What are optimistic updates and how do you implement them safely?

Question

Accepted Answer

An **optimistic update** updates the UI **immediately**, *assuming* a server request will succeed, instead of waiting for the response. This makes the app feel instant — but you must handle the case where the request actually fails by **rolling back**.

## The contrast

```text
Pessimistic (normal): click → show spinner → wait for server → THEN update UI (feels slow)
Optimistic:           click → update UI INSTANTLY → send request →
                      if it fails, ROLL BACK the change (feels instant)
```

## Why it matters for UX

For high-confidence actions (liking a post, toggling a todo, sending a chat message), the request almost always succeeds, so waiting for the round-trip just adds perceived lag. Optimistic updates make these feel immediate.

## Implementing it safely (the rollback is critical)

```jsx
async function toggleLike(post) {
  const previous = post.liked;                 // 1. SNAPSHOT the old state for rollback
  setPost({ ...post, liked: !post.liked });    // 2. update UI optimistically

try {
    await api.toggleLike(post.id);             // 3. fire the request
  } catch (err) {
    setPost({ ...post, liked: previous });     // 4. ROLL BACK on failure
    showError("Couldn't update — please retry"); // 5. tell the user
  }
}
```

The key to safety: **snapshot the previous state before the optimistic change**, and restore it if the request fails. Skipping rollback leaves the UI lying about server state — the dangerous bug optimistic updates can introduce.

## With React Query (built-in support)

```jsx
useMutation({
  mutationFn: toggleLike,
  onMutate: async (newData) => {
    await queryClient.cancelQueries({ queryKey: ["post"] });
    const previous = queryClient.getQueryData(["post"]); // snapshot
    queryClient.setQueryData(["post"], newData);          // optimistic update
    return { previous };                                   // pass snapshot to onError
  },
  onError: (err, vars, context) => {
    queryClient.setQueryData(["post"], context.previous);  // rollback
  },
  onSettled: () => queryClient.invalidateQueries({ queryKey: ["post"] }), // reconcile with server
});
```

React Query formalizes the pattern: `onMutate` (snapshot + optimistic update), `onError` (rollback), `onSettled` (refetch to confirm truth).

## When to use / avoid

```text
✓ Use: high-success, low-stakes actions (likes, toggles, reorder, chat)
✗ Avoid: critical/uncertain operations (payments, irreversible actions) — wait for confirmation
```

## Why it matters

Optimistic updates are a powerful UX technique that makes apps feel instant by not waiting on the network — but they're only *safe* if you implement rollback (snapshot-then-restore) and surface errors.

The pattern (snapshot → update → on failure roll back → reconcile) is essential knowledge, and understanding when it's appropriate (likely-to-succeed, low-stakes actions) versus dangerous (payments) is what separates a polished, trustworthy UI from one that silently misrepresents server state.