The Numbers That Matter
Amazon calculated that every additional 100ms of latency cost them 1% in sales.
The Numbers That Matter
Amazon calculated that every additional 100ms of latency cost them 1% in sales.
Google observed that going from 0.4 to 0.9 seconds load time reduced traffic by 20%.
These are extreme cases with enormous volumes, but the trend applies at any scale. Slower means less engagement, more drop-offs, worse results.
Why Speed Matters So Much
Human patience is limited. In a world of instant gratification, waiting feels like unnecessary friction.
Speed is a signal of quality. A fast website conveys competence, investment, care. A slow website conveys neglect, limited resources, lack of attention.
Google uses it as a ranking factor. It's not the most important factor, but it's one more. And all else being equal, the fastest wins.
Where Time Is Lost
Unoptimized images. The most common and easiest problem to solve. Modern formats, appropriate sizes, lazy loading.
Excessive JavaScript. Every added library has a cost. That carousel with 200KB of JavaScript for something that could be CSS.
And much of that JavaScript isn't even yours: third-party scripts are often the worst performance problem on a website.
Poorly loaded web fonts. Flash of unstyled text while fonts load, or worse, invisible text until they load.
Cascading requests. Loading something that needs to load something else that needs to load something else. Each hop adds latency.
Slow server. All the rest optimized is useless if the server takes two seconds to respond.
Metrics to Measure
Largest Contentful Paint: how long it takes for the main content to be seen. Should be less than 2.5 seconds.
First Input Delay: how long it takes the page to respond to the first interaction. Should be less than 100ms.
Cumulative Layout Shift: how much the content moves while loading. Should be less than 0.1.
These are Google's Core Web Vitals, but beyond SEO, they are reasonable indicators of user experience.
The Necessary Balance
Optimization has practical limits. There comes a point where improving one more millisecond requires disproportionate effort.
The goal isn't perfection but good enough. A website that loads in 2 seconds doesn't need to obsess over reaching 1.5 if the effort is enormous.
Prioritize the big gains first. Optimizing images can reduce seconds. JavaScript micro-optimizations can reduce milliseconds.
Web performance isn't an isolated technical problem. It's part of the user experience, search engine ranking, brand perception.
No need to be obsessive. But ignoring it has measurable consequences in metrics that matter.
Every millisecond counts. Literally and figuratively.