How would you decide what data to cache on the frontend and when to refetch it?

Frontend caching, stale data, and refetching

Explains client cache keys, freshness, invalidation, refetching, optimistic updates, and server truth.

1Say itAnswer out loud first

2CompareCompare with Core answer

3RefineRefine with Stronger and Advanced

TL;DR

I would cache data that is reused across components or routes, expensive to fetch, or safe to keep briefly stale. I build cache keys from the inputs that change the result, such as user, tenant, filters, sort, and page cursor. After mutations, I update or invalidate the affected cache entries. The server remains the source of truth, so the UI reconciles with the response or refetches when correctness matters.

Full answer below

Practice focus

Cache keys, stale time, invalidation, mutation effects, server truth, optimistic updates, and user expectations.

Answer first

Answer path

Core answer first, then checkpoints toward a stronger answer.

Start
Core answer
Start here. This is the direct spoken answer to practice first.
Why this question matters
Frontend caching makes apps feel fast, but stale data bugs can make users distrust the product. The interview signal is knowing when cache is helpful and when it must yield to server truth.
I would cache data that is reused across components or routes, expensive to fetch, or safe to keep briefly stale. I build cache keys from the inputs that change the result, such as user, tenant, filters, sort, and page cursor. After mutations, I update or invalidate the affected cache entries. The server remains the source of truth, so the UI reconciles with the response or refetches when correctness matters.
Checkpoint
Stronger answer
Optional upgrade with examples, trade-offs, and sharper phrasing.
I would choose freshness based on product expectations. Reference data can stay fresh longer than an admin approval list or inventory count. Some pages can show stale data while refetching; others need a blocking refresh. Optimistic updates need rollback and server reconciliation. Broad invalidation is simpler but can cause extra traffic; precise updates are efficient but easier to miss. Query libraries help standardize these choices.

Review signals

Use these to catch weak phrasing, shallow reasoning, and risky claims.

Common mistakesCandidateCommon mistakesMistakes that weaken the answer or signal shallow understanding.Open details

Details

Using cache keys that omit filters, tenant, or user scope.
Treating cached data as final truth after a mutation.
Invalidating everything after every small action.
Never refetching data that can change outside the current tab.

How would you decide what data to cache on the frontend and when to refetch it?

Practice focus

Answer path

Why this question matters

Review signals

Follow-up prompts

Continue practice

Go deeper

Tech stacks

Skill areas

Suggest an improvement