Filtering traffic on your Help servers

So you've moved your help content onto a server. Wonderful! Server-based content offers a lot of advantages, including the ability to analyze how users are accessing that content. Such feedback allows you to optimize your help topics accordingly.

However, how can you tell if those help visits are from internal testing or actual users?

One way is to filter the traffic by IP address. Much like your street address or the URL for your website, an IP address is a unique identifier for the machine accessing your help servers. The IP address of each visitor is recorded in the .LOG files on your servers, and can be viewed in your traffic analysis tool.

To filter by IP address, set your analysis tool so that it excludes all IP addresses assigned to your company. Usually, a company will have a range of IPs assigned to it. By filtering out these IPs, any visits from machines inside your company will be excluded from reports of traffic data.

If you are unable to filter IP addresses, you may be able to use cookies instead. A cookie is a small bit of code that gets saved on a PC when it visits a website. Create the cookie so that it never expires. Then, have all internal users visit the page with the cookie. After the cookie is set on their machines, set your analysis tool so that it filters out all visits that include the cookie.

Both of these methods will allow you to sort out whether traffic on your Help servers is coming from internal testing or actual end-users. If you need technical assistance, try contacting the team that manages your servers. They should be able to help you evaluate which solution is best and assist with implementation.

When you deliver help topics on a server, knowing where your traffic is coming from is very important. Internal testing behavior will differ greatly from how actual users navigate through your help topics. You don't want to make improvements based on data from internal testing.

If your help is not delivered via a server, you might want to consider whether such a move would be worthwhile. Server-based help allows you to monitor how your help is used, update content without assistance from developers, and more.

Related: Can help authors afford to ignore Google?