Stretch the timer
Content that barely changes should export a longer revalidate window. For evergreen routes we typically use 168h (one week). That is cheaper than letting visitors trigger re-renders.
Vercel Free
You are charged based on the data read from and written to the ISR cache plus the regions that serve those hits. This guide keeps your free-tier usage predictable via deterministic renders, planned revalidation, and focused monitoring.
Give slow-changing routes multi-hour or weekly windows and reserve on-demand revalidation for real changes.
Serve deterministic payloads so cache hits stay identical and avoid cross-region replication when not needed.
Watch the ISR dashboard grouped by project and region to see where cache pressure spikes first.
Content that barely changes should export a longer revalidate window. For evergreen routes we typically use 168h (one week). That is cheaper than letting visitors trigger re-renders.
Ship a webhook or dashboard action that calls POST /api/revalidate?tag=<segment> whenever real data changes. On-demand revalidation is the most predictable way to align ISR writes with business events.
When a revalidation runs but HTML stays identical, Vercel does not count an ISR write. Make sure your output hash is stable so accidental changes do not force cache churn.
You pay per gigabyte read from the ISR cache, multiplied by the regions that serve it. Use the Vercel dashboard to group reads different ways and decide where to tune caching or add a CDN rule.
Breaks down ISR read units per deployment so you can see which property consumes the free tier the fastest.
Shows where the cache is read most frequently. If one geography spikes, consider pinning that route nearer to its audience or using CDN redirects.
We can audit your routes, revalidation triggers, and caching layers so you avoid surprise overages on the free/Pro tiers.