Home > Tracking Campaigns > Other Systems (CPVLab, iMobiTrax, P202, Track Revenue, Click, Google Analytics, etc.)

---

Segments vs Filters in Google Analytics (3)

---

Filters in Google Analytics operate differently from filters in general, where generally it's a temporary and partial view of your data. Below is an outline of some of the *key* differences:

1. Time to take effect:
   
   • Filters: Filters take up to 24 hours to take effect.
   • Segments: Segments take effect immediately.
   
   
2. Affected data:
   
   • Filters: Filters affect only the data from when the filter is applied, it does not apply retroactively.
   • Segments: Segments affect all of the data.
   
   
3. Effect on view:
   
   • Filters: Filters affect the view permanently, so a bad filter will lose you data forever.
   • Segments: Segments affect the view temporarily.
   
   
4. Scope:
   
   • Filters: Filters are applied on the view level.
   • Segments: Segments are applied on the session, user, or hit-level, more on that here.
   
   
5. Data sampling:
   
   • Filters: Data sampling doesn't apply to filters.
   • Segments: Data sampling is applied if, according to Google:
     Ad-hoc queries of your data are subject to the following general thresholds for sampling: 500k sessions at the property level for the date range you are using
     Essentially once you meet the condition, the data will be sampled and Google will display a certain % of the data, which will result in less accuracy. How much less depends on the sampling and data set size. While it can be only a minor drop, it is avoidable, so might as well set it up to avoid sampling.

The ideal setup would be to have a view which collects all data, then setup additional views that use filters for higher level or longer term segmentation, and segmentation for shorter term / ad-hoc analyses. A common usage of filters for example is splitting converted users off from the population. For any kind of analysis where you'll likely need to view the population as a whole, go with filters. It doesn't hurt to create a new view early on in the process.

---

How are these best used for AM?

---

Filters would work better for situations such as separating out types of traffic sources into different views from the start.

Segments work well for more in-depth campaign analyses, like page funnels, and can be used to quickly hone in on specific cohorts. In particular, Advanced Segmentation allows you to specify multiple conditions simultaneously or to segment users based on a sequence of events.
Attachment 16261

In addition, there's a shared gallery of analytics solutions, including segments, which apply to many areas, for example, eCommerce:
Google Analytics Solutions Gallery

---

Home > Tracking Campaigns > Other Systems (CPVLab, iMobiTrax, P202, Track Revenue, Click, Google Analytics, etc.)