Coding Interview Blank Pad: What to Do in the First Two Minutes
The blank pad is where candidates lose before they start. You feel the clock. You want to type. Typing without a plan is how you paint yourself into a corner.
Here is a first-two-minutes script that works whether the pad is a shared doc, CoderPad, or a plain textarea.
Minute 0–0:30 — Mirror the ask
Repeat the problem in your words. Confirm I/O types, whether order matters, and what “invalid” means. Write the function signature immediately so the page is not empty.
Minute 0:30–1:00 — Constraints out loud
Ask or infer: n size, value ranges, duplicates allowed, memory limits, online vs offline. Constraints choose the algorithm. Saying them shows you are not guessing.
Minute 1:00–1:40 — Two-line plan
Not a novel. Two lines max:
- Approach in plain English (“hash count, then heap of size k”)
- Why it fits the constraints (“n up to 1e5 → avoid O(n²)”)
If you cannot fit it in two lines, you do not have an approach yet—keep talking, do not code.
Minute 1:40–2:00 — One tiny example
Walk one example on the side of the pad. Boxes, arrows, whatever. This is your oracle for the dry run later.
Then start coding.
Why this beats “just start”
| Habit | Outcome |
|---|---|
| Dive into code | Mid-solution rewrite |
| Signature + plan + example | Partial credit even if code is incomplete |
| Silent thinking for 3 minutes | Interviewer assumes stuck |
Practice the opening cold
Drill openings separately from full solutions. Ten problems where you only do the first two minutes. Stop. Next problem. Opening fluency matters more than another memorized pattern.
A Chrome side panel that keeps the live question visible while you write the plan—like Luna Interview—can help rehearsals feel closer to the real screen. Keep integrity rules intact: practice to narrate your own approach.
Closing thought
Empty editor is a psychological trap. Put structure on screen early. The first two minutes are for orientation, not for heroics.
Rehearse with an AI interview copilot, then browse commonly reported interview questions.