Hi everyone, I changed to Cloudflare couple of hours ago and since then getting slower performance than before, checking google pagespeed insight shows lower score from 80 to 29 for mobile: https://developers.google.com/speed/pagespeed/insights/?url=https%3A%2F%2Fwww.yogatodo.com%2F&tab=mobile
Reduce server response time - cf probably add 100ms to this number so not a lot to do
Leverage browser caching - again you can fix your code,but in cf you can change it in: caching -> Browser Cache Expiration
Optimize images - again nothing to do with cf, but you can fix it with cf , by going to: speed -> polish
can you tell what went wrong specific in the test before moving to cf? which points changed? and to what?
First of all, thank you for your reply I appreciate it!
To see the difference before cf we can compare it to our production server: https://developers.google.com/speed/pagespeed/insights/?url=https%3A%2F%2Fwww.yogatodo.es%2F&tab=mobile
- Staging server (.com) - W3 Total Cache + Cloudflare (dedicated server US)
- Production server (.es) - WP Fastest Cache + No Cloudflare (shared hosting Spain)
Before moving to cf I was installing “W3 Total Cache” on the staging server, which got better results than “WP Fastest Cache” on Gtmatrix and PageSpeed Insight, then changed to cf to see if we can improve the performance even more but for some reason it’s slower.
you are comparing different servers with different plugins etc…
did you test your server response time? so you can say in 100% confident “my server responding in 500ms but with Cloudflare its taking is 2.5sec to response” or something like that?
Yes it was faster for sure but I don’t remember the exact numbers, what I can do is create another copy of the website on the same server without cf and then we can compare.
Check these similar threads:
End result… speed ‘scores’ are almost nonsensical. You have to look at the overall speed of the fully loaded page.
Best thing to do to get a better score is to host as close as possible to the target audience and not use any CDN at all.
No no… a CDN can help dramatically with speed - even the ‘score’. With proper tuning of the CDN and the site you can definitely get a 100%.
I don’t get a 100% because of google’s own analytics!
This is not a discourse website… for some websites I suppose CDN can help…
But checked the topic and it’s not about discourse so idk
pagespeed insight score is relative but doesn’t tell how fast your site is
here’s a clear example where @jules linked insight score for his site is 98/99 score versus my forum at 82-92 score (i use adsense + google analytics) but WPT filemstrip shows page’s visual render time in clear view as to page load speed experienced by visitors
instead of pagespeed scores, focus on the proper metrics like first paint, first contentful and meaningful paint and document load times. As those are metrics Google is looking for https://developers.google.com/web/fundamentals/performance/user-centric-performance-metrics and where Cloudflare will help you optimise for when it’s enabled and properly configured and optimised for your specific sites’ style elements.
No matter what, speed scoring is not a perfect science. When I load in firefox I get much faster results - caches disabled. And you’ll see they don’t jive with online speed tools.
Like Einstein and @eva2000 said: It’s all relative.
That’s 4.5MB transferred in less than a second. Includes two 1080p video clips. That most likely cannot happen without a CDN.
Of course relative to the speed of your ISP connection if you’re testing via browser dev tools network tests Webpagetest default is 5Mbps cable speed profile usually.
This topic was automatically closed after 30 days. New replies are no longer allowed.