If performance goes down, something is wrong Monica, but to be able to tell you what is wrong I will have to see a page on a site of yours where this is the case. Can you share a URL?
best,
frank
Thanks a lot Frank, it’s http://www.monicakingsley.co – upon installing and activating the plugin, Performance on PageSpeed moves from 60-62 usual to 34.
OK, I can now see the page and score without LYTE active, can you activate LYTE for a short time to I can see what changes (and what might be going wrong)?
Done right now, it’s activated. Default settings, no changes.
Thank you.
OK, the problem is that on mobile by default the YouTube embed is loaded immediately after the LYTE javascript executes. You should be able to change that by activating the “cache thumbnail locally” option so maybe try that.
That being said; as the video is lazyloaded already even when LYTE is not active, there will be little to no performance improvement to be expected from LYTE.
hope this clarifies,
frank
I did try activating the “cache thumbnail locally” option (Yes), but it had no good effect either (only somewhat less worse effect):
– Performance down to 57-58 from 60-31
– LCP from roughly 5.2s to 5.8s
– DOM size 870 from 858
Summing up, no improvement on status quo ante at all.
Do you agree with uninstalling and deleting the plugin, or any more good tricks up your sleeve?
Btw I use Speed Optimizer (formerly SG Optimizer), but it offers no power over influence whether a video is lazyloaded or not (unlike images).
Thank you and good evening.
If performance is the only reason you installed LYTE (and not privacy/ GDPR) then indeed you can deactive/ delete it.
Re other tricks; I don’t know “speed optimizer”, but check if there are options to defer the JS and CSS, if so that should improve FCP & LCP considerably?
Yes, performance was the only reason – I only seek to cut out the very extras that native YT embed code contains, but the solution to only get a thumbnail clicking which activates that extra YT code, is elusive.
Perhaps that’s equivalent to lazyloading the whole video that Frank confirmed is the case anyway, what do you think please?
I’m already defering JS, and even have defined scripts to exclude from deferral of render blocking JS… CSS is also both combined and minified, and within the PageSpeed test doesn’t offer a great savings potential (unlike excluding hero image from lazy loading, which I already did).
So, I only seek to slim down the load hit associated with single embedded video… any more ideas as to this single area please?
Thanks!
As far as I remember there was no mention of any youtube resources in the pagespeed insight report, which indicates it had not even loaded by virtue of it being lazyloaded. Based on that I don’t think there’s any further optimization to be done in that single area, no 🙂
Thank you Frank – I’m truly perplexed though. Upon removing the plugin (with that “cache thumbnail locally” option still toggled to Yes), I got first 57 Performance result only, and then twice in a row 77.
No other work or tweaking done.
Can there be an explanation to it please?
Have a nice day both.
The only way to know what happened is comparing the main KPI’s and the detailed recommendations really 🙂
Thank you Frank, it turned out 77 was a fluke, not repeatable. Back to the previous 60-62 score.
I also tried “AMP for WP – Accelerated Mobile Pages” plugin, but it didn’t result in any mobile improvement. I take it as having done everything that could have been done (apart from WebP images).
Thanks.
