Compare equivalent pages under the same test conditions. A static landing page and a logged-in shop are different workloads, even when they share a framework name.
Practical checks
-
Record the URL, build, device profile, network settings, tool version and timestamp.
-
Repeat lab tests and examine image, font, CSS and JavaScript costs rather than selecting the best run.
-
Use real-user LCP, INP and CLS where enough data exists; lab TBT is not field INP.
What to record before a decision
Write a short decision record: the current problem, one proposed change, the evidence needed to accept it, the person responsible and a review date. Test with fictional data before exposing customer information. If evidence is missing, record it as unknown rather than presenting a forecast as an achieved result. Keep a working fallback so a customer can still reach staff when an integration fails.
Scope and limits
Both platforms require appropriate hosting, asset control, accessible design and maintenance. WordPress optimisation and Next.js static export each have documented options and limits. Framework choice does not guarantee a speed score, search ranking or conversion increase.
