What is a good strategy for approaching coding-interview problems?

Question

Accepted Answer

A reliable strategy is to **understand, plan, then code** — never jump straight to typing. A structured approach surfaces edge cases early and communicates your thinking clearly.

## A step-by-step framework

```text
1. CLARIFY  -> restate the problem, ask about inputs, ranges, edge cases
2. EXAMPLES -> work a small example by hand; note edge cases (empty, 1, dups)
3. BRUTE FORCE -> state the obvious solution and its complexity
4. OPTIMIZE -> spot the bottleneck; pick a pattern (hash, two-pointer, DP...)
5. CODE     -> write clean, modular code while talking through it
6. TEST     -> trace examples, check edges, fix bugs
7. ANALYZE  -> state final time and space complexity
```

## Pattern recognition

| Signal | Likely technique |
|--------|-----------------|
| sorted array / pair | two pointers / binary search |
| "subarray / substring" | sliding window / prefix sum |
| "count / seen before" | hash map |
| "all combinations" | backtracking |
| overlapping subproblems | dynamic programming |
| graph / grid reachability | BFS / DFS |

## Example mindset

```python
# Always START from the brute force, THEN optimize.
# Brute force two-sum is O(n^2); a hash map makes it O(n).
# State this trade-off out loud before coding.
```

## Pitfalls

Don't code in silence, don't ignore edge cases, and don't optimize prematurely. Get something correct first, then improve it.

## Why it matters

A repeatable process keeps you calm and prevents the panic-coding that produces buggy solutions.

Verbalizing each step lets the interviewer follow your reasoning and give hints — they evaluate *how* you think, not just the final answer.

The same understand-plan-test discipline makes you a better engineer on real, ambiguous problems.

Signal	Likely technique
sorted array / pair	two pointers / binary search
"subarray / substring"	sliding window / prefix sum
"count / seen before"	hash map
"all combinations"	backtracking
overlapping subproblems	dynamic programming
graph / grid reachability	BFS / DFS