Configuring Visitor Identification

There are a number of settings that influence how Logaholic will count visitors and visits.

If you change any of the visitor identification settings, the changes only affect new traffic analyzed by Logaholic, not historical data.



Visit Timeout:

This is a setting in minutes. It determines how many minutes a visitor can be inactive before the next request from that visitor will be counted as a new session/visit.


Unique Visitor Identification Method:

This setting has a big influence on how your visitors are counted. Choose the method that is most appropriate for your situation.

Logaholic supports a number of ways to determine what a unique visitor is:

IP Address only

This method uses the user’s IP address as the ID to identify the visitor. For sites with little traffic, this is the recommend setting.

When using this method, 2 different users/browsers/devices coming from the same IP address will be counted as 1 unique visitor.

However, on low traffic sites there is little overlap in IP addresses (multiple users using the same IP address to visit your site), which makes this a good method, because usually the visitor connecting with different browsers or devices will actually be the same person (or company) and it makes sense to count those as one entity for accurately measuring things like conversion rates.

IP Address and User Agent

This method combines the user’s IP address and full browser user agent string in a unique hash, which is used as the ID to identify the visitor.

This means 2 different users coming from the same IP address will usually be counted as 2 separate unique visitors. For log file based profiles, this is the default setting in Logaholic.

Cookie based

This method uses a unique Logaholic visitor id cookie, which is set in the user’s browser by the Logaholic javascript tracking tag.

For profiles using javascript based tracking, this is the default setting.

If you are using log file based profile, you can still choose the Cookie based method, but you’ll need to take some additional steps to make it work (we call this “Hybrid Mode”):

  1. Copy the javascript snippet that appears when you select the ‘Cookie based’ option from the drop down menu and paste that into the HTML of your website pages. This will make sure the Visitor ID cookie is set in the users browser.
  2. Change your log file format so it also includes the user’s cookie information
For more information on how to set up your logs to include cookie data, please read Configuring Log File Analysis Hybrid Mode.

Cookie based visitor identification will generally be more accurate than IP/Useragent, because a visitor can still be identified even if the IP address changes. However, users may also delete or block cookies. If a cookie is not found for a request, the method falls back to IP/Useragent

Custom Cookie

Logaholic also supports identifying users based on your own custom cookie. If you are already setting an identifying cookie (for example, for logged in users) you can tell logaholic to use that cookie to identify unique visitors. If the cookie is not present, the IP/Useragent method will be used.

In tis case, we recommend using a human readable ID, like an email address as the custom cookie as this will make it easier to identify individual users in several Logaholic reports, like Click Trails etc.

This method is only supported in log file based profiles using log files that contain the cookie information. See Configuring Log File Analysis Hybrid Mode for instructions on how to include cookie info in your logs.


Human/Bot Detection:

In Logaholic, visitors that are marked as “bots” are not included in most reports. The traffic from bots is stored separately but is available in reports via the advanced settings button in the report options.

Logaholic has two ways to detect whether a visitor is human or a bot (a script, program, crawler or other automated process):

  1. User agent based. This method uses a library to analyze the user agent string to see if it can be identified as a bot. This will catch most “honest” bots, like GoogleBot and other crawlers by legitimate services.

  2. Behaviour based. This method uses a broad algorithm that will initially identify all users as bots, unless “human like” behaviour is detected. This method is the default setting for log file based profiles as of Logaholic version 7.

How behaviour based bot detection works and when you should use it

The vast majority (99%+) of bots do not request images or javascript files when they visit a website. On the other hand, the vast majority (99%+) of web pages will embed at least one image or javascript file.

Although our algorithm may be refined in future releases, it essentially works by identifying a visitor as human if there has been a request to a piece of content (html/php whatever) and at least one image or javascript request during the course of the session.

Behaviour based bot detection only makes sense when using a log file based profile, so it’s not available for javascript based profiles.

You should use this setting if you want a more realistic view of your actual visitor numbers.
If your site contains no images (or very very few) and no javascript files, i.e HTML only, do not use this setting.

So, depending on how your site or app is put together, some traffic may be wrongly identified as being a bot. Use the “All Traffic” switch in the Advanced Report Options section to show the numbers across all your traffic. If there are too many false positives, we recommend changing this setting to “User Agent based”.