Issue with GA tracking and Native AMP
-
Hi everyone,
We recently pushed a new version of our site (winefolly.com), which is completely AMP native on WordPress (using the official AMP for WordPress plugin). As part of the update, we also switched over to https. In hindsight we probably should have pushed the AMP version and HTTPS changes in separate updates.
As a result of the update, the traffic in GA has dropped significantly despite the tracking code being added properly. I'm also having a hard time getting the previous views in GA working properly.
The three views are:
- Sitewide (shop.winefolly.com and winefolly.com)
- Content only (winefolly.com)
- Shop only (shop.winefolly.com)
The sitewide view seems to be working, though it's hard to know for sure, as the traffic seems pretty low (like 10 users at any given time) and I think that it's more that it's just picking up the shop traffic.
The content only view shows maybe one or two users and often none at all. I tried a bunch of different filters to only track to the main sites content views, but in one instance the filter would work, then half an hour later it would revert to no traffic. The filter is set to custom > exclude > request uri with the following regex pattern:
^shop.winefolly.com$|^checkout.shopify.com$|/products/.|/account/.|/checkout/.|/collections/.|./orders/.|/cart|/account|/pages/.|/poll/.|/?mc_cid=.|/profile?.|/?u=.|/webstore/.
Testing the filter it strips out anything not related to the main sites content, but when I save the filter and view the updated results, the changes aren't reflected. I did read that there is a delay in the filters being applied and only a subset of the available data is used, but I just want to be sure I'm adding the filters correctly.
I also tried setting the filter to predefined, exclude host equal to shop.winefolly.com, but that didn't work either.
The shop view seems to be working, but the tracking code is added via Shopify, so it makes sense that it would continue working as before.
The first thing I noticed when I checked the views is that they were still set to http, so I updated the urls to https. I then checked the GA tracking code (which is added as a json object in the Analytics setting in the WordPress plugin. Unfortunately, while GA seems to be recording traffic, none of the GA validators seem to pickup the AMP tracking code (adding using the amp-analytics tag), despite the json being confirmed as valid by the plugin.
This morning I decided to try a different approach and add the tracking code via Googles Tag Manager, as well as adding the new https domain to the Google Search Console, but alas no change.
I spent the whole day yesterday reading every post I could on the topic, but was not able to find any a solution, so I'm really hoping someone on Moz will be able to shed some light as to what I'm doing wrong.
Any suggestions or input would be very much appreciated.
Cheers,
Chris (on behalf of WineFolly.com) -
Lots going on here, so, a laundry list of follow up questions and thoughts for you...
Are you seeing AMP results showing up in the Search Console? Are you seeing them indexed as intended?
If you're doing Native AMP, you won't be able to diagnose pages by /amp URL types of formatting. It might be worth trying to fire off an event, or custom dimension in GA, for AMP = Yes / No or something like that.
For the sitewide view, have you tested loading pages on a private browser and incognito mobile browser and seeing if they show up in GA realtime in each of the 3 views when they're supposed to?
It looks like you might be using Cloudflare - I haven't dealt with an AMP site that uses it, but have you checked whether there are compatibility issues or anything you need to activate?
Are any Google Tag Manager pages set to fire on HTTPS only?
Are any GA filters in place that specify HTTP/HTTPS that need to be broadened?
Your Amp Analytics code seems to match the one on a site that is functioning as intended, so I don't think it's a formatting issue.
For the GA view filter - it seems like you should be able to simply include/exclude traffic to shop.winefolly.com - why the added complexity beyond that?
Got a burning SEO question?
Subscribe to Moz Pro to gain full access to Q&A, answer questions, and ask your own.
Browse Questions
Explore more categories
-
Moz Tools
Chat with the community about the Moz tools.
-
SEO Tactics
Discuss the SEO process with fellow marketers
-
Community
Discuss industry events, jobs, and news!
-
Digital Marketing
Chat about tactics outside of SEO
-
Research & Trends
Dive into research and trends in the search industry.
-
Support
Connect on product support and feature requests.
Related Questions
-
Only fraction of the AMP pages are indexed
Back in June, we had seen a sharp drop in traffic on our website. We initially assumed that it was due to the Core Update that was rolled out in early June. We had switched from http to https in May, but thought that should have helped rather than cause a problem. Until early June the traffic was trending upwards. While investigating the issue, I noticed that only a fraction (25%) of the AMP pages have been indexed. The pages don't seem to be getting indexed even though they are valid. Accordingly to Google Analytics too, the percentage of AMP traffic has dropped from 67-70% to 40-45%. I wonder if it is due to the indexing issue. In terms of implementation it seems fine. We are pointing canonical to the AMP page from the desktop version and to the desktop version from the AMP page. Any tips on how to fix the AMP indexing issue. Should I be concerned that only a fraction of the AMP pages are indexed. I really hope you can help in resolving this issue.
Technical SEO | | Gautam1 -
Weird Google indexing issues with www being forced
IM working on a site which is really not indexing as it should, I have created a sitemap.xml which I thought would fix the issue but it hasn't, what seems to be happening is the Google is making www pages canonical for some of the site and without www for the rest. the site should be without www. see images attached for a visual explanation.
Technical SEO | | Donsimong
when adding pages in Google search console without www some pages cannot be indexed as Google thinks the www version is canonical, and I have no idea why, there is no canonical set up at all, what I would do if I could is to add canonical tags to each page to pint to the non www version, but the CMA does not allow for canonical. not quite sure how to proceed, how to tell google that the non www version is in fact correct, I dont have any idea why its assuming www is canonical either??? k11cGAv zOuwMxv0 -
Href lang issues - help needed!
Hi, I have an issue with Google indexing the US version of our website rather than the UK version on Google.co.uk. I have added hreflang tags to both sites (https://www.pacapod.com/ and https://us.pacapod.com/), have updated and submitted an XML sitemap for each website and checked that the country targeting in search console is set-up correctly but Google are still indexing the wrong website. I would be grateful for any assistance with this issue. Many thanks Eddie
Technical SEO | | mypetgiftbox0 -
Interesting indexing issue - any input would be greatly appreciated!
A few months ago we did SEO for a website, just like any other website. However, we did not see crawl/indexing results that we have with all of our other SEO projects - the Google webmaster tool was indicating that only 1 page of the site (although only 20 pages) was indexed. The site was older & originally developed in Dreamweaver, so although that shouldn't have been an issue, we were desperate to solve the problem & ended up rebuilding the site in WordPress. While this actually helped increase the number of pages on the site that Google indexed (now all 20) - we are still seeing strange things in the search results. For example, when we check rankings manually for a particular term, the new description is showing, however, it is displaying the old title text. Does anyone know what the problem could be? Thank you so much!!
Technical SEO | | ZAG0 -
Google Places & Multiple Accounts?
As an agency that manages multiple accounts should I have all my Google Place accounts under one account or should I create a separate account for each client with a unique username and password for every client? Thanks,
Technical SEO | | fun52dig
Gary Downey0 -
Same Video on Multiple Pages and Sites... Duplicate Issues?
We're rolling out quite a bit of pro video and hosting on a 3-party platform/player (likely BrightCove) that also allows us to have the URL reside on our domain. Here is a scenario for a particular video asset: A. It's on a product page that the video is relevant for. B. We have an entry on our blog with the video C. We have a separate section of our site "Video Library" that provides a centralized view of all videos. It's there too. D. We eventually give the video to other sites (bloggers, industry educational sites etc) for outreach and link-building. A through C on our domain are all for user experience as every page is very relevant, but are there any duplicate video issues here? We would likely only have the transcript on the product page (though we're open to suggestions). Any related feedback would be appreciated. We want to make this scalable and done properly from the beginning (will be rolling out 1000+ videos in 2010)
Technical SEO | | SEOPA0 -
Domainchange & Relaunch
Hi SEOMOZERS, We are planning to do a relaunch (new content, new design) + a domainname change (because it contains a keyword). Some of my team members think it is best to do the domain change already on the old-version to see if there might be some problems. Whereas I think it would be best to implement the domainchange on the new version because otherwise it can be suspected we did the domainchange only for SEO purposes. What are you guys thinking? Thanks Gregor
Technical SEO | | GregorHendrych0 -
Canonicalization - duplicate homepage issues
I'm trying to work out the best way to resolve an issue where Google is seeing duplicate versions of a homepage, i.e. http://www.home.co.uk/Home.aspx and http://www.home.co.uk/ The site runs on Windows servers. I've tried implementing redirects for homepages before (for a different site on a linux server) and ended up with a loop, so although I know I can read lots of info (as I have been doing) and try again, I am really concerned about getting it wrong. Can anyone give me some advice on the best way to make Google take just one version of the page? Obviously link juice is also being diluted so I need to get this sorted asap. Thanks.
Technical SEO | | travelinnovations0