All notes
NetlifyPerformanceOperations· 6 min

What breaks when a Netlify site goes viral

A field report from managing 19 production Netlify sites across clients and our own SaaS products.

JS

Joe Saba

January 14, 2026

When a client's site gets picked up by a newsletter with 100k subscribers, you learn which parts of your stack actually hold and which parts were only ever load-tested against two engineers hitting refresh.

Netlify functions: the underappreciated chokepoint

Static assets scale linearly with Netlify Edge. Serverless functions do not. Every client site that survived a traffic spike had pre-rendered the critical path and used functions only for form submission, auth callbacks, and edge personalization. Sites that called a function on every page load hit cold-start cascades.

The 3 things we now bake into every launch

1) A static fallback for every dynamic read. 2) Netlify Image CDN on every hero and project image — the savings are free. 3) A plain HTML /status page served outside the SPA, so when the SPA breaks, visitors still get a signal that the business is alive.

What monitoring actually catches

We run Better Stack on every production site. It caught 7 outages last year that nobody on our Slack would have seen for hours. None of them were our fault. All of them were upstream vendor issues that we communicated about within 10 minutes — which is the kind of thing clients remember.