I keep seeing this everywhere lately.
“My websites score 90+ on performance tests.”
And every time I think… okay, but does it actually work?
Don’t get me wrong, performance matters. A slow website is frustrating. People leave. It absolutely has an impact. But somewhere along the way, we started treating tools like Google PageSpeed Insights as the ultimate judge of quality… and they’re just not.
I’ve seen “perfectly optimized” websites that load fast, score beautifully, and still feel off. Empty. Clunky. Hard to follow. And I’ve seen websites with more average scores that feel smooth, clear, and actually bring in leads consistently.
I think part of the issue is that developers are often pushed (or choose) to focus on what’s easy to measure. Scores, milliseconds, audits. Things you can screenshot and put in a proposal. But a website doesn’t live inside a testing tool. It lives in the hands of a real person, on a real device, usually in a hurry.
And this is where the gap shows.
A good website is never just “dev” or just “design”. It’s the result of both working together. If development chases performance at the expense of layout, hierarchy, or storytelling, the site might be fast… but it won’t connect. If design ignores performance completely, the site might look great… but it won’t hold attention.
The balance is where things start to work.
That’s also why I’ve been leaning more into building systems rather than just pages. Not in a complicated way, but in a way where structure, performance, and design decisions support each other from the start. So you don’t end up fixing speed later by breaking the experience, or fixing design by slowing everything down.
Because users don’t care about your score. They care about how quickly they understand what you do. How easy it is to navigate. How it feels on their phone. Whether they trust you enough to take action.
Sometimes chasing that perfect number does more harm than good. You start stripping away design, delaying important elements, over-optimizing things that didn’t need fixing in the first place. All for a nicer-looking metric.
In reality, for most websites, if it loads in around 2–3 seconds, feels responsive, and doesn’t jump around or lag… you’re already doing very well. That alone puts you ahead of a big part of the internet.
Performance is important, yes. But it’s not the goal. It’s just the baseline.
The real goal is a website that communicates clearly, feels good to use, reflects the brand, and actually converts.
A high score is nice.
A website that works is better.