Page speed – a business metric, a ranking factor and a technical challenge

Definition

Page speed is how quickly a web page's content loads and is displayed to the visitor. It is one of the most directly measurable links between technical quality and business results in web development.

Page speed is a business metric, not just a technical measurement

Every second of delay affects the visitor's decision to stay or leave, and the knock-on effect shows up in conversion rate, bounce rate and revenue. According to Deloitte and Google, making a mobile page a tenth of a second faster raised the conversion rate in retail by 8.4 per cent. The probability of a visitor leaving the page increases by 32 per cent when loading time rises from 1 to 3 seconds – and by 90 per cent at 5 seconds, according to Google Developers.

Pagespeed vs Core Web Vitals – what is the difference?

Pagespeed is the broad term for how fast a page feels to the visitor. Core Web Vitals are Google's specific metrics for quantifying three aspects of that experience:

• LCP (Largest Contentful Paint) – how quickly the page's primary content appears (target: under 2.5 seconds)
• INP (Interaction to Next Paint) – how quickly the page responds to clicks and interactions (target: under 200 ms)
• CLS (Cumulative Layout Shift) – how stable the layout is while loading (target: under 0.1)

Page speed is primarily about LCP and the total loading time. INP and CLS measure other dimensions of the experience. Google uses Core Web Vitals as a ranking factor through its Page Experience framework.

What affects a page's loading time?

• Unoptimised images – Uncompressed PNG and JPEG files are the single most common cause. WebP cuts file size by around 30–40% compared with JPEG, with no noticeable loss of quality. AVIF goes further: in most use cases the files are more than 50% smaller than JPEG at the same image quality. AVIF has been supported in all modern browsers since 2024 and is now the first choice for photographs and complex images on websites.
• Render-blocking JavaScript – Scripts loaded synchronously in the <head> stop the browser from rendering the page. Third-party scripts – chat widgets, ad networks, analytics tools – are the most common culprits.
• Hosting and TTFB – Time to First Byte measures how long the server takes to respond. Poor hosting or the lack of a CDN (Content Delivery Network) increases TTFB markedly.
• No caching – Without browser caching, every new visit has to reload all resources from scratch.
• Images that are not lazy-loaded – Images outside the viewport that load as soon as the page opens delay time-to-interactive unnecessarily – switch on lazy loading for them.

How do you measure page speed?

• Google PageSpeed Insights – The only tool that shows real CrUX field data from Chrome users alongside lab simulations. The starting point for every pagespeed analysis.
• Lighthouse – Built into Chrome DevTools. Runs synthetic tests and suggests fixes ranked by impact.
• GTmetrix – Combines Lighthouse with historical data and the option to test from different geographical locations.
• Google Search Console – The Core Web Vitals report shows aggregated field data per URL group over time.

Page speed and the choice of platform

The platform you build on sets the baseline for how fast the website can be. Webflow serves websites through the Fastly CDN, with automatic image optimisation and modern image formats – a strong starting point with no manual configuration. In practice, WordPress websites with lots of plugins on shared hosting are among the most common causes of poor Core Web Vitals results.

Statically generated Next.js websites served from a CDN are technically the fastest option when implemented correctly. There is more about that stack in Building a website with AI: how Claude Code, Next.js and Sanity work together.