Getting Feedback on your Technical Documentation

A question that is often asked of Technical Writers is, "How useful is your documentation?". All too often the question results in a shrug of the shoulders and an "I don't know" answer. That's a shame, as it is only by understanding how our users use our content, and the feedback we get from that, that we can truly answer that question. The problem is, how can you get that feedback.

There's really no substitute than a face to face conversation with your users. Admittedly it is time consuming and expensive, but it can lead to a wealth of valuable information. Your skill at asking a Product Manager questions that unearth useful nuggets about functionality in a future release, can just as easily be applied to asking open and probing questions to users about their use of your "product".

The problem is that most Technical Communication departments are "protected" from meeting customers by a combination of budget, organizational set-up, and lack of a cohesive content strategy. This makes the chances of most Technical Writers speaking to their users near impossible. So what other options are there?

If you can't get out to your users, and you must never give up trying, the answer can come through the use of technology. Both the main players in the technical communication tool space offer solutions that provide statistics on your content's use.
  • Adobe RoboHelp Server has been around for years, but has recently undergone a major rewrite. RoboHelp Server requires a Java runtime environment (e.g. JRE or Apache Tomcat) and a database (e.g. Oracle or SQL Server) but the installation and setup is very easy. When I tried, it took me little more than 30 minutes See the Adobe RoboHelp Server System Requirements page for full details.
  • Madcap Central is the newer kid on the block and offers similar functionality. Unlike Adobe's offering, Central doesn't rely on external software or databases. You just need a browser and an internet connection. Instead, it has partnered with Microsoft Azure to host your content. See the Madcap Central System Requirements page for full details.
Both of these offering have very similar functionality. Their series of reports display statistical data of the number of page visits, search terms used, search terms used that returned no results, context sensitive help calls, etc. All very useful, but the offerings aren't the golden bullet you may expect. Here's why:
  • They only work with specific content. Adobe RoboHelp Server only works with Responsive HTML output. Madcap Central also has limitations with the Flare targets that can be published to Central, but also offers content not being authored in Flare to be hosted.
  • The reports offered by both solutions are fairly basic. For example, they don't allow much in the way of customization or filtering.
  • Statistics on their own provide only part of the story. It requires time and effort to analyze the results and action change. For example, repeated context sensitive help calls from a dialog could mean there's a need for a UI or copy change.
There are other analytics offerings. Google Analytics offers a wide range of statistical information, although it isn't tailored specifically to the technical communication industry. It requires you to add a tracking code to your content, and depending on the content size and usage, may require you to pay Google for the privilege of using it.

The takeaway is that getting feedback isn't easy, and even when you get it, it needs additional time and effort to use the data effectively. It may require buy in from other teams. For example, if the UI needs changing. Get the process right, and you've a great resource at your disposal.

Let me end with a tale from my past. 

I once had a manager who just wanted to know how many hits the documentation got, and the more the merrier. He used these statistics to justify our existence. He wasn't interested in the analysis behind the statistics. I could have spent all day pressing the F5 key to boost the stats, and he'd have been happy. The end result of this, is that I didn't stay long at that gig!