Examining the Landing Page is (not set) Issue
In Google Analytics, when you have landing page is (not set) showing in your reports you have a problem with session integrity. Session integrity refers to GA’s ability to accurately represent the visits that occurred on a website. When issues like landing page is (not set) occur, it causes many sessions to be created, thus corrupts the integrity of the session metric. This article will review how to identify issues that negatively impact session integrity.
We’ve got a problem
We’re going to examine a scenario in which the “(not set)” value is showing up in your Landing Page reports. This is an impossible condition that is somehow appearing in your data! How is it impossible? The landing page should be the first web page visited during a session. For the landing page to be “(not set)” would mean that a session was created that didn’t have a page, or in this case the first hit wasn’t a page. This is what the error looks like in GA:
Behavior > Site Content > Landing Page
Thinking back to Joe, we see that he has come to our website three times in a period of two days. However, the Tuesday sessions look suspicious. We see that the second session has a Landing Page (not set) issue, and the first session looks like a bounce; what a nightmare!
To make matters worse, it turns out the marketing team ran a big ad campaign on Tuesday and now they’re asking you how many sessions had transactions as a result of their campaign spend. This would normally be answered by looking at the visits that had both the campaign name and transaction in the same session.
The problem is that the session tied to the campaign spend looks like a bounce!
To make matters worse, we have a junk session that contains a transaction with no landing page. This adds complexity and inaccuracy to our reporting.
Why would a session break? This appears to be the best explanation for what happened to Joe and his transaction. It isn’t logical that he made a transaction without a page view first, and it makes sense that the bounce session might have been corrupted.
When a session is created, the subsequent hits need to have the same client ID. Also, the campaign name, source and medium need to remain consistent. Knowing this, we can examine a few scenarios where new sessions are created automatically by GA:
- Session Settings: There is a setting in the Admin > Property > JS Tracking screen that allows you to set a default timeout period for your GA sessions. This defines the length of time required before GA terminates a session due to inactivity.
- Midnight: At midnight in the timezone the view is configured, GA will terminate sessions and assign new IDs. Google does this to keep the data partitioned by date. It would be wise to investigate if this is an issue on your site. Conversions by hour of day is a great report to run to understand the scope of this problem on your site!
- Attribution: Campaign Name, Source, Medium or Client ID changes suddenly during session. The inbound attribution information is preserved while the previous session and new session is created using the new attribution information. This is almost always a mistake and can wreak havoc on session integrity. If client ID changes, that’s a huge problem that affects user and session scopes.
Thinking again about Joe and his bounced visit, what if it wasn’t a bounce at all? What if, instead, right before making his purchase, Joe simply stepped away from his computer; maybe looking for his wallet!
If that break took more than the default timeout of 30 minutes, then GA would have counted the first session as a bounced visit. Any subsequent hit would be added to a new session. Once the break occurs there is a good chance the session metrics will be impacted negatively.
By making simple adjusts to our measurement strategy, we can account for these different scenarios and correct them. To solve this particular issue, it could be as simple as increasing the Session Timeout window from 30 minutes to perhaps a more forgiving 45 minutes, or an hour.
The business has to think about the rules as they are defined in GA and how that impacts the measurement. We know we want to be able to attribute the dollars to the correct sources and campaigns, therefore it’s important to address the issues that lead to fragmentation.
Where to Start?
You first need to identify the problem. By auditing your GA reports, and reviewing the measurement code implementation you’ll be able to single out and prioritize critical issues. If you need help with this then please reach out to us! We are your Analytics professionals, so put us to work!
Once you identify the problem, assigning the priority is important. We can use the scopes we reviewed in a previous article to help assign priority. The user scope is critical, since this can have an additive effect on metrics. Session corruption comes after that as it causes additive issues in data. Finally our hit corruption needs to be addressed to ensure we collect individual behaviors accurately. The session and user scope issues are prioritized because of the math involved with adding and averaging metrics in GA reporting.
We want to make sure that we identify issues with scope early on. We can identify issues with session integrity by looking for common errors, like when landing page is (not set). When we do the work, and ensure our clickstream is intact, and aggregations across scopes are consistent, then we’re left with incredibly actionable data!
We want your data to be as actionable as possible in Google Analytics. We’re a great resource to help you and your team along the way. Please reach out to us to learn more. Stay in touch on Twitter @AnalyticsPros and let us know about your issues with session integrity and the many ways it can negatively impact your dashboards and reports!