The Cookie Consent Dilemma: Unpacking Bounce Rate Changes in Adobe Analytics

[DhanA2]

Hello fellow Adobe Analytics enthusiasts,

I recently implemented a cookie consent banner on our website to ensure GDPR compliance and to respect user privacy. However, I've noticed some fluctuations in our bounce rate metrics since the implementation. I'm curious if others have experienced similar changes or if you have any insights into why this might be happening. Here are a few questions I have:

1. Did you notice any changes in bounce rate after implementing a cookie consent banner? I'd like to know if this is a common occurrence or if it's specific to my setup.

2. What are the potential reasons for changes in bounce rate? Could it be related to how users interact with the cookie consent banner or any technical issues associated with its implementation?

3. Are there any best practices or recommendations for tracking bounce rate accurately when using cookie consent banners? I want to make sure I'm measuring bounce rate correctly in this new environment.

4. How can we mitigate any negative impact on bounce rate while still complying with data privacy regulations? Balancing user consent and accurate analytics is crucial, and I'd love to hear your thoughts on achieving this balance.

5. Have you tried any specific Adobe Analytics configurations or customizations to address this issue? I'm open to suggestions on how to fine-tune my Adobe Analytics setup to account for the changes.

Your experiences and insights will be greatly appreciated. I'm sure this is a topic of interest for many of us, given the increasing importance of data privacy and the impact it can have on our analytics.

Thanks in advance for your help!
CC: @Jennifer_Dungan  @aanchal-sikka @yuhuisg  @Mahedi_Sabuj  @Harwinder-singh  

Best regards,
Arun S

[Jennifer_Dungan]

We have not implemented a cookie banner yet, but here is my thoughts:

Adobe calculates "Bounce Rate" by looking at "Single Server Calls within a Visit"

I don't actually like Adobe's default Bounce Rate, I created my own calculated metric for "Single Page Views within a Visit" (as in there could be multiple server calls, but only one actual page)

I am not sure how you have set up your consent, but I suspect that you are either tracking the first page as a page view (prior to consent) or maybe you are tracking the Cookie consent banner as an action (then after consent is given tracking the page or not if consent is rejected).

Now, if you are doing the first.. tracking that first page, there's not much that you can do about Bounce Rate since even a custom "Bounce Rate" metric would still be triggered as a single page view. But you could exclude visits where the consent banner is triggered, so that those visits where users reject consent, are not included in your calculations? (This might be a good starting point, at least for Bounce Rate)

However, if you do an action call on the banner, then do a page view only after consent is granted, then creating your own "Page specific bounce rate" could help.

[Damonwhall · Accepted solution]

Hello @DhanA2 , 

We have had to get creative with "Bounce Rate" as well.  I agree with @Jennifer_Dungan  that the out of the box bounce rate is not really helpful.  This is for several reasons: 

• If you have iFramed information on your site
• If you have an SPA site or parts of your site
• If you are using something like Path Factory that can cause an iFrame-like experience.  

Some things to consider are if your "banner" is firing a call that is set as "page view" instead of "link-clicks". this will automatically reduct your bounce to almost zero.  So, you can consider altering that.  Also, if your banner IS on its own page, you can make rule exclusions in Launch to manually address this issue.  

Alternatively, I have generally found that the idea of "Bounce Rate" (number of 1 page view sessions divided by sessions) is not too helpful IF the goal is to understand "targeted/valuable audience ratio".  I Instead have found that there are plenty of visits that are 1 page view, but are still relevant.  Example: someone hits a page, is on the page for 3 minutes reading and engaging, and then when they are done, they leave.  In this case, the visit had meaning, and then they left.  Doesn't mean they should be excluded as a "bounce".  I created an alternative metric I call "Calc Bounce Rate" (calculated bounce).  that looks like this: 

And this allows me to understand the number of "quick leavers" on top of 1 page view visitors.  This tended to solve the issue of iFrames and such on our site (although admittedly, it is not the traditional Bounce Rate metric.)