Using AI for Mock Interviews
As a job interview coach with previous roles in the tech industry, it’s a given that I’m always interested in AI and playing with new technology. I’ve been intrigued by everyone talking about having AI mock interview you and using that feedback to work on your interview skills, as well as the fact that so many companies are using AI these days to identify and narrow down their overloaded candidate pool. I conducted my own small study, and these are the takeaways I hope both job seekers and hiring managers will find helpful when considering using AI for their own purposes.
TLDR: Using AI is helpful at a high level for assessing functional skill, but it was not able to identify nuances, yellow or red flags in my responses. It was positive and encouraging, but in a way that was “nice” vs. “kind”. When you’re looking for help with interviewing, this is only going to get you so far and may give you the wrong impression of your actual skill level.
You could go really in depth with using AI for this purpose, so I tried to find the middle ground with the script I used knowing that some folks may spend more or less time on this. I told Chat GPT I was going to be interviewing for a Senior Program Manager position at a large tech company, and asked it to conduct a mock interview with me, asking 3 behavioral interview questions plus follow up questions, followed by feedback I should use to improve my interview skills. (When trying this out, you can also provide details like the job description of the actual role, information about the company, or your resume, to get a little more personalized)
Then, I tested answering questions multiple ways, and this is how it responded to me:
1. Answering a question as I truly would during an interview using the STAR Method (if you don’t know what this is, check out my previous blog posts, google it, or reach out to me!)
a. AI told me my answer needed no adjustment and it gave me specific praise about what worked well. This was helpful validation.
2. Answering a question with a little less detail, but still using the STAR Method structure
a. AI told me I had a great response and gave me a little feedback about areas to “consider” expanding on and providing more detail around. Helpful food for thought, but didn’t give me the impression I needed to change anything about my response.
3. Answering a question in great detail while including major yellow and red flags specifically about relationships with coworkers
a. AI gave me 100% positive feedback with follow up questions that did not dive deeper into these aspects. Human interviewers understand that “soft skills” like earning trust are more difficult to teach someone and they would absolutely want to hear more about this if I mentioned it an interview.
4. Answering a question very badly – throwing a coworker and her manager under the bus
a. AI thanked me for my honesty and said I did the best I could at the time, and did not give any feedback about how an actual interviewer might react to that response or how I could change how I phrased it. This was far from a realistic response you might receive from an interviewer.
5. Asked the AI if it would hire me for the role given ALL of these responses – obviously it was not going to give me a real human answer here, but I was interested in the result
a. It gave me completely positive feedback and detailed how I hit the marks and had a solid foundation of what is expected of a Senior Program Manager. It again reminded me to consider other details to include, but did not give me any indication that I truly needed to work on my interview skills. I even asked it to be more direct with me, and it simply rephrased, still told me I was a strong candidate. To be clear, a Senior PM with major yellow and red flags about relationships coming out in an interview with no clear indication of learning or growth is not realistically a strong candidate.
Reasons to do an AI Mock Interview:
- AI might be a good solution for you if you have a lot of interview experience – if all you need is the practice of trying out some new behavioral interview questions and getting a high level, positive response.
- AI can help identify very high-level functional skills based blind spots you may want to focus on in your interview preparation.
- Lower cost and less time investment.
Reasons to do a Mock Interview with a Coach:
- If you want actual interview practice, it’s best to work with a human being who can help you simulate the feeling of a real interview – time bound, with pressure, and even chemistry of talking to the person across the table.
- To improve upon your skills, you need to work with an experienced interviewer who can spot “yellow flags” and follow the threads to better understand your examples. They can also be kind, but honest and thorough about areas you need to work on to ace your real interview. It’s hard to get direct feedback about your interview skills from a company you interviewed with, friends or family, this is the best opportunity to figure out what you’ve been missing.
- It may feel like it takes less time to work with AI than engage with a coach, but your time will be better spent diving deep into coaching with someone who can meet you where you are.
If you’re interested in evaluating and identifying areas you should focus on to improve your job interview skills, reach out to me! It’s sad to think that folks out there may be relying on AI for mock interviews, becoming even more confused about why they’re getting stuck at the interview stage and not landing any roles when it’s telling them they’re good to go. Similarly, I wonder how many companies screening with AI are finding they are still wasting time at the interview stage with candidates who are just not hitting the mark…I’ll continue to dive deep on this topic and report back what I learn. What a fascinating time to be in the world of job interviews!
Happy Interviewing,
Carolyn