Application Analytics – Integrating Feature Usage Tracking in a WPF application

It’s important to note that there are plenty of analytics monitors available in multiple technologies. In our case, because we’ve implemented a WPF application we’ve selected the .NET monitor for our application. The current downloadable monitors at the time of this writing are as follows:



Now that we’ve registered our application you are presented with a product key that is unique to your application. (Please note that the key in the screenshot below is fictional)

In the App constructor, assign the generated analyticsMonitor variable to this new static property:

 

In the Get Started pane of your dashboard, directly below your product key, you can see the step-by step instructions to enabling analytics in your registered application.


The first step as indicated in the screenshot above is to download the analytics monitor for .NET. Once this is complete, unzip the contents of the file to a location of your choosing on your hard drive. Right-click on your project file and browse to add a reference to EQATEC.Analytics.Monitor.dll that you just downloaded.

Open App.xaml.cs and replace the code listing with the following (Note: change the namespace to the base namespace of your application):

In the constructor we created an instance of the analytics monitor using the AnalyticsMonitorFactory, this value is then assigned to our static Monitor property so that the monitor can be accessed from anywhere in the application. We also overrid the OnStartup and OnExit methods to start and stop the monitor.

To walk through this example, you will need to download the original source code to the WPF RSS Reader. This WPF client uses the RadControls for WPF Application template and controls, therefore if you don’t already have it installed, please download and install a trial.

The application we are going to instrument is a simple one, it reads the RSS feeds of the user’s favorite blogs and then provides the ability to dive into a specific article for viewing within the application. The favorite feeds are stored in an XML data file that is maintained through an administration screen.

In order to gain a better understanding of the behaviors of the users of our application we will add feature tracking to a few key areas of the application.  

As a first step, we want to see if the users are actually using the menu system to switch between blog feeds. Since the application is delivered to the users with a couple of the most common feeds already in the data file, we are curious to see if they are actually using the application to read more than the first blog feed. The currently visible blog feed is switched in the application through the ChangeCurrentFeedCommand command. Open Commands\ChangeCurrentFeedCommand.cs file and append the following line of code to the Execute method:


Next, we would like to see if users are actually drilling into individual articles to read. This functionality is implemented through the ChangeCurrentArticleCommand. Open Commands\ChangeCurrentFeedCommand.cs and add the following line of code to the Execute method of the class:


Another aspect of the application that we’d like to track would be the administration of the feeds. We can track the accessing of the screen through the Commands\GoToFeedAdminCommand.cs file by adding the following line of code to the Execute method:


Finally, we can benefit from seeing how the users are actually using the administration screen. This screen is implemented with the RadDataForm control with custom commands that are wired up through a DataForm command provider. Our implementation of this command provider can be found in Commands\FeedAdminCommandProvider.cs. Let’s track saving of records by adding this code to the CommitEdit method of the class:


Then append the following line of code in the Delete method of the class:


It will be interesting to see the comparison between how many people access the administration screen versus actually using it.
It’s now time to save the application code, run it and interact with it so that we can see some of the collected data within our dashboard.

VIEWING LIVE DATA