How to ignore a custom link server call from Bounce Rate calculation like Google Analytics does?

[Aiyanna_Kuttapp]

As per Adobe Analytics documentation, Bounces are defined as :

"A visit that consists of a single server call. For example, a single page visit is a bounce if a visitor does not interact with the page in a way that sends data to Adobe, such as clicking a link or a video start. If more than a single hit is received in a visit, a Bounce is not counted."

Suppose I have a custom link server call which triggers immediately after the pageload call when I enter my site . Adobe Analytics will document this visit as a non bounced visit, since there were multiple server calls. 

Is there any configuration/coding technique which would make Adobe Analytics to conclude such a visit (with only a pageload call and one custom link server call in the session) as a Bounced Visit? Basically can we disable the custom link from Bounce Rate calculation?

I know we can do this in Google Analytics where in we can tag events with 'opt_noninteraction' to make it not part of the Bounce rate calculation

_trackEvent(category, action, opt_label, opt_value, opt_noninteraction)

https://developers.google.com/analytics/devguides/collection/gajs/eventTrackerGuide

Thanks,

Aiyanna Kuttappa

[brian_au · Accepted solution]

Hi Aiyanna,

You can potentially leverage the 'Single Access Metric' which does not account for custom link calls to calculate bounces. Extensive details on the differences between different bounce measures is linked below:

https://marketing.adobe.com/resources/help/kb/en_US/analytics/kb/comparing-bounces-and-single-access...

>>>

This article provides a comparison between bounces, single access, and single page visits, although their names and perceived definitions may appear similar, their actual functions are very different.

• Bounces are defined as the number of visits to your site that contain a single image request. If a user comes to your site and fires more than a single image request, even a page refresh, they aren't counted as a bounce. If "count page reloads" is enabled in Ad Hoc Analysis, image requests flagged as reloads are omitted. Visits consisting of one image request + reloads is considered a bounce in this specific case.
• Single Access
  • In Reports & Analytics, single access is defined as the number of visits to your site that contained a single unique variable value (most commonly Page Name). Provided the variable value in question does not change, they can fire many image requests and still count toward Single Access.
  • In Ad Hoc Analysis, single access is defined as the number of visits that consisted of a single page view (and possibly one or more custom link image requests). 
• Single Page Visits is a metric used specifically in ad hoc analysis, though there is also a report with a similar name in reports & analytics. If you have ad hoc reporting set to count page reloads, it reports closer to Bounces. If you have set ad hoc reporting to not count page reloads, it reports closer to Single Access.

There are two types of Bounce Rate metrics available:

• Bounce Rate available as a standard metric is the percentage of users who didn't make it past that single image request on your site.
• Bounce Rate applied as a calculated metric is defined by the end user. The most common bounce rate calculated metric is Single Access divided by entries.

Best,

Brian

[nibedita91]

HI Briyan,

[nibedita91]

HI Briyan, The link which you have shared is not accessible. Could you please help. Thanks!

[Gigazelle]

I second Brian's suggestion to use the Single Access metric. Based on your description, single access is exactly what you're looking for.

[LAE]

This works for "bounce rate", but not if you are looking at bounce rate for other associated things, like marketing channel or eVars. If the non-interactive event fires BEFORE the pageview -- which happens to us about 20% of the time -- the Entry is lost to the page. That causes Entries to not be counted, and can lead to "bounce rates" (single access rates) of greater than 100%.

In 2019, we still need a non-interactive event.

[frederiquer4521]

Hi LAE. may i ask how you figured this out? super interested as we are seeing same issue of BR based on SPV greater than 100%

[LAE]

Hi @frederiquer4521, I looked at the hit-level data from our Adobe data feed. We had the issue with bounce rate relative marketing channel, but I discovered the issue applied to eVars as well.

Essentially, pre-pageview secondary server calls (non-pageview event server calls that fired before the initial pageview server call) did not include important post_eVar information needed to associate eVars with Entries. If there were not a high number of visits to begin with, the discrepancy was amplified.

These non-pageview server calls were related to a number of things and we did some work to delay those, though we could not control the Optimizely server call related to testing. As a reult, we opted to push that into Google Analytics instead, which removed those problem Optimizely server calls.