Choose representative journeys
Include the homepage, an important service page, a resource and the enquiry route. Test mobile conditions as well as desktop. Note third-party scripts, large media and interactions that may affect the customer’s experience.
Mobile page experience
A mobile test snapshot with the original performance and page-quality scores.
Separate lab and field evidence
Laboratory tests help reproduce problems; real-user measurements show what visitors have experienced. Use both where available and state when field data is missing. Avoid describing an unmeasured redesign as a guaranteed performance improvement.
Fix the largest avoidable costs
Size media for its placement, reserve space for content and reduce unnecessary JavaScript. Review whether each external widget earns its loading cost. Recheck the same journey after changes rather than comparing unrelated test conditions.
Put the guidance to work
Capture your starting point, assign an owner to the next action, and decide what evidence will confirm that it worked.
Discuss this with us