AdWords is now Google Ads. Our new name reflects the full range of advertising options we offer across Search, Display, YouTube, and more. Learn more

3.5K members online now
3.5K members online now
Discover how to select who is eligible to be in your experiment, and when they’re served experiment variations.
Guide Me

Too few sessions on experiment

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭



I am running an a/b experiment but the sessions in the experiment are way less than expected (less than 10% of total sessions) taking into account the target selected. The only thing I suspect that could be working wrong in some cases is that I am using a cookie for targeting, and maybe the cookie is not set at the moment Google Optimize runs.


The website is:


Thanks for your help.

1 Expert replyverified_user

Too few sessions on experiment

Google Employee
# 2
Google Employee

Hi Martin


One mistake that I can see in your page is that you are adding the async-hide snippet through GTM.


This will not help much and causes a "blinking effect" on your page. The async-hide snippet is there to hide the page, until GTM and then Optimize get loaded.


Since GTM is loaded asynchronously, users may have already seen your page anyway so you will not avoid the flicker.


You need to either add the async-hide snippet inline in your page (if you plan loading Optimize via GTM then use your GTM and not the Optimize container ID as its parameter), or if you are not worried about flicker then just not use it at all.



On the main question about the cookie: I don't think that you have a problem there - as far as I can see, the cookie is set on the initial response. In fact parameters are not used in the path rule, so one of the values on your path rule is ignored (but I don't think it makes any harm).


I believe things you just work using the cookie rule and just the path "/" rule, so I couldn't spot anything else going wrong.


You may do the following to troubleshoot though: Use the preview functionality (you can do it on a started experiment) and change your url and cookies in that tab to double check that you see the previewed variant.


You should see it as long as you stay in that tab on any URL and page reload where targeting rules match. You can use the chrome dev tools to clear up that specific cookie that you are using.


The most likely reason for the issue, is that with the way that you have setup your GTM container, the analytics hit is send before Optimize runs and doesn't include the experiment information, which is send later as a separate hit.

See this article about ways to ensure correct order of execution (look at tag sequencing):


If you want to load Optimize via GTM, then the native "Google Optimize" tag would probably be simpler. Consider moving the GTM installation to happen quicker to minimize flicker (if you face that flicker problems). At the moment its loaded by a script inside the body which is quite late so flicker will be probably visible. 


Keep in mind that Optimize unlike other tags, is actually affecting your own page, and this is why its recommended to install as early as possible in the page (i.e. inline in the head) see: