Installation View

We have updated the Installation view with a new visualization of the user loyalty. The new visualization features the loyalty plotted over time to allow you to better understand how your user base changes over time. You will still be able to see the overall distribution for a selected time period but will also be able to see the relative distribution between the different loyalty groupings as if changes during the time period. The existing visualization of new users on the product is maintained but the underlying algorithm for determining new users have been fine tuned. The new Installation view is illustrated in the screenshot below.

You can also access the average runtime for each of the loyalty group by selecting the View average runtimes item in the right hand menu of the Loyalty chart. Selecting this item will display a window with the individual average runtimes as well as a total average runtime. This is illustrated in the screenshot below.

You can still access the list of identifiers by selecting the Get Identifiers item in the menu and select the desired identifier. The identifiers are extracted in the background and an email with the results are sent as soon as they are ready.

Feature Value and Feature Timing View

We have also updated the way you can view and interact with your feature timing and feature value tracking. Just as with the installation view described above, we have moved away from a static snapshot view of the data and over to a visualization focusing on the evolution of the data, the change over time. This is facilitated by dividing the values and timing data into logical and semantic groups of data and visualizing how these distribution of these groups change as time passes. Seeing these groups of data and their distribution change over time greatly improves how data can be interpreted and acted upon. The visualization of data is exemplified in the screenshot below.

From looking at the screenshot above you can see both the overall distribution of data between the groups in the full period as well as how the distribution between the groups evolve over time. Furthermore, the sparkline chart shows the volumes of data collected over time and below the piechart you can see some aggregated data for the period selected. Overall you should be able to easier understand and work with this visualization of data.

You can define the semantic groupings of data by selecting the Setup intervals menu item which brings up a dedicated dialog. The dialog allows you to enter specific limits to setup intervals and you can use the sample data to quickly select one of the predefined percentile distributions to get a good starting point. You could enter your limits to track specific service level agreements you may have or you could setup the intervals to simply get a feeling for the experience of e.g. the 95th and 90th percentile of your users. The setup dialog is shown in the screenshot below.

You can change the semantics groups at any time to changing requirements or user behavior.

Improvements and fixes

We have awesome customers that continue to provide us with feedback, suggestions and issues that needs to be fixed. As always we have tried to address as many of these as possible, some of which are mentioned below:

• More views can now be added to the workbench, including the new feature timing and feature value views
• More views can be queried with Advanced Queries for more in-depth looks at your data
• The list of countries and releases in the filters are now limited to the releases actually received as part of public data, filtering away those releases you have only seen as internal data
• Improvements to performance on the log message page and the session view page
• Lots of minor issues