Site Kit By Google

Thanks for reaching out. Can you confirm the plugin or theme feature you’re using for your post views? This may not be linked to your Google Analytics or Search Console account, therefore the numbers wouldn’t match.

1 – Could this be due to AMP analytics script inserted on my theme. (note that the theme is optimized for amp)? Also

No, I don’t think this is the case. If you share the post views plugin I’m happy to check this however

2 – Noticed this notification on my theme AMP section (For Google Analytics please consider using Site Kit by Google…

This is a recommendation so that you don’t have to manually insert your Analytics snippet within the AMP plugin. You can indeed remove the snippet you suggested if you are using Site Kit. Just be sure to keep the default option of allowing Site Kit to insert your Analytics snippet on your site.

Hi James

Thanks for your fast reply

What do you intend by – Can you confirm the plugin or theme feature you’re using for your post views?

The theme I use has your plugin (Official AMP) integrated. In the APM settings section I have added the script I’ve mentioned earlier.

I removed the script and tested the results from analytics LIVE and didn’t notice any difference. It seems like all the results (amp and non-amp) are displayed the same with no /amp/ tag in the end of the search. Correct me if I’m wrong please.

So you are saying that it won’t matter if I remove or not the script the results are the same? If yes, then how the discrepancy is explained?

Can you please conduct a test?

Here in your post (https://amp-wp.org/success-on-the-web-with-amp-and-google-site-kit/) you’ve clearly explained the setup but as i mentioned i didn’t notice the diff

• This reply was modified 3 years, 5 months ago by EddieG.

Hi @eddievet,

What do you intend by – Can you confirm the plugin or theme feature you’re using for your post views?

I should have been more clear. When I check your site I notice the “views” above the title. I am wondering is this a theme feature or are you using a plugin. I would expect this isn’t linked to your Analytics or Search Console account. If true it’s normal that the figures in the screenshot you shared wouldn’t match.

I removed the script and tested the results from analytics LIVE and didn’t notice any difference. It seems like all the results (amp and non-amp) are displayed the same with no /amp/ tag in the end of the search. Correct me if I’m wrong please.

If you removed the script and allowed Site Kit to insert your Analytics snippet then all visits to your site will be attributed to the same Analytics property, whether to your canonical (non AMP) or AMP visits. I’m unsure what you are referring to in relation to search however if you get visits to your site from search engines this is the same, whether visitors land on AMP URLs or non AMP URLs they will be attributed to the same Analytics property.

So you are saying that it won’t matter if I remove or not the script the results are the same? If yes, then how the discrepancy is explained?

Correct. If you’ve allowed Site Kit to insert your snippet it will do so on both AMP and non AMP URLs. The discrepancy exists because there are different tracking methods used for your “Page views” feature.

Can you please conduct a test?

While we don’t support third party plugins or theme page visit features you can check with the respective theme or plugin to check how they calculate page visits. Typically it’s normal they won’t match with a users Analytics or Search Console data.

Hopefully the above makes sense, let me know if you have any further queries. If you want to perform your own test using a separate AMP compatible page views counter you can use the Page Views Counter plugin. It will only start collecting counts based on visits after activation, not on historic, Search Console or Analytics data.

Hi James

First I want to thank you for the continuous support.

I got in touch with the theme makers as per your suggestion. However, i need to ask you about some strange results noticed in the last 24 hours.

I have removed the script and two things are
1 – The views/traffic has reduced by at least 50% and when i look at analytics live data i see that most users are from mobile devices (see image – https://pasteboard.co/NqluxrHqp8c4.jpg) but i can’t tell which are AMP and non-AMP.

2 – Apart the traffic lose, I got better Bounce rate (10-15% increase) and time spent (70%-80% increase) on site now.

Could the script removal be the reason of these changes (bearing in mind you saying that it won’t make any difference)?

• This reply was modified 3 years, 5 months ago by EddieG.

Not a problem. While I’m awaiting a response I’d be happy to provide a temporary fix for you however please do allow me some time to address this. I’m awaiting a response from my query from the theme author.

In relation to your bounce rates and traffic these would be best answered by the product experts at the Analytics Help Center. You can use the below guide to setup an AMP specific report:
https://support.google.com/analytics/answer/9264222?hl=en

The product experts at the Help Center should be able to provide any further insights when it comes to any discrepancies. Removing the AMP Analytics snippet and replacing that with a Site Kit placed Analytics snippet doesn’t have any impact on your sites bounce rate, it’s just a replacement insertion method.

Hi James

I replaced the script and now I get different result compare to when I removed the script.

I understood that when i removed the script i hadn’t AMP on, neither i allowed analytics/Site-Kit to insert a script Automatically in the header, so i was running on normal mobile view

You may interested to know these results.

The traffic has improved and has increased above 50%+ but the bounce rate jumped to 86%+ compare to about 72% without the script, the time on site also decreased by about 50%.

These numbers are the same as before removing the script in the past 24 prior to replacing it.

Is giving false or inaccurate reading? I don’t believe that the users behavior changed over night.

That leaved me to believe that there is something to do with AMP. Looks like the the AMP has improved the traffic but it has produced bad Bounce rate and bad Time on site.

Thanks,
Eddie

Hi Eddie,

Thanks for sharing the analysis you’ve been performing on your site. While I can’t be certain of any bounce rate differences the theme you’re using have it’s own AMP templates, which are different in appearance to your non AMP URLs (ie. no sidebar widgets). As there are various factors when it comes to bounce rates you may find better assistance when it comes to such queries using the Analytics Help Center.

While we can’t assist with bounce rate queries can you confirm that I do see the same Analytics property for both your canonical and AMP URLs, so the setup is correct.

Hi James

My attention was to point out the differences with the AMP on/off. My main focus was not on the bounce rate but on its results but when AMP is on/off.
Those data changed overnight but users don’t.

However, thanks for keeping up supporting!
Eddie

Hello James

I have finally found out what causes the issue. Based on your message (Can you confirm the plugin or theme feature you’re using for your post views?) I made contact with the plugin developer (WordPress Popular Posts) and he said that it is due to his plugin as it is not compatible with AMP.
So the issue is found. Now, Is there a way to fix this? (apart from the developer to make the his plugin compatible)

Thank you!

@eddievet Nice troubleshooting, thanks for the update. Unfortunately it’s up to plugin and theme developers to ensure their products are AMP compatible. I believe the Page Views Count is AMP compatible and could be used as an alternative.

Hi James

Thanks for the support. Good idea to use the suggested plugin but i can’t as the one I use has other features included.
The plugin dev is under way to develop a version with AMP support. So, will wait to see how it develops.

p,s
Can you suggest me a way to do an A/B testing to continue with the understanding the GA results and compare them with AMP results.
Basically i would like to test non-AMP with AMP versions

Many thanks again!

No problem, happy to help. If you’re just looking to compare AMP traffic versus your canonical traffic within Analytics you can use the guide below to setup and view such reports:
https://support.google.com/analytics/answer/9264222?hl=en

For AB testing you may also consider Optimize.

As we didn’t receive a response I’ll mark this as resolved. Feel free to open a new support topic if you require any further assistance.