Saas

Priority patterns

1. Per-request memoization with React.cache(). Server Components called from multiple places in the same request tree often re-query the database. React.cache() dedupes within the request. Cite vercel-react-best-practices:server-cache-react.
2. Parallel data loads in Server Components. Dashboards typically load user + org + billing + recent-activity. Run all four in parallel via Promise.all. Cite vercel-react-best-practices:async-parallel and :server-parallel-fetching.
3. Fluid Compute. Auth-gated routes have higher cold-start sensitivity (every cold start is a user waiting). If cold-start signal shows up in observability, Fluid Compute is usually the right account-level rec.
4. Async work after response. Activity logs, audit trails, analytics events — anything that doesn’t block the user — should run via after() (Next 15+) or waitUntil() from @vercel/functions. Cite vercel-react-best-practices:server-after-nonblocking.
5. Suspense boundaries around expensive widgets. The dashboard shell renders fast; widgets stream in. This shifts perceived latency without changing the underlying queries.

Frequent gotchas

• N+1 ORM queries. A list page that loops over results and fetches related records per-item. Especially common with Prisma’s .findUnique inside a .map. Use include or batch via DataLoader.
• Sequential session+permission checks. await getSession() then await checkPermissions() then await loadData() — these can often be parallelized when the permissions check doesn’t depend on the data load.
• No connection pooling on serverless. Prisma without a pooler exhausts the database under load. Connection pooling is mandatory.
• Polling for state from the client. Every poll is a function invocation. Replace with SWR + on-demand revalidation, or with revalidateTag triggered by the mutation that actually changes state.

Cross-references

• vercel-react-best-practices:server-cache-react — per-request dedup
• vercel-react-best-practices:server-parallel-fetching — restructure for Promise.all
• vercel-react-best-practices:async-suspense-boundaries — stream the dashboard shell
• vercel-react-best-practices:server-after-nonblocking — defer audit/analytics writes (Next 15+)
• vercel-react-best-practices:client-swr-dedup — replace polling with SWR
• https://vercel.com/docs/fluid-compute — when cold starts hurt