5 terrible job interview questions, and 5 science-backed alternatives

Not all interview questions are created equal. If you’re a veteran recruiter or talent acquisition manager, this may sound obvious, but the reality is that many companies (and their hiring managers) run interviews like an informal conversation, in which so-so questions are selected at random, and candidate responses are recorded haphazardly. 

This has led to a general lack of confidence in interviewing as a practice: An Aptitude Research and Sapia.ai report from 2022 found that 33% of companies aren’t confident in the way they interview, and 50% believe they’ve lost talent due to their processes. Statistically speaking, it’s likely that your company or clients fit into these cohorts.

There are two proven fixes to shoddy interviewing: The right structure, and the right questions. Before we share our list of interview questions as recommended by our personality scientists, let’s quickly cover the structured interview and its benefits.

Why you should only be using structured interviews

Many (if not most) of us believe that resumes and past experience are the best indicators of future employee performance. Surprisingly, that’s not true: The best way to find top candidates (with 26% of predictive success) is the structured interview; conversely, past job experience accurately predicts just 3% of an employee’s on-the-job performance.

When you run AI-enhanced structured interviews, you devise a consistent set of questions that are asked to all candidates without any variation. Responses to questions are generally entered into a rubric, and are then scored against consistent criteria. The benefits of structured interviews are numerous, but the crucial reason for using them, especially in the context of AI interviews, is that you’re minimising variables; all candidates are compared fairly to one another.

Check out this quick video to learn more about the benefits of structured interviewing.

Our steadfast belief that past experience dictates the future, while understandable, has led to the proliferation of academic questions that trap and stymie candidates, mainly because they’re based on difficult concepts that are near-on impossible to articulate without preparation. 

For example, it’s not uncommon for marketing candidates to be asked something like, “Tell me how you’d adjust ROAS (Return on Ad Spend) numbers to compensate for unexpected reduction in impression share”. Should the candidate know what ROAS and impression share are? Sure. Can you reasonably expect them to solve an unexpected puzzle while you stare them down? No.

Instead, you want to ask questions that speak to behaviours. And, ideally, you want a reliable framework for assessing responses so you can adequately analyse behaviours (more on this later).

1. Don’t ask: Tell me how you’d respond if [specific situation] occurred.
   Instead, ask: Tell me about when something went wrong at work, and you had to fix it. How did you go about it?
   

   This is similar to the marketing example above. Although it’s tempting, you want to avoid questions that attempt to gauge technical proficiency. Such questions are best saved for pre- or post-interview assessments.

2. Don’t ask: Why do you want this job?
   Instead, ask: If you could do anything in the world, what might it be?
   

   No one has ever had an original response to the question of ‘why’. Responses are laden with bias, and most hiring managers expect some grandiose (and frankly, sycophantic) response. Conversely, the question of ‘if’ is far more telling – it speaks to a candidate’s openness, which is a valuable (and measurable) trait for problem-solving.

3. Don’t ask: What would your former manager say about you?
   Instead, ask: Describe a time when you struggled to build a relationship with someone important. How did you eventually overcome that?
   

   Former managers are not reliable character witnesses when their feedback is relaid second-hand by the candidate. You’ll never get a true response here. Instead, try to understand how a candidate overcomes difficult relationships. Answers will clue you in to a candidate’s levels of emotionality, extraversion, and empathy.

4. Don’t ask: What is your biggest weakness?
   Instead, ask: Tell me about a situation in which you have had to adjust to changes over which you had no control. How did you handle it?
   

   Candidates will not accurately describe their weaknesses to you. As with the questions above, you’re asking a question that will result in a biased sample. Instead, ask about a common point of pressure that we all experience in working life. If someone is adaptable and flexible to change, their responses will show you – and you’ll have learned something much more valuable.

5. Don’t ask: What is your biggest strength?
   Instead, ask: How would you change an institutional “this is always how we do it” attitude, if you felt there was a better approach?
   

   These questions don’t seem interchangeable on the face of it, but it’s far more valuable to know how a candidate can use their skills and behaviours to propel your company forward, than to know how they’d big-up themselves. Legacy thinking is a company-killer, and you want people who can think laterally to solve problems with new solutions. So find out if they can do that.

Go beyond great questions

These questions will serve you well in your interviews, but at a certain point, they’re only as good as your interview structure and analysis capabilities. Sapia’s Ai Smart Interviewer is designed to ask these questions – plus hundreds of others – and assess candidate responses to build accurate behavioural profiles. This means you get high quality candidates without having to bother with a long and inconsistent interview process.

Want even more questions designed by our people scientists and proven in the field? Download our HEXACO job interview rubric for free, here.

About Author

Laura Belfield

Head of Marketing