Google Optimize Redirect Test Q&A: Conversions and sessions discrepancies, URL paths, anti-flicker, utm_expid, SEO questions, etc

If you want to test a lot of changes on a page, a completely different design or landing page concept, using a URL redirect test is more appropriate than a standard A/B test.

And with most of our clients using Google Optimize (due to its simplicity and reliability), we get a lot of questions regarding Google Optimize redirect tests (a.k.a. split URL test).

So in a hope to help other users facing similar difficulties, we have collected these questions and posted our answer in this article.

In this way, everyone facing similar misunderstandings or issues, can find a working solution here.

List of questions:

  1. Why conversions are not appearing in Google Optimize but are appearing in Google Analytics?
  2. How to keep the same path in all new URLs in a redirect test?
  3. Variant preview mode of our A/B redirect test shows the original version.
  4. Redirecting based on URL parameters is not working.
  5. Getting error “Analytics tracking code not found. The web page is not tagged with Google Analytics tracking code”
  6. Do I need to install the anti-flicker snippet if I run only redirect tests?
  7. How can I remove utm_expid from pages tested in Google Optimize?
  8. I have targeted 100% of traffic in Google Optimize but when comparing to Google Analytics data, only 80% of sessions appear to be in experiment.
  9. Original page never redirects to variation page in live test
  10. Will a redirect test have a negative impact on the SEO performance?

1. Why conversions are not appearing in Google Optimize but are appearing in Google Analytics?

Question: We are running a redirect test for some time and see no conversions in Google Optimize reports but when checking Google Analytics, we see conversions. In Google Optimize we are seeing just the experiment sessions.

Answer: Have you linked your Google Optimize account with a User ID enabled view? If so, can you link it to a regular view as at the moment, Google Optimize does not support User ID enabled views and this causes conversions to show as 0.

Another reason for no conversions appearing in Google Optimize vs Analytics is timing. It takes about 1 hour to see data in Google Analytics (although Google states that it takes up to 24 hours) but in Google Optimize, data appears much later, after about 4 hours.

2. How to keep the same path in all new URLs in a redirect test?

Question: Can someone help us pass all sub-paths from domain A to domain B in a redirect test? We are redirecting domain.tld to beta.domain.tld and it works fine for homepage, but when adding a path to our domain, it is not redirecting.

We want all paths from domain.tld to redirect to beta.domain.tld (ex: domain.tld/path/ should redirect to beta.domain.tld/path/). Thanks!

Answer: You can achieve this behavior with an “Advanced Redirect”. Advanced redirect tests allow you to modify parts of a URL by replacing, modifying, or adding strings to it.

To create an advanced redirect test, when adding a variant, choose “Advanced Redirect” as the redirect type.

advanced redirect google optimize

You can read more about advanced redirect tests here.

3. Variant preview mode of our A/B redirect test shows the original version.

Question: When I preview the variant from our A/B redirect test, I always see the original version instead of the variant. Can you help me view the variant instead of original version? Thanks!

Answer: When previewing the variant, instead of clicking on “Web Preview”, click the “Debug” button. When doing so, you will not be automatically redirected and the original page will still load, however a debugging panel will load, showing you where the page should be redirected to. If the Debug panel does not show that the page will be redirected, then check the targeting conditions of your experiment.

If Debug panel is not loading when “Debug mode” is enabled, check to see if Google Optimize is installed on that page and you are not using an ad-blocking browser extension (some browsers block Google Optimize by default when using incognito mode).

4. Redirecting based on URL parameters is not working.

Question: I am trying to run a redirect test where the original page and the variant page is differentiated only by URL parameters:

  • Original page: example.com/?run=register
  • Version 1 page: example.com/?run=register2

While testing this, it seems that Google Optimize is not differentiating based on the URL parameters and redirects all pages starting with example.com/. How can I fix it so it would only redirect traffic from example.com/?run=register ?

Answer: Are you using the default “URL matches” targeting condition? If so then this might explain the behavior as “URL matches” ignores query parameters when checking for URL matches.

To target pages based on URL parameters, click on the “Customize” button from the “Audience targeting” section and select “Query parameter” as the rule type.

query parameter targeting optimize

On the next screen, from the variables list, click on “Create new” and create a new query variable. In your case, the query variable should be “run”. After this, from the targeting condition select “equals” and enter the value of the query parameter you want to target. In your example you should write “register”.

optimize targeting

After this, the experiment should only run for users who go to example.com/?run=register

5. Getting error “Analytics tracking code not found. The web page is not tagged with Google Analytics tracking code”

Question: We created an experiment in a hope to run a redirect test but when running the Optimize diagnostics to see if everything works correctly, we get an error saying: Analytics tracking code not found. The web page is not tagged with Google Analytics tracking code for the Analytics property configured for this experiment. You won’t be able to measure usage of your site nor monitor the performance of Optimize experiments. To run experiments, you must first install the Analytics tracking code (either analytics.js or gtag.js ).

analytics tracking code not found optimize

We are firing our Optimize and Google Analytics tags via GTM and I followed the official installation guide when installing Optimize so I think it should be installed correctly. Can you please help us fix this error?

Answer: Usually this means that Google Analytics tracking code is not installed on the page, but if you followed every step from the official installation guide and both Google Analytics and Optimize are installed correctly, then the most likely reason for getting this error is using an ad-blocking extension.

You could try and disable any ad-blocking/privacy extension from your browser or if you use an anti-virus or a VPN software, check if it has an option to block ads/tracking and disable it too.

If you disabled all ad-blocking extensions or software and see the Google Analytics hit being sent in your browser console, but still get this error, then the second most likely reason could be an internal error from Google. This happened in the past a few times but the Optimize team fixed the issue in few hours.

6. Do I need to install the anti-flicker snippet if I run only redirect tests?

Question: We plan to run only redirect tests and I was wondering if we still need to add the anti-flicker snippet. Or redirect tests are not prone to flickering as the other types of tests?

Answer: Yes, the anti-flicker snippet needs to be installed even if you run redirect tests. Before the redirect happens, Google Optimize needs to load on the page and in this short time interval, the initial page will be visible, thus the need for anti-flicker snippet. Here is the official guide on how to install the anti-flicker snippet.

7. How can I remove utm_expid from pages tested in Google Optimize?

Question: We run a couple of redirect tests in Optimize and noticed that when users get redirected to the variant, the utm_expid and utm_referrer parameters are being added at the end of the URL.

Is there a way to remove them?

I read that they may interfere with gclid parameter and other UTM parameters used for tracking campaigns.

Answer: Indeed, when running redirect tests, Google Optimize will add the utm_expid and utm_referrer parameters at the end of the URL but they should not interfere with gclid or UTM parameters.

You can easily test this by going to the original page with some UTM parameters (or the gclid) added in the URL and see them being re-added to on the variant URL, after the original page redirected to the variant page.

So domain.com/originalpage/?utm_source=test&utm_medium=test&utm_campaign=Test &gclid=EAIaIQobChMIj7znspiB6gIVRIuyBh3dfweLEAAYAiGAEgLuM_D_BwE will be redirected to domain.com/newpage/?utm_source=test&utm_medium=test&utm_campaign=Test &gclid=EAIaIQobChMIj7znspiB6gIVRIuyBh3dfweLEAAYAiGAEgLuM_D_BwE&utm_expid=.cOf44FksRzyZDgBCGueSEA.0&utm_referrer

8. I have targeted 100% of traffic in Google Optimize but when comparing to Google Analytics data, only 80% of sessions appear to be in experiment.

Question: I have created a redirect test where I targeted 100% of traffic but when I am comparing the numbers with Google Analytics, only 80% of what is reported in GA appears to be in the experiment as reported by Google Optimize.

So in GA I see 709 sessions for original page and variant 1 while in Optimize I only see 567 experiment session for the 2 pages.

Answer: There can be 2 causes for the sessions discrepancy reported in Google Analytics vs Google Optimize: reporting delays and anti-flicker snippet time out.

Since the data you see in Optimize is first processed in Google Analytics and then pushed to Optimize, GA reports tend to have fresher data, which means more experiment sessions being reported. And since the data push can take up to 12 hours (usually it takes less than 4 hours), this only increases the discrepancies between the GA and GO.

Another reason for the discrepancy is anti-flicker snippet time out. If you use the anti-flicker snippet, a visitor will not be added to the experiment if the anti-flicker snippet times out.

9. Original page never redirects to variation page in live test

Question: I have created a redirect test in Optimize and set the traffic allocation to 50% but no matter how many times I go to the original URL, I am never redirected to the variation page. I ran the Optimize diagnosis tool and it showed that Optimize is installed correctly and the preview links also work correctly.

How can I be redirected to the variation page instead of seeing the original page?

Answer: If the diagnosis tool showed that Optimize is installed correctly, then you can clear your cookies and revisit the original page.

Or, you could open an incognito window in your browser (make sure no previous incognito windows are opened) and visit the original page.

Repeat this process until you will be redirected to the variation page. If the page you are testing is visited by a lot of users at that time, you might have to repeat this process several times.

On high traffic pages, you can visit the original page up to 15-20 times until you are redirected to the variation page.

10. Will a redirect test have a negative impact on the SEO performance?

Question: We want to run a redirect test but we are not sure of the impact it would have on SEO performance. Will it decrease our rankings for the testing page? Should we add a noindex and nofollow meta tag to the variant page?

Answer: If configured correctly, a redirect test should have no negative impact on SEO. You should not add any noindex or nofollow tags on the variant page, however you do need to place a canonical attribute on the variant page and set the value to the original page.

So on the variant page, in the head, you need to add the following tag:

In this way, search engines will be informed that the variant page has no added value compared to original page and only the original page needs to be displayed in search engine results.

These were some of the questions we got about Google Optimize redirect tests. If you have other questions on this topic or something is still not clear, feel free to add a comment below.

6 Comments Leave a reply

  1. Ah, so sad that we can’t remove the utm_expid parameter from the URL as if a user bookmarks the page or shares it in the social media, it will have that parameter added to the URL.

    1

    0
  2. We are having trouble with the 302 redirect being applied – both original and variant have 200 status codes and have now therefore caused both versions to be indexed. Any thoughts? All other protocols have been followed.

    0

    0
    1. Have you defined the canonical URL on the redirected page? Every time you run redirect tests, make sure to add the canonical tag in the header of the redirect page, with the value set to the original page.

      ex: <link rel="canonical" href="https://yourwebsite.com/original-page/" />

      This tells Google that original page and the current page are the same, and in its search results, it should display the original page over the variant page. You can read this help page for more details about canonical URL.

      0

      0
      1. Thanks Victor! We have applied the canonical as well. The dev team is also confused why we would manually apply a 302 redirect to the variant page as theoretically that would make it not visible to anyone.

        0

        0
        1. Hmm, this is strange indeed. Do you use Google Optimize? If so, when you create a redirect test you do not need to set up any 302 redirects as Google Optimize redirects users by itself.

          0

          0
          1. Yep, we are indeed using Google optimize, although it is not applying the 302 – instead both original and variant are 200 status codes.

            0

            0

Leave a Reply

Thanks for choosing to leave a comment. Please keep in mind that comments are moderated and your email address will NOT be published. Please Do NOT use keywords in the name field. Let's have a personal and meaningful conversation.