7 out of 11 candidates gave me the exact same answer


Hey Reader,

Welcome back to Guided Growth where we talk about the skills that accelerate your career beyond just writing great code.


This is The Unwritten Rules #2: a 4-part series where I share the career rules that exist but nobody shared them with you.


Last quarter, I interviewed 11 candidates for a senior backend role. I asked all of them: "Design a URL shortener."

7 out of 11 gave me the same answer. Same structure. Same buzzwords. Same diagram that looked like it came from the same YouTube video.

"Read-heavy system. NoSQL database. Base62 encoding. Redis cache. Load balancer. Scale horizontally."

It's a correct answer. On paper.

But when I can predict what you'll say next, it tells me you borrowed this knowledge. You didn't build it.

One candidate, let's call him Rohan, started differently:

"Before I design anything, how many URLs per day are we expecting? And is this a public tool like Bitly or an internal tool for a marketing team?"

Two questions. That's all it took. Because the design for 1,000 URLs a day is completely different from 100 million. Most candidates skip this step and jump straight to drawing boxes.

It's like a kid who memorizes "7 × 8 = 56" but can't figure out how many chocolates to buy for 7 friends who each want 8.

Three things tell me you learned from videos, not from building:

You over-engineer from the start. Microservices for 500 users. Kafka for a system that doesn't need event streaming.

You can't answer "why." Why Redis and not Memcached? Why NoSQL and not Postgres? If you pause, I know you picked it from a template.

You never mention failure modes. Real systems break. Builders know this because they've seen it happen.

My advice: build the thing after you watch the video. That one weekend project teaches more than 10 tutorials. When I ask "why Redis?" you'll say "because I tried without it, and the latency was terrible."

That answer can't be faked.


Know someone preparing for system design interviews? Forward this to them.

That's The Unwritten Rules #2. Next Thursday: the non-coding question I ask every candidate and the answer that decides everything.

Thanks,

Abhishek

The Guided Growth

The Guided Growth

Ex-Google | Stanford LEAD | Ex-Founder | Sr. Engineering Manager. Career systems for engineers who want FAANG offers, faster promotions, and leadership roles. 650+ interviews conducted, 1,100+ engineers mentored. Read by 178,000+ engineers across platforms.

Read more from The Guided Growth

Hey Reader, Welcome back to Guided Growth where we talk about the skills that accelerate your career beyond just writing great code. This is The Unwritten Rules #3: a 4-part series where I share the career rules that exist but nobody shared them with you. What do you think matters MOST in the final interview round? Coding skills System design depth Cultural fit Communication skills Last month, I gave two candidates the same debugging problem. A piece of code with a subtle concurrency bug....

Why Most Engineers Never Hear Back From Recruiters

Hey Reader, Welcome back to Guided Growth where we talk about the skills that accelerate your career beyond just writing great code. This is The Unwritten Rules #1: a 4-part series where I share the career rules that exist but nobody shared them with you. A reader replied to my newsletter last month. "200+ applications. 4 months. 6 callbacks. 0 offers. What am I doing wrong?" What do you think matters MOST in the final interview round? Coding skills System design depth Cultural fit...

The 4 Projects That Actually Get You Promoted (and the 20 That Don't)

Hey Reader, Welcome back to Guided Growth where we talk about the skills that accelerate your career beyond just writing great code. You spent 3 months on the analytics dashboard. Nights. Weekends. Shipped perfectly. Promotion cycle: Not mentioned. Your peer wrote a design doc in 2 weeks. Got promoted to Staff. I've reviewed 47 promotion packets. Pattern is clear: Only 4 types of projects get you promoted. THE 4 PROJECTS THAT MATTER 1. Multiplier Projects Make your team 2x more effective....