AI Can Find the Questions. You Still Have to Find the Right Answers.

I coach a lot of clients for interviews, but two sessions I had on the same day last week stayed with me.

My second client of the day was one of the strongest interviewers I have coached in quite a while. Without any prompting, he answered questions in a STAR format. He explained the situation, his role in it, what he did, and then closed most of his answers with some version of, “And the result of this was…”

That may sound basic, but I’ve found it is not how most people interview. When I ask candidates behavioral questions, many answer in broad terms. They tell me what they would do, how they tend to approach something, or what their philosophy is on the topic. When they do give an example, they often leave out part of the story. I may hear what happened and what they did, but not the result. Or I hear the result without enough detail about their own role in getting there. This client did a very good job of giving me the full story.

A Strong Interview Does Not Guarantee the Job

He has worked at Google and Meta and is now interviewing with another FAANG company. The first time we worked together, his search moved quickly. He landed a new role within weeks. That role later ended in a layoff, along with many others in the tech sector, and this search has been different.

He has been getting interviews, although not at the same volume as before, and he has moved through several rounds with different companies. More recently, he was rejected late in the process for another role. That creates a very normal question in a candidate's mind: Am I doing something wrong in these interviews?

From what I heard in our session, I do not think poor interviewing is the problem. The role he is pursuing now is also a step up. He has not held the title before, and that creates one real gap in his experience. He has fewer examples of owning something from the first idea through the final result. In many of his roles, he has been the person who takes a concept or strategy and drives it through execution. While those may not be the same thing, they are close enough that we had strong examples to work with. There was no reason to dress his experience up as something it was not. If a company wants someone who has already led work from concept through completion many times, that will show up in the interview process anyway. So we focused on the closest examples he did have. At this stage, the company already knows his background; they have seen his résumé, know the titles he's held, and still invited him to interview. If this role does not work out, I would not assume it is because he gave a bad interview. We do not know who else is in the process, whether there is an internal candidate, who has deeper domain experience, or what tradeoffs the hiring team is making.

What I can say after coaching him is that he explains his work well. Sometimes another candidate is simply a closer match.

Both Clients Had Already Used AI

The interesting part of that day was that both clients had already used AI to prepare before meeting with me, and both received fairly generic interview questions.

For one client, the problem was more specific. She was preparing for an initial recruiter phone screen, but many of the questions AI gave her were questions I would expect much later from a hiring manager. A recruiter screen and a hiring manager interview usually serve different purposes.

The recruiter may be confirming basic fit, motivation, salary expectations, location, timing, and whether your background lines up with what the team asked them to find. A hiring manager is more likely to dig into the substance of your experience, how you think, what you have accomplished, and how you would operate on the team. If you prepare for the wrong interview, you can spend a lot of time practicing questions you are unlikely to hear.

For both clients, I researched the roles beyond the job descriptions. I use AI tools I have built with Claude Code to help with this. They let me find details that are hard to uncover through a basic search: interview stages, team structure, candidate reports, company hiring practices, evaluation criteria, and sometimes even the types of questions candidates have been asked. The research gave us a much better idea of what each interview was likely to look like.

By the end of the day, my first client emailed me to say she was moving forward to the hiring manager interview. She also told me that the recruiter screen went almost exactly as we had discussed. The questions we prepared for came up, and in nearly the same order.

AI was useful in that process, but the more important work happened after we knew what she was likely to be asked.

Knowing the Question Is Not the Same as Knowing the Answer

One of the first questions I asked her during our session was the most predictable interview question of all: Tell me about yourself.

She had the information. Her answer covered the right parts of her background. But compared with the rest of the session, it did not sound nearly as strong. Later, I asked her about some projects she worked on in her previous role, and the difference in how she spoke was obvious. She loved that job. When she talked about the company and the work she had done there, her voice changed. She had more energy. She gave me better detail. It was easy to hear that she cared about the work and understood why it mattered. That was missing from the answer she had prepared for “Tell me about yourself.” So I asked her to stop using the version she had been practicing.

Instead, I asked what she loved most about her previous job and told her to describe her role from that point of view. Her next answer did not sound like she was trying to remember an interview script. She was talking about work she understood well and enjoyed doing. We used that conversation to rebuild her introduction; the facts didn't change much, the way she communicated them did.

AI Can Find the Question. It Cannot Find Your Best Story for You.

This is where I think people sometimes expect too much from AI interview preparation. AI can do an impressive amount of research. With the right prompts and tools, it can help uncover a company's interview process, common question types, leadership principles, competency models, candidate experiences, or even scorecards and evaluation criteria discussed publicly. In some cases, it can get surprisingly close to the questions you will hear.

But knowing the question does not mean you know your best answer. AI wasn't in the room when you led the project. It does not know which accomplishment you are most proud of unless you tell it. It does not hear the difference in your voice when you talk about one project instead of another. It may not recognize that the example you think is your strongest is too far away from what the interviewer is trying to measure. It can help organize your experience, but it cannot fully judge how compelling you are when you explain it, and that is where another person can still add value.

You Do Not Need an Interview Coach to Prepare Well

Interview coaching can help, but I do not believe everyone needs to hire a coach, and not everyone can afford one. There is a lot you can do on your own, especially if you use AI as a research and practice tool instead of treating it as an answer generator.

Here are a few things I would do before an interview:

Tell AI who you are interviewing with. Do not only paste in the job description. Tell it whether you are meeting with a recruiter, hiring manager, department leader, peer, executive, or panel. Ask what that person is most likely trying to assess at that stage.

Ask it to research the interview process. Look for recent candidate reports, company career pages, interview guides, leadership principles, Glassdoor discussions, Blind posts, Reddit threads, or other public sources that may give you clues about how interviews are structured.

Ask what the role is really measuring. Have AI identify the five or six main competencies behind the job description. Then ask what evidence an interviewer would want to hear for each one.

Build a story bank instead of memorizing answers. Prepare six to ten examples from your background that you can reuse across different questions. Know the situation, your role, what you did, and what changed because of it.

Ask AI to challenge your examples. After you give it an answer, ask what is missing. Is your role clear? Is the result clear? Are you talking too much about the team and not enough about your own contribution? Did you answer the question?

Practice out loud. Reading an answer and saying it are completely different experiences. You will hear problems out loud that you will never catch on the page.

Do not memorize AI-written responses. If the language does not sound like something you would naturally say, rewrite it. The goal is not to sound polished enough to pass an interview. The goal is to explain your experience clearly enough that another person understands your value.

The Goal Is Not the Perfect Answer

The best interview answers make it easy for the interviewer to understand what you know, what you did, why you did it, and what happened because of your work, and not always the most polished ones.

AI can help you get there faster: it can help with research, surface questions, identify themes in a job description, organize your stories, and point out gaps in your answers.

A coach can use those same tools and add what AI still struggles with: listening to how you communicate, recognizing when an answer isn’t working, asking the follow-up question that uncovers a better example, and helping you decide which parts of your experience are worth leaning into.

You don't need a coach to prepare well, but if you have been getting interviews and not moving forward, or you are preparing for an opportunity that feels especially important, having another person listen to your answers can reveal things that are hard to catch on your own. And sometimes the problem is not that you are interviewing poorly at all. Sometimes someone else simply has experience that is closer to what the company needs.

If you want help figuring out which one it is, I am always happy to talk.