Hi @therob,
I think you’re looking at cached .js files, because I see the code injected just before the body closing tag. I have changed the plugin version in the git so your browser will pull the latest .js files. Can you check again?
But I think the test itself already used the updated files, as the test probably doesn’t cache, and you already saw some improvement, so apparently moving it to the end of the body tag helps a bit, but not enough in your case.
I have tested on six sites of ourselves using Complianz GDPR, among which complianz.io and really-simple-ssl.com, and found no significant CLS values.
I’ve run this test on several other sites with Complianz running, most just return a CLS of 0, or near it. Of the 6 sites, only one returned a significant CLS. After trying some banner positions, it dropped to 0.03. After that I haven’t been able to reproduce the issue on any tool, not even after testing 6-7 times. This site also used a theme which causes a layout shift of itself. I’m wondering if this might be the result of a theme with a CLS issue in combination with complianz.
Summarizing:
– Test results greatly differ from test to test, with same setup. This leads to questions regarding in how much we can depend on these results to be trustworthy.
– As a user, I don’t see an actual layout shift from the banner. Possibly the measurement technology is failing here.
– Of 6 sites, only one showed a significant CLS the first two tests. After that, the CLS dropped to 0.03, even after repeatedly running the test.
– To be able to make improvements, I need a consistent test, which shows the same results for the same setup each test.
Next steps:
– I’m wondering if this can be related to a theme you’re using. Do you have the same results with other themes (e.g. a default WordPress theme)?
– If you run 10 tests, how many times do you get a bad CLS?
– Do you use tools which defer CSS? This might cause different layout behaviour.
– Do you get the same results with different banner positions?