You’re welcome @ScottWelch, and thanks for sharing the link.
I took a look at the CSS for your CTAs compared to other sites and it looks like the legacy CTA tool had an option to set border-width, which when left unset, defaults the CSS to 0 px.
It’s definitely CSS bloat, unfortunately, I couldn’t find a way to remove it - with that said, one line of CSS is likely not the culprit of performance issues, even with multiple CTAs on the same page.
The new CTAs aren’t running the same script as legacy CTAs - I don’t have a page with multiple, but checking the source code for the pop-up, it is loading individually…however, it is ld+json and lightweight compared to the old script. It’s unique per CTA rather than containing a huge amount of excess code.
My recommendation would be to test swapping out a single CTA on the page to compare the performance before and after. Then swapping all of the CTAs on the same page to further compare the results before deciding whether to implement across all pages.
With that said, I ran this page through Google Lighthouse and you have MUCH bigger fish to fry than the CTA script and CSS when it comes to performance - I wouldn’t spend a lot of time on CTAs before resolving the other issues first that will have a much bigger impact.
Your Largest Contentful Paint (LCP) is 8.6 seconds on mobile:
This has to do with how the background image is handled - it’s still there even though it’s not fully visible like it is on desktop.
Hiding this image on mobile would have an immediate impact.
On desktop it’s not as big of an impact:
It’s set to a background image and the file size is relatively small already:
it might seem counterintuitive, but if you set the background image to be the full height of the viewport, Google will view it as a true background image and not ‘relevant’ content so it won’t impact LCP.
LCP is your biggest opportunity:
I would prioritize LCP, FCP, Speed Index, then CLS when optimizing this page.
Hope this helps, happy to help further if you have questions!