You've been conducting interviews for 20 years, and there are still candidates who disappoint—they accept the position, and after six months you're repeating the process. When you review your notes, all you had was "good attitude" and "SaaS experience." This gap between the interviewer's perception and the candidate's actual performance has a technical name: lack of predictive validityAnd the most robust methodology for closing it is called BEI: Behavioral Event Interview or interview for critical incidents.
At Voicit, we speak weekly with recruitment consultancies and professional headhunters who use BEI as a core technique in their process. The difference between those who apply it well and those who apply it superficially is enormous: predictive validity of 0.48 compared to 0.38 for a standard competency-based interview. This guide provides the complete framework: What is BEI, the 4 phases of a critical incident, the 5 levels of competence, how to formulate BEI questions by competence, and how to automate the process with behavioral indicators.
What is the BEI (Behavioral Event Interview)? The BEI interview, or critical incident interview, is a professional competency assessment methodology developed by David McClelland (Harvard, 1973) that asks the candidate to recount specific and real situations from the past where they have demonstrated a specific competency. It differs from the standard competency-based interview in that it delves into a single incident with 4 phases (Situation, Task, Action, Result) and assesses the level of mastery through observable behavioral indicators, not the candidate's opinions.
What you will find in this guide
- What is BEI and why does it work better than other techniques?
- Origin and scientific validity (McClelland)
- The 4 phases of a critical incident
- The 5 levels of competence
- Behavioral indicators: the key to rigorous evaluation
- 30+ BEI questions per competency (dropdown menus)
- How to evaluate and score BEI responses
- BEI vs Competency-Based Interview vs STAR
- 7 common mistakes when applying for the BEI
- How to automate BEI with AI + behavioral indicators
- Frequently Asked Questions
What is BEI and why does it work better than other techniques?
BEI stands for Behavioral Event InterviewIn Spanish it is also known as critical incident interviewIt is a competency assessment methodology based on a seemingly simple idea, but one that is radical in its implications: The best predictor of a candidate's future behavior is their past behavior in similar situations —not their opinions, not their intentions, not their model answers.
What differentiates BEI from other interview techniques:
- Look for specific incidents, not generic patterns. Instead of asking "How do you work in a team?", the EIB asks, "Tell me about a specific situation from the last year in which you had a major conflict with a colleague." One answer is worth more than ten generic ones.
- Go deeper with structured follow-up questions. Each incident is explored using four phases (Situation, Task, Action, Result). The interviewer doesn't accept the candidate's initial account: they delve deeper to understand exactly what happened.
- Evaluate against observable behavioral indicators. Not "good communication" as an impression, but specific predefined behaviors: "explained the situation in terms understandable to non-technical people", "verified understanding by asking questions".
- Quantify the level of competence. On a scale of 1 to 5 (Insufficient → Excellent), allowing comparison of candidates with objective criteria.
The result: an interview with scientifically validated predictive validity This significantly reduces the number of bad hires. That's why it's the technique used by professional headhunting firms, established talent acquisition teams, and high-level recruitment processes.
Origin and scientific validity of BEI
The BEI interview was developed by David McClelland, psychologist at Harvard University, based on his seminal article «Testing for Competence Rather Than Intelligence» (American Psychologist, 1973). McClelland argued that traditional intelligence tests and CVs were poor predictors of job performance, and that it was necessary to evaluate skills: combinations of knowledge, skills, and behaviors that are manifested in real-life situations.
His disciple Lyle M. Spencer He later formalized the BEI methodology in the book "Competence at Work" (1993), where he detailed the four phases, evaluation criteria, and behavioral indicators. Today, it is the standard methodology in competency assessment programs at companies such as McKinsey, BCG, Accenture, and most reputable recruitment consultancies.
The predictive validity of BEI has been confirmed by multiple meta-analyses:
- Schmidt & Hunter (1998): BEI-type structured interviews achieve a predictive validity of 0.51, compared to 0.38 for unstructured interviews.
- Huffcutt, Conway, Roth & Stone (2001): Behavioral questions about the past outperform hypothetical situational questions in validity for skilled positions.
- Spencer & Spencer (1993): 70% success rate in predicting superior performance when BEI is rigorously applied with behavioral indicators.
The 4 phases of a critical incident
The core of the BEI interview is the exploration of critical incidentA specific situation from the candidate's past in which the competency we want to assess was demonstrated. The exploration is done in four successive phases that progressively deepen the understanding.
The interviewer asks the candidate to identify a specific past situation where they demonstrated the competency. Describe the actual incident, provide the approximate date, and give the context.
What exactly was the candidate's responsibility in that situation? Separate "the team did" from "I was responsible for."
The specific steps the candidate took. Use first-person verbs. Ask follow-up questions until you have real, not generic, details.
Measurable impact of the candidate's actions. Metrics if available, lessons learned, what changed. Without this, it's just narrative.
The 5 levels of competence in the EIB
Each competency is assessed on a 5-level scale ranging from absence or counterproductive behavior to expert mastery. Level assignment is not subjective: it is based on the behavioral indicators observed in the candidate's response (we will see this in the next section).
The target level for each competency depends on the position: for a junior IC, level 3 is sufficient in most competencies; for a middle manager, level 4; for a C-level, level 5 in the critical competencies of the role.
Behavioral indicators: the key to rigorous evaluation
Here's the real difference between a well-applied BEI and a superficial BEI. behavioral indicators These are observable behaviors, predefined by competence and level, that the interviewer looks for in the candidate's response. They are not the interviewer's opinions: they are verifiable facts in what the candidate says.
For example, for the competency "Customer Communication" at level 4, the behavioral indicators could be:
30+ BEI questions per competency
These are the batteries of BEI questions organized by competency. Each one is formulated to extract a real critical incident. If the candidate answers in the abstract, ask questions with: "Can you tell me about a specific case from the last 12 months?".
LeadershipLeadership and people management
- Tell me about the best team you've ever led. What exactly did you do to build it that way?
- Describe the most difficult conversation you've had with a team member in the last year. How did you prepare for it, what did you say, and what happened afterward?
- Give me a specific example of a time you had to fire someone. How did you arrive at the decision and how did you handle it?
- Tell me about someone you promoted in the last two years. Why did you choose them, and what did you do beforehand to prepare them?
- Describe a situation where a technically brilliant member was toxic to the team. What did you do, and how long did it take?
DecisionDecision making under pressure
- Tell me about the most difficult decision you've made in the last year. What information did you have, how did you decide, what happened?
- Describe a decision you made with clearly incomplete information. How did you assess the risk?
- Give me an example of a data-driven decision you made against your intuition. What happened?
- Tell me about a decision you made against the team's consensus. How did you defend your position?
- Describe a decision that was unpopular at the time but proved to be correct over time.
TeamTeamwork and collaboration
- Tell me about a major conflict you had with a colleague in the last year. What did you do, and how was it resolved?
- Describe a cross-functional project where you had to collaborate with people outside your team. What was your specific contribution?
- Give me an example of a time you changed your opinion thanks to a colleague's input. What changed in you?
- Tell me about a time you defended an absent colleague at a meeting where he was being criticized.
- Describe a situation where a colleague wasn't fulfilling their responsibilities and it directly affected you. What did you do?
CustomerCustomer focus
- Tell me about a particularly difficult client you've dealt with. What happened, how did you handle it, and where did the relationship end up?
- Describe a time you had to say "no" to a reasonable request from a client. How did you do it?
- Give me an example of a time a client asked you for something you knew wasn't good for them. What did you do?
- Tell me about the last time you lost an important client. Why did it happen, and what would you have done differently?
- Describe a time when your intervention turned a complaint into a recommendation. What exactly did you do?
AdaptationAdaptability to change
- Tell me about the most important change your company has experienced in the last year. What was your role in it?
- Describe a time your role changed overnight. What did you do in the first 30 days?
- Give me an example of a time you had to learn something completely new under pressure.
- Tell me about a strategic decision your company made that you disagreed with. How did you implement it?
- Describe a significant professional failure from the past year. What did you learn from it? What changed in you?
CommunicationCommunication and influence
- Tell me about a time you had to explain a complex concept to someone with no technical background. How did you approach it?
- Describe a presentation you gave that didn't go as planned. What happened? What would you change?
- Give me an example of a time you managed to influence a decision where you had no formal authority.
- Tell me about a difficult email you had to send (firing, bad result, conflict). How did you write it?
- Describe an important negotiation in which you participated. Your role, the concessions, the outcome.
How to evaluate and score BEI responses
The EIB assessment is not subjective. It follows a 4-step process:
- Verify that the incident is real and specific. If the candidate responds in the abstract ("I always try..."), there is no incident and therefore no evidence. Follow up until you get details. If after two follow-up questions they still respond in the abstract, note: "does not provide behavioral evidence" and move on to the next section.
- Identify positive behavioral indicators. Mark the predefined behaviors that the candidate explicitly mentions. Do not invent or infer them: only those that appear literally in the answer.
- Identify negative behavioral indicators (derailments). Behaviors that contradict the competition carry more weight than positive ones because they reveal limitations.
- Assign level (1-5) according to the combination. More positive indicators without negative ones = high level. Negative indicators present = medium or low level depending on severity.
BEI vs Competency-Based Interview vs STAR
There is frequent confusion between these three techniques. They share common DNA (behavioral questions) but differ in depth, rigor, and purpose.
- Delve into 2-3 incidents in 60 minutes
- Evaluate using predefined behavioral indicators
- Level 1-5 by competency
- Predictive validity 0.48
- Used in professional consulting firms and headhunting
- Framework for structuring responses
- It covers more skills in less depth
- Without a standardized level scale
- Predictive validity ~0.40
- Standard in general HR interviews
- An open-ended question based on competition
- Without structured in-depth study
- Subjective score of the evaluator
- Predictive validity 0.38
- Most internal processes
In practice: BEI is the standard for reputable recruitment consultancies and critical skills assessment processes. If you're hiring someone who will have a direct impact on results (any skilled role), BEI will give you a better decision than the other two techniques.
7 common mistakes when applying for the BEI
After reviewing thousands of interviews with professional consultants, these are the most frequently repeated mistakes—and those that invalidate the methodological rigor of the BEI:
- Accept hypothetical or generic answers. "I always try..." is not an incident; it's an opinion. If no specific case emerges after two follow-up questions, competence has not been demonstrated.
- Not having behavioral indicators defined before entering. Without predefined indicators, the evaluation is opinion, not methodology. Define them with the client or hiring manager before the first interview.
- Delving too deeply into too many skills. BEI requires 15-20 minutes per incident. In 60 minutes, no more than 2-3 competencies can be rigorously assessed. Two good ones are better than five bad ones.
- Skip the Task phase. This is the most important phase to distinguish between "I was on the team that..." and "I was responsible for...". If you don't delve deeper, the candidate may take credit for the team's achievements.
- Give points during the interview. It generates confirmation bias. It records indicators live, scores them later.
- Do not look for indicators of derailment. Derailments (counterproductive behaviors) outweigh positive ones. A single serious derailment can invalidate several positive indicators.
- Confusing BEI with STAR. STAR is just one way to structure the questions. BEI is a complete methodology with predefined indicators, levels, and an evaluation process. Applying only STAR without the evaluation component is a standard competency-based interview, not BEI.
How to automate BEI with AI + behavioral indicators
The traditional barrier to the EIB for many consultancies has been the operating costImplementing it effectively requires a trained interviewer, rigorous live note-taking, post-interview evaluation with indicators, and a structured report to the client. A well-conducted BEI interview can take 90 minutes of total work (60 minutes for the interview + 30 minutes for the report).
AI, when applied correctly, changes this equation. The current professional workflow:
- Pre-interview (10 min): The consultant configures the 3 critical competencies in Voicit, with the behavioral indicators (positive and derailments) predefined for each level.
- Interview (60 min): The consultant asks the BEI questions with explicit consent for recording. Their only task is to listen and delve deeper; they do not take notes.
- Post-automatic interview (5 min): Voicit transcribes, identifies speakers, and applies the evaluation engine. The AI detects positive behavioral indicators and deviations in the transcription, weights them by level, and delivers a pre-filled scorecard with literal textual evidence of the candidate (the exact phrases that justify each score).
- Consultant review (10 min): You review the level assignment, adjust if necessary, and add observations about nonverbal communication or context that the AI does not capture.
- Customer report (automatic): generation of a professional report using the documented BEI methodology.
From 90 minutes to 25 minutes without sacrificing methodological rigor. And with an added benefit: the recording serves as auditable evidence of the process, defensible to the client or in any subsequent review.
Automated BEI with behavioral indicators
Voicit has just integrated the complete methodology: configure your competencies with behavioral indicators, and AI rigorously evaluates each candidate's response, not just as a generic summary. It's professional BEI applied at scale.
Frequently Asked Questions about the EIB Interview
What is the difference between BEI and the standard competency interview?
BEI is a more rigorous methodology within the competency paradigm. Standard competency-based interviews use behavioral questions but lack predefined indicators and a standardized rating scale. BEI predefines behavioral indicators (both positives and derailments) for each competency and level, reducing interviewer bias and increasing predictive validity from 0.38 to 0.48.
How long does an BEI interview last?
A full BEI interview lasts 60-90 minutes to assess 2-3 competencies in depth. Each critical incident requires 15-20 minutes of exploration using the 4 phases (Situation, Task, Action, Result). To assess more competencies, it is best to divide the assessment into two sessions.
How many skills can be assessed in an BEI interview?
In a 60-minute interview: 2-3 critical competencies maximum. In 90 minutes: 3-4 competencies. Wanting to assess more is the number one cause of poor BEI implementation: it only scratches the surface, and behavioral indicators are not rigorously collected.
Is BEI suitable for junior profiles or only for senior positions?
BEI works for both, but with adjustments. For junior profiles, the incidents will come from academic projects, internships, or first jobs—less in-depth but equally valid. For senior profiles, the incidents are richer and allow for easier assessment of levels 4-5. The methodology is the same; what changes is the depth of the available examples.
What do I do if the candidate doesn't remember a specific incident?
Three tactics in order: (1) narrow the time horizon ("and in the last 3 months?"), (2) slightly change the scope of the task ("think of a project where you had autonomy"), (3) accept that the task is not proven and move on to the next section. If the candidate can't find any real incidents, that's information in itself.
Can the BEI be applied in online interviews or only in face-to-face interviews?
It works equally well in person, online (Meet, Teams, Zoom), and by phone. The methodology is independent of the format. Online, the main advantage is the ease of recording with consent, which allows for calm re-analysis of incidents after the interview.
Can I apply for the BEI without specific training?
Basic BEI is accessible, but advanced BEI requires training. The principles (behavioral questions, four phases of an incident) are readily available. The training-intensive aspect lies in designing behavioral indicators by competency and level, and rigorously assigning scores. Tools like Voicit, with behavioral indicators predefined by the methodology, reduce this barrier for consulting firms applying BEI without an in-house team of industrial psychologists.
How do I combine BEI with psychometric tests in the selection process?
Recommended workflow in professional consulting firms: (1) CV screening + phone screening, (2) psychometric cognitive aptitude tests for the top 30%, (3) BEI interview for the top 10-15 candidates focusing on critical role competencies, (4) final panel with the client. BEI and psychometric tests are complementary: the tests measure potential aptitude; BEI measures how that aptitude has been demonstrated in real-world situations.
Actionable summary
If you only take away three ideas from this guide:
- The BEI assesses the actual past, not opinions. Each competency is explored through a specific critical incident, delving into the four phases (Situation, Task, Action, Result). If there is no specific incident, there is no evidence.
- Behavioral indicators are what separates BEI from standard competency-based interviews. Defining them before entering the room turns the evaluation into a methodology, not an opinion.
- AI changes the operating cost of implementing BEI at scale. Voicit It integrates BEI questions and behavioral indicators into its assessment engine: it records the interview, identifies positive indicators and derailments in the transcript, and delivers a pre-filled scorecard with literal textual evidence from the candidate.
