554705, email info@acecorpsers.com, or contact us to have your interview process reviewed before your next round of hiring.