These clubs remind me of this tweet (as someone that has always been the perf guy wherever I worked):

---

Also, hot take: users care way less about performance than you think. They want "fast enough", but we're over-indexing on "as fast as possible" instead of caring about other things that matter more to users.

Source: https://mobile.twitter.com/DavidKPiano/status/15787403709971...

luckylion

I agree completely regarding "users want fast enough".

Also: for regular sites, html/css/js optimization is less important than server location. If your server is in Europe and your user is in the US, that's the big one, not your HTML, CSS or JS.

And if you're fetishizing over Lighthouse scores, stop. It's only a very rough measure and shouldn't be treated as a goal itself.

lelandfe

> It's only a very rough measure and shouldn't be treated as a goal itself

Lighthouse is a synthetic testing tool. It can be incredibly useful when you connect it to your CI build process and track scores over time – revealing how code changes affect (some) front-end performance metrics.[0] Of course, some scores are non-deterministic so you'll need to run it multiple times.[1] And you should be tracking real user data as well to validate it.

Using PageSpeed Insights to benchmark a site is akin to loading your website a single time on a single device and saying "wow, my site is slow"

Using Lighthouse in your browser to benchmark a site is akin to wasting your time.

[0] https://github.com/GoogleChrome/lighthouse-ci

[1] https://developers.google.com/web/tools/lighthouse/variabili...