Results day is not the day to find out you are slow.

A university site is not one site. It is a decade of departments, each with its own pages, news posts, staff profiles and PDF handbooks, all on one WordPress install that nobody has fully audited since it was built.

For most of the year that is fine, because most of it is barely visited. Then admissions closes, or results are published, and thousands of people arrive within an hour looking for the same handful of pages — and the install that was comfortable at ten visitors a minute meets a thousand.

Almost all of that content is static: it has not changed in months and will not change today. Served as cached HTML and prewarmed before the deadline, the spike becomes a non-event, and the only pages doing real work are the ones that genuinely need to.

Using xSpeed, prepare our education site for a deadline-day traffic spike. Benchmark it now, enable page caching, run the preloader across the whole site, and schedule prewarming for the admissions and results sections. Exclude any logged-in student area. Then re-benchmark and tell me how much headroom we have and which pages are still slow.

What happens: Your AI preloads the site, schedules the warming, and reports which pages remain uncached and why — before the spike, not during it.