The following table shows different methods of excluding bots and how they compare to each other.
Method | Bot Rules | Exclude by IP Address | Customer attributes | Segmentation | 3-rd party scoring + Segmentation | Suppress Server Call for Bots at Runtime | Custom DB VISTA rule |
---|---|---|---|---|---|---|---|
Description of method for excluding data | Exclude based on user-agent, IP address, or IP address range | IP address | A flag in customer attributes that identifies ECID as a bot | Criteria in an Analytics segment that identifies known bots based on bot behavior | A 3rd party such as Perimeter X or Akamai Bot Manager assigns each page view a score on how likely it is to be a bot. Score is sent into Analytics and segments can be used to filter data out based on the score. | Client-side logic stops the Analytics server call from being executed for bots. | A VISTA rule will move traffic from bots that meet certain criteria to a separate report suite. |
Bot names are available for reporting? | Yes | No | No | No | No | No | Yes |
Able to see which pages are being visited by bots? | Yes | No | No | No | Yes | No | Yes |
Incurs server call cost for bots? | Yes | Yes | Yes | Yes | Yes | No | Yes |
Bot data available in data feeds? | No | No | Yes | Yes | Yes | No | Yes |
Able to report on bot traffic as if they are actual server calls? | No | No | Yes | Yes | Yes | No | No |
Can retroactively remove data from a data set? | No | No | Yes, once declared IDs are implemented | Yes | Yes, once scores are implemented | No | No |
Is subject to Uniques limits in criteria? | No | No | No | Yes | No | No | No |
Is subject to additional cost? | No | No | Possibly, depending on Analytics SKU | No | Yes | No | Yes - cost to implement and maintain a VISTA rule |