Database Row Limits
Google Analytics uses database tables to store all the information it gathers for reports. Each table has a row limit. At last count, the limit was 50,000 rows. This means that each table can hold up to 50,000 unique entries.
This is how it works
1. A URL is visited and data for the pageview is received (eg. which URL was viewed, how long was spent on the page, etc.)
2. Google Analytics checks to see whether a row for this URL has already been entered in the database
This process continues happily until there are 50,000 entries in the table. At that point, a new row is created, titled “(other)”. Now if a new URL is encountered, the data for that pageview gets lumped into the (other) row.
This does not mean that only 50,000 pageviews or visits can be tracked, because there is no limit on the number of visits that can be recorded for a single page. It does mean that only 50,000 unique URLs can be tracked.
What data is affected by database row limits?
There is a separate table for every dimension. For example, there can be up to 50,000 URLs and also 50,000 page titles. Also affected by this row limit are dimensions like browser, operating system, search keywords, e-commerce products, geo-locations, etc.
How often are new database tables created?
New database tables are created each day. When new tables are created, the row limits are reset.
How do I get rid of (other)?
The table row limits are typically only exceeded in content reports. The cause is usually a combination of unnecessary query parameters and case sensitivity. Each of these URLs is seen as unique:
Get past this by creating a filter to set all URLs to lowercase and by specifying query parameters that Google Analytics should exclude.
Is there any way to see what pages are contained in (other)?
No. This shows up as a result of storage limitations. It is enabled on the back-end, not by the reports themselves. There is no way to tell what URLs contributed to the (other) totals.
Will advanced segments break out the (other) URLs?
No. Advanced segments repackage existing information. They do not reprocess old data. There is no way to segment the different URLs that may have contributed to the (other) row.