The Amazon VP Who Interviewed Hundreds Reveals His #1 Interview Rule

Neil Roseman spent years as Technology VP at Amazon and Zynga, interviewing hundreds of engineers. His biggest frustration? Interviewers who ask questions they can't properly evaluate.

"One of the things that really pisses me off is people asking questions that they don't even have a good handle on themselves," Roseman says. "They don't have a good distinction between a great answer and a crappy answer."

Sound familiar? You're not alone if you've ever walked out of an interview unsure whether the candidate actually knew their stuff or just gave confident-sounding answers.

The 60-Second Rule That Changes Everything

Here's Roseman's counterintuitive approach: Once you form an initial impression of someone—which usually happens within the first 60 seconds—spend the rest of the interview trying to invalidate that impression.

This flips the script on confirmation bias. Instead of looking for evidence that supports your gut feeling, you actively hunt for contradictory data. It's debugging your own decision-making process.

From Resume Claims to Technical Reality

The magic happens in how you read resumes. Roseman doesn't just scan for buzzwords—he hunts for specific, measurable claims.

"Grew revenue by 50%" or "decreased downtime by 30%" become interview gold mines. During the session, he'll have candidates whiteboard the actual system they claim to have optimized, then drill down: "Show me your specific contribution. Walk me through the architecture. What was the before-and-after?"

The key distinction: We vs. I. Participant vs. owner. Exposure vs. expertise.

Most candidates with impressive-sounding accomplishments will crumble under this scrutiny—they were participants, not architects. The exceptional ones will light up and dive deep into technical details because they actually built what they're describing.

The Technical Bar Comes First

Roseman leads with hands-on technical questions because "the reason most people don't get hired is because they simply don't have the skills—they don't pass the technical bar."

His non-negotiable rule: Never ask a question you haven't tested on non-candidates first. You need to know what constitutes a poor, good, and exceptional answer before you ask it.

He'll often use layered questions—start with a basic coding problem, then add constraints like limited memory to see how deep their understanding goes. Even if they've seen the surface-level question online, the follow-ups reveal true competency.

The Question That Reveals Character

After technical validation, Roseman asks everyone the same culture question: "Do you consider yourself lucky?"

He's looking for people who embody "fortune favors the prepared"—engineers who create their own opportunities rather than blaming external factors for setbacks. In startups especially, this mindset separates builders from blamers.

Your next technical hire will either elevate your team's standard or dilute it. Roseman's approach ensures you're hiring people who actually did what their resume claims—and have the mindset to do it again.

Question from Messa: How many of your recent hires could recreate their biggest "accomplishment" from scratch if asked.

This post draws on insights from Neil Roseman’s 2013 HBR article, “Fortune Favors the Prepared, and Hiring Managers Favor the Fortunate.”