This content has been marked as final. Show 3 replies
A feedback form or link is always a good idea as you've mentioned. Unfortunately users don't always use them. So, a "passive" traffic analysis of your site can be helpful. There are several approaches. Some free. Some not. Here are some ideas.
All web servers have log files that keep a record of each file that is "served". Most web administrators have applications that can analyze these log files and produce reports that will tell you how often a particular topic (URL) was delivered. This would be the least expensive way to obtain information. It does not require the user to do anything. The info is collected anonymously. The author just wants to know if and how often the topic is accessed. Though limited, this is better than nothing and it's almost free.
Next would be to evaluate an upgrade to RoboHelp Pro and its RoboEngine server application. It employs a natural language search engine and whatever text a user puts in the search field is captured verbatim in a database. Reports (like the ones I mentioned earlier, but more sophisticated) can be generated to give the author feedback on not only which topics are being accessed, but which search queries are not being answered. That way you can address it in your next help version. RoboEngine is installed on the web server, not the author's computer. Read more about it here.
Hope this helps.
Are your recommendations only applicable with a web-help system? Since I am developing in HTML help, and thus my help system resides on an end-users machine as opposed to a web server, what are my options for measurement?
I'm afraid what I outlined it is a browser-based webhelp only solution. If HTML Help is used, there's no practical way to programatically report info back to a database because the .chm is on a local hard drive. Practically speaking you are limited to the more traditional feedback form or email link approach.