Archived topic
Confusing Longest Contentful Paint Result
7 replies · Started by Michael on June 19, 2020
Hello guys,
I am getting a strange result regarding the site logo in the Google Web Dev test.
For some reason our bespoke page is getting a significantly higher score (10 points) than our faq page in spite of basically being the same page layout with a similar amount of content.
The difference appears to be that the faq page is saying the logo is flagged as largest contentful page. It's confusing me because the logo is identical on both pages, and is not significantly large (only 150px x 150px and compressed jpg):
https://storage.googleapis.com/support-forums-api/attachment/thread-54220664-3485006926089899050.jpg
https://storage.googleapis.com/support-forums-api/attachment/thread-54220664-3485006926089898181.jpg
Here is the code I am using for the logo image:
<picture>
<source srcset="https://reclaimdesign.org/wp-content/uploads/reclaim-design-site-logo.webp" type="image/webp">
<source srcset="https://reclaimdesign.org/wp-content/uploads/reclaim-design-site-logo.jpg" type="image/jpeg">
<img src="https://reclaimdesign.org/wp-content/uploads/reclaim-design-site-logo.jpg" srcset="https://reclaimdesign.org/wp-content/uploads/reclaim-design-site-logo-mobile2.jpg 0.5x, https://reclaimdesign.org/wp-content/uploads/reclaim-design-site-logo.jpg 1x, https://reclaimdesign.org/wp-content/uploads/reclaim-design-site-logo-large.jpg 2x" alt="Reclaim Design Logo" />
</picture>It also seems like the webp image is not mentioned in the largest contentful paint result for the faq page. In my understanding it should be showing up rather than the jpg?
And here are the full results:
Bespoke page web dev result - https://lighthouse-dot-webdotdevsite.appspot.com//lh/html?url=https%3A%2F%2Freclaimdesign.org%2Fbespoke-furniture
FAQ page web dev result - https://lighthouse-dot-webdotdevsite.appspot.com//lh/html?url=https%3A%2F%2Freclaimdesign.org%2Fhome-decor-products-faq
Thank you for your help. It's driving me mad as I can't seem to improve the performance score at all! I have asked on Google Webmaster Forums but haven't found any resolution there so wanted to ask for your advice in case you can see a way to resolve.
Michael.
Hi there,
looking at your TTFB for both pages - the FAQ is significantly longer. Might be worth running Query Monitor to see what the notable differences are.
Regarding the webp image - Chrome is actually loading the webp so not an issue. But i can explain why lighthouse is flagging it incorrectly or as the LCP
Hi David,
Thanks very much for getting back to me. I installed the query monitor and I can't see any noticeable differences, except that the FAQ page makes 1 less HTTP API call. Both pages have 18 php errors, 1 duplicate query, 17 scripts, 14 styles and 3 transient updates
https://www.dropbox.com/s/ax3vy994vh52it9/bespoke1.jpg?dl=0
https://www.dropbox.com/s/jca1rp0tpdihhej/bespoke2.jpg?dl=0
https://www.dropbox.com/s/yfyvuhe7fb5q7fm/bespoke3.jpg?dl=0
The php errors mention generatepress but I am not able to understand if these are actually errors or not!
You'll definitely want to address those PHP errors, as they won't help with performance. They won't be from GP itself, but may be from a hook you've added.
As for the other issue - are the scores similar if you use a regular logo/disable the logo temporarily?
Thanks for getting back to me Tom. If I disable the code snippet for the logo and then test in web dev the result seems to be the same (or similar - within a couple of points - seems to vary each time I run, I guess to do with the CDN/server). I've found the snippet that caused some of those errors and disabled it, and the other notices seem to be involving the Ecwid plugin so I have written to them.
I've tried all sorts today to improve the page result and basically nothing has worked. Amazing. I would have thought something would have made a significant difference (if not an improvement then the opposite!)
If you go into developer tools and go to the "Network" page, then refresh, you'll get a good idea of what's going on with the page.
You'll see the two pages are basically identical, but the actual document is taking longer to load on the FAQ page. I'm assuming this is something to do with some PHP running on that page that isn't running on the other pages.
Thanks a lot Tom. You're right the network tool is good for seeing what is running, and I spent all day dissecting it yesterday but could not find anything that was jumping out from the FAQ page that is causing the lag. Anyway, I appreciate your input and pointers! Have a good Sunday.
Michael.
Thanks, you too!
