Understanding Bounce rate

[AshwinC5I]

Hi could someone help me decode this..

After redesigning my webpage, I noticed a significant drop in the bounce rate. Additionally, I incorporated an autoplay video onto the page. Could the reduction in bounce rate be attributed to the autoplay video?

[bjoern__koth]

Hi @AshwinC5I 

if your video is tracking any event, Analytics will not consider you as bounce anymore.

In Google Universal Analytics (I don't think it's possible in GA4 anymore), you could flag a tracking requests as "non interaction hit", which makes perfectly sense and did not influence the bounce rate.

Adobe Analytics unfortunately does not support this, but has a workaround: Single Page Visits

Essentially, a session with just one page view that can have user interactions.

Obviously, this is a bit of a compromise but will likely solve your issue. Alternatively, think about not having your video on autoplay or disable tracking on it.

Cheers from Switzerland
Björn

[Jennifer_Dungan · Accepted solution]

Yes, I agree with @bjoern__koth 

Adobe's default definition of a "Bounce" is a single hit not a single page, and the definition of "Bounce Rate" is bounces / entries

For our purposes, we don't like this definition... particularly when there are other actions that can occur on the page (videos, overlay impressions, cookie consent, etc)

So for our needs, we actually created custom Bounce (i.e. Single Page Visits), and a custom Bounce Rate (i.e. Single Page Visits / Entries)

This aligns closer to what is globally considered a bounce or bounce rate...  based on single page, not on single hit...  

[DeniseSp]

Do you know how I can define a bounce rate in CJA? Specifically, if I want the definition of a bounce to be similar to that of GA4:

A session is considered a bounce if it doesn't meet any of the following criteria: 

• The user stayed on the page for less than 10 seconds 
• The user didn't trigger an event (link clicks, video views...) 
• The user didn't move on to another page or screen. 
•  

[Damonwhall]

Agree with both @bjoern__koth  and @Jennifer_Dungan , 

Additionally, if you have something like iFrames that will also add additional complexity to the bounce metric.  

Using just a page view based metric for bounce addresses most of this.  I have also added further adjustment in my own metric that removes 1 second or less visitors to get past things like bots.  This will provide more of an indication of the page value (as it is unlikely that someone that was on the page 1 or less seconds was human, or intentional at all in their visit.  )