I used to think mock interviews were the golden ticket. Every career coach, every Reddit thread, every LinkedIn influencer said the same thing: "Practice with a friend! Do mock interviews!" So I did. For months. And I kept bombing real interviews anyway.
Let me explain what I mean.
The Mock Interview Trap
Back in 2023, I was grinding through the job search after getting laid off from a mid-stage fintech startup in Austin. I had a solid resume — 4 years of backend engineering, decent system design chops, a couple of open-source contributions. On paper, I should've been fine.
My buddy Marcus and I set up a routine. Every Tuesday and Thursday evening, we'd hop on a Zoom call and take turns grilling each other. He'd ask me things like "Design a URL shortener" or "Tell me about a time you resolved a conflict with a teammate." I'd answer, he'd give feedback, we'd switch.
It felt productive. It felt like we were doing The Right Thing.
Then I walked into my onsite at Stripe and completely froze during the system design round. The interviewer asked me to design a real-time payment notification system, and my mind went blank. Not because I didn't know the material — I'd actually studied this exact topic. But the pressure was completely different from anything Marcus and I had simulated.
What's Actually Wrong With Mock Interviews
Here's the uncomfortable truth I've come to accept after going through this cycle with multiple companies (Stripe, Datadog, Coinbase, and a handful of Series B startups):
1. Your friend is too nice.
Marcus never interrupted me mid-sentence. He never asked a sharp follow-up that made me reconsider my entire approach. He never sat in uncomfortable silence while I fumbled through an answer. Real interviewers do all of these things. Not because they're mean, but because they're evaluating you under realistic conditions.
2. You're practicing the wrong things.
When Marcus asked me behavioral questions, I'd give these perfectly polished, rehearsed stories. They sounded great in practice. But in my actual Amazon LP interview, the interviewer kept drilling deeper: "What specifically did YOU do? What was the metric impact? What would you do differently?" My rehearsed stories fell apart because I'd practiced delivering monologues, not having conversations.
3. The timing is artificial.
In mock interviews, there's no real clock pressure. If I took 45 minutes on a coding problem that should take 25, Marcus would just let it slide. In my Meta phone screen, the interviewer literally said "We have about 5 minutes left" when I was only halfway through my solution. That moment of panic? You can't simulate it with a friend who's rooting for you.
4. There's no real-time course correction.
This is the big one. In a real interview, you need to read the room as it's happening. Is the interviewer nodding along or looking confused? Did you miss a key constraint in the problem? Are you going down a rabbit hole? Your mock partner can tell you after the fact, but by then, the moment is gone.
The Day I Changed My Approach
After my Stripe rejection (which stung — I really wanted that job), I took a step back and thought about what was actually going wrong. I realized my problem wasn't knowledge. It was real-time performance.
I started looking into tools that could help with the actual moment of the interview, not just the preparation phase. A colleague from my previous company mentioned she'd been using AceRound AI — an AI tool that listens to your interview in real time and provides suggestions as you're talking. I was skeptical at first. It sounded almost too good to be true.
But I tried it during a practice session, and the difference was immediately obvious. Instead of getting feedback 30 minutes after I'd already messed up, I was getting nudges in the moment: "Consider mentioning trade-offs between consistency and availability" or "You haven't addressed the scalability constraint yet." It was like having the world's most attentive interview coach sitting right next to me.
What Actually Works
I'm not saying mock interviews are useless. They have their place. But here's what I'd recommend based on my experience:
Use mock interviews for warmup, not mastery. They're good for getting comfortable talking about your experience out loud. That's about it.
Record yourself. Seriously. Watch your actual interview recordings (if the company allows it) or at least your practice sessions. You'll notice verbal tics, long pauses, and tangents you never knew you had.
Practice under real constraints. Set a hard timer. Use a tool that can simulate realistic pressure. Don't let your friend give you extra time "just this once."
Get real-time feedback. This was the game-changer for me. Whether it's through a tool like AceRound AI or by reviewing your thought process in the moment, the ability to course-correct while you're still in the interview is infinitely more valuable than a post-mortem.
Study the company, not just the topic. I started reading engineering blogs from the companies I was interviewing at. When I interviewed at Notion, I'd read their blog post about their data model. When I talked to Figma, I'd studied their multiplayer architecture. This gave me specific context that no amount of generic mock interviews could provide.
The Result
After changing my approach, I went from a roughly 15% onsite-to-offer conversion rate to landing offers from two companies in the same month. The difference wasn't that I suddenly became a better engineer overnight. I just got better at performing during the actual interview.
Your mock interview partner means well. But meaning well and actually helping you get a job are two very different things.
If you're stuck in the mock interview loop and not seeing results, it might be time to try a different approach. I'd recommend checking out AceRound AI — it's the closest thing I've found to having a real-time interview coach. It won't replace your preparation, but it'll help you actually perform when it counts.
Top comments (0)