30-Day Study Plan

Updated Nov 16, 2026 · 4 min read
On this page

    With a month to prepare, the risk is not too little time; it is spending it badly. A common pattern: three weeks of reading articles and watching videos, feeling well informed, then a first mock interview in week four that goes poorly because the candidate has never actually designed anything out loud under a clock. Knowledge was never the problem. Practice was.

    A 30-day plan should turn the same thinking process into a habit by repeating it on many problems, adding depth as you go, and saving the last week for timed mocks. That is what this plan does.

    How it works

    Four weeks, each with one focus: week 1 foundation with the framework, building blocks and URL shortener; week 2 problems such as rate limiter, news feed, chat and ticket booking; week 3 depth with to
    Four weeks, each with one focus: week 1 foundation with the framework, building blocks and URL shortener; week 2 problems such as rate limiter, news feed, chat and ticket booking; week 3 depth with toolkit topics, patterns and redoing two problems; week 4 mocks, scoring and final review.

    Week 1: Foundation

    Week 2: Problems

    One problem a day, same loop each time: 20–25 minutes on paper, read the walkthrough, fix the gaps, explain out loud.

    1. Design a Rate Limiter
    2. Design FB News Feed
    3. Design a Chat System
    4. Design a Ticket Booking System
    5. Design Dropbox / Google Drive

    By the end of the week, starting a new problem should feel routine, even if the deep dives are still rough.

    Week 3: Depth

    • Days 1–4: the Deep Dive Toolkit, one or two topics a day: locking, sharding, consistent hashing, rate limiting algorithms, idempotency, backpressure and circuit breakers, TTL. For each topic, note when it comes up, what it solves, what it costs, and which problem used it.
    • Day 5: Patterns: read-heavy, write-heavy, real-time messaging, fanout, large files, distributed sync. Notice which problems you've done fit each pattern.
    • Days 6–7: redo two Week 2 problems from memory, adding one deeper discussion each (for example, celebrity fanout in the feed, or hold expiry in ticket booking).

    Week 4: Mocks

    Three or four full mocks, 50 minutes each, on problems you haven't studied: for example Design Uber, Design YouTube, Design Job Scheduler, Design a Proximity Service like Yelp. Speak out loud, draw as you go, and score each mock right after.

    Mock scorecard: five areas scored from 1 to 5; in this example, reaching the first deep dive by minute 22 scores lowest at 2, so the next mock uses a 22-minute alarm.
    Mock scorecard: five areas scored from 1 to 5; in this example, reaching the first deep dive by minute 22 scores lowest at 2, so the next mock uses a 22-minute alarm.

    Fix the lowest score before the next mock. Finish with the System Design Interview Cheat Sheet, Back-of-the-envelope Numbers and Common Interview Phrases in the last 48 hours. No new topics then.

    Trade-offs and costs

    • Mocks feel worse than reading. That discomfort is the point; it shows the gaps while there's still time to fix them.
    • Not every topic gets depth. Multi-region, stream processing and consensus get light coverage. Add them only if your target roles emphasize them.
    • Consistency beats intensity: 75 minutes on 25 days builds more than three 8-hour weekends.

    When to reach for it

    • Interviews 3–6 weeks away, 60–90 minutes a day.
    • Less time: the 7-Day Crash Plan, or compress this plan by doing Week 2 problems on alternate days.
    • More time, senior target: add a fifth week of hard problems (distributed key-value store, message queue, ad click aggregator) and two more mocks.
    • Mid-level target: spend more of Week 3 on building blocks and fewer on advanced toolkit topics.

    Example

    One Week 3 day, idempotency:

    • 20 min: read Idempotency & Deduplication.
    • 10 min: notes: comes up in payments, order creation, retries, queue consumers; solves duplicate side effects; costs key storage and an extra write; used in ticket booking.
    • 30 min: reopen the ticket booking design and add a payment deep dive: idempotency keys from the client, the key stored in the same transaction as the booking, a pending state while the provider responds.
    • 10 min: explain that deep dive out loud in 3 minutes.

    Common mistakes

    • Reading for three weeks and mocking for three days.
    • Only easy problems, because they feel good.
    • Not scoring mocks, so the same weakness repeats.
    • Learning new topics in the final 48 hours.
    • Skipping rest days. Two days off a week keeps the pace sustainable.

    Say this in the interview

    By the end of 30 days, this opening and closing should come naturally:

    "I'll start by clarifying the core use cases and scale, then propose a simple design and deep dive into the main bottlenecks." … "To wrap up, here's one request end to end; the biggest remaining risk is X, and with more time I'd add Y."

    Used in