Doubts about Bounce Rate

[JBriceno]

Hi, I have doubts about the bounce rate and its formula. It is currently defined as Bounces (visits containing exactly one hit) divided by Entries (number of times a given dimension item is captured as the first value in a visit).

On a website I am analyzing, I see that since February of this year, we went from an average of 10% to 30% without any clear explanation (no changes in tagging, no cookie modifications, no increase in traffic), so I want to know if changes have been made in the formula, specifically in the Bounces which is the metric that is increasing.

Also, when the page loads, 2 hits are received: one click and one page view. In this case, shouldn't the bounce rate decrease, or is it only considered for page view hits?

Thanks

[bjoern__koth]

Hi @JBriceno 

that is correct, if 2 hits are received, you should by definition not have a Bounce.

Note that Adobe also has a "Single Page Visits" metric that should solve your issue (unless you explicitly send a second request to not have a Bounce).

Can you see a similar behavior if you do Single Page Visits divided by Entries?

[Nick_Walter]

There are many things that could cause this. From increase in bot traffic, to change in implementation. I am currently not aware of anything changing with how Adobe calculates this metric. Personally I like to look at engagement rate rather than bounce rate, as I think it gives more value.

[Jennifer_Dungan]

I agree with both @bjoern__koth and @Nick_Walter.

I am unaware of any changes on Adobe's side, but changes in implementation can have a huge impact... if you recently added Cookie Consent, and are tracking the acceptance and the page... all of a sudden that will be multiple hits (and no longer considered a bounce).

I have used my own custom Bounce definition using "Single Page Visit" for years...  

[jeff_bloomer]

Hi @JBriceno,
I have also never preferred to use the bounce rate that incorporates bounces with entries, as it does not properly accomodate for what your customers are doing.  More than quite a while back, I ran into this linked video and began using this kind of metric instead as a way to understand things differently.  Again, it discusses the SPV approach, and you can see the difference it makes.
https://youtu.be/cem3n0mgpEA?si=txQayajsAfLdOmcX 
However, I also agree with @Nick_Walter, that an engagement rate is a good way to begin looking at things differently, especially if you have any future plans to migrate to CJA.