There is an A/B/n test in which there are 3 experiences - Control, Var1, Var2. As I create its A4T reporting, I am using the dimension Target activities and then I broke it down with Target Activity > Experiences. Now for further analysis I tried breaking down the Target Activity > Experiences dimension with Page URL dimension and what I saw was, there are many extra URLS other than the actual activity URL. If I have created the activity with only 1 page URL, let's say - Page A, then I should just see the URL of Page A in my A4T reporting. Why do I see the other page URLS in the reporting?
Solved! Go to Solution.
hey @nogupta1791, great question. When you consider what Analytics is capturing (and able to report on) it goes beyond just the page where the experience was delivered. This is a very good thing! Consider if you run a test on one page but want to see how that impacted behavior on another page further down the funnel, you can do that with A4T. This allows you to go much deeper in deriving insights and interpreting success for your test. In your example, the "other" URLs that are showing up in your breakdown are the pages that visitors ALSO saw who were part of your Control, Var1, Var2 experiences. I would look to see if there are major differences in those pages to see if your experiences impacted "what else" the user may have done because of the experiences delivered.
hey @nogupta1791, great question. When you consider what Analytics is capturing (and able to report on) it goes beyond just the page where the experience was delivered. This is a very good thing! Consider if you run a test on one page but want to see how that impacted behavior on another page further down the funnel, you can do that with A4T. This allows you to go much deeper in deriving insights and interpreting success for your test. In your example, the "other" URLs that are showing up in your breakdown are the pages that visitors ALSO saw who were part of your Control, Var1, Var2 experiences. I would look to see if there are major differences in those pages to see if your experiences impacted "what else" the user may have done because of the experiences delivered.
Thanks for your reply @ryan_pizzuto . Below is the screenshot of the reporting that I created. The activity URL is yellow highlighted. The 'other' URLs are orange highlighted. So, as per your above reply, there are users who visited the activity URL, saw one of the experiences then they might have browsed to one of the 'other' URLs which is why Analytics captured those 'other' URLs. Is there any documentation around it?
Views
Replies
Total Likes
That's right. Analytics is capturing all the activity that a user has done within their sessions. For some primers on A4T I would suggest the following links:
Views
Replies
Total Likes
Hi @ryan_pizzuto,
Thank you for the information and validating my thinking about additional URLs in the report. I have a follow up question.
When you use A4T, the report defaults to 'Activity conversions' based on the Goal I've set up for the test (e.g., click on the button by selecting/highlighting a link/button). If I don't break down the Exp by Canonical URL, I think it gives a misleading information because the total number of Activity conversions includes other conversions not only on the activity URL but also other pages that users view within that Experience. So it is important to actually break it down by the URL to see the true conversion on the activity URL. And to your point, still review other URLs in the report in case there is a valuable insight.
There is also an option to update the default Activity Conversion to 'Same Touch' attribution model that gives 100% credit to the very hit where the conversion occurred. When I do that, Activity conversions on other URLs in the Experience turn into 0. But in addition to that, whatever Activity conversions that exist get allocated to 'Unspecified' and not to the Activity URL. Plus, the conversions number differs from the number when using the default 'Activity Conversion'.
Do you have any insights into 'Why'? And which model and reporting I should be relying on?
Thank you!
Leyla
Views
Replies
Total Likes
Views
Likes
Replies
Views
Likes
Replies