Cache, Assets, and Performance Diagnostics

Step 1: Measure a Real Route
Choose a representative public route and record cold and warm response time, rendered content, transfer size, requests, long tasks, layout shifts, largest content, interaction behavior, and server logs. Test anonymously and use the same route and viewport for the before and after comparison.

Step 2: Reduce Unused Work
Optimize the media actually rendered, use responsive sources, preserve useful lazy loading, and enable only the Builder modules the page needs. Keep component styles and scripts tied to their widgets or module contracts rather than loading every commerce, learning, slider, form, or administration asset on every page.

Step 3: Protect Cache Boundaries
Cache public read-only pages where appropriate, but preserve bypass behavior for administration, previews, signed-in sessions, carts, checkout, account pages, forms, query-sensitive routes, webhooks, and personalized content. A fast cached response is wrong when it contains another visitor state or stale post-save output.

Step 4: Warm and Recheck Behavior
Purge only the affected route or site scope, request it cold, confirm the new content fingerprint, then request it warm and compare. Test menus, FAQ, sliders, video, forms, search, cart, responsive layout, and reduced motion. Keep the change only when public behavior and fresh logs remain correct as well as the performance evidence.
