Abstract: Evaluating user experience for an application includes collecting, for each of a plurality of sessions of the application, a list of user event data items experienced during that session. The list of user event data items includes any of a device utilization quantifier, user action-response time pairs, and an application fault indicator. For each session, a value is assigned to each collected user event data item, and a session score is derived based on those assigned values. A user experience score is derived based upon a plurality of the derived session scores. The user experience score is reported.

Abstract: In one example of the disclosure, incident data indicative of a crash of an application is obtained, the crash occurring during an access of the application by a user at a computing device. A session depth at which the crash occurred is determined based upon the incident data. Time period data is obtained, the time period data indicative of a time period, following the crash, until the user next accesses the application at the computing device. A severity score for the crash is determined based upon the session depth and the time period.

Abstract: Example embodiments relate to measuring user interface (UI) responsiveness related to a UI of an application running on a computing device. An example method may include accessing a list of actionable controls, where the list indicates at least one UI control of the UI as being actionable. The method may include detecting a user interaction when a user interacts with a UI control of the UI. The method may include determining a list of change controls. The change controls are UI controls that change, refresh or load in response to the user interaction, which will ultimately be displayed on a next screen of the UI after the user interaction, and which are also on the list of actionable controls. The method may include determining a UI responsive time based on the list of change controls.

Abstract: A monitoring system includes a monitoring device for monitoring an event occurred on an original screen and a change from the original screen to a target screen as a result of the event. The monitoring device may include a report monitor to obtain a plurality of event reports, and a report analyzer to group the plurality of event reports into a plurality of report groups and to generate an event summary for at least one of the plurality of report groups.

Abstract: Examples include a user segmentation system for staged application rollout. In an example, an application usage engine receives data for users including a number of screens used in an application version and a number of user actions used in the application version. A staged rollout engine determines, for each user, an early adopter probability based on the number of screens used in the application version and the number of user actions used in the application version, and in some examples a preconfigured factor or weighting. For a subsequent application version, staged rollout groupings are determined. Users are assigned to staged rollout groupings based on the early adopter probabilities.

Abstract: A plurality of application requests from an application may be monitored, with each of the application requests associated with an instance of a plurality of instances of a user interaction with the application. A request response time for each of the application requests may be determined. A user interaction time for each of the plurality of instances of the user interaction with the application may be obtained. From the plurality of instances, slow instances that have a user interaction time greater than a slow threshold may be selected, with each of the slow instances comprising a selected application request of the plurality of application requests. For the slow instances, the request response times of the selected application requests may be averaged to generate an average request response time. A comparison request response time for the selected application request may be determined.

Abstract: Examples relate to calculating user experience scores for applications. The examples disclosed herein collect, for each of at least two frames of each of at least two user sessions of an application, a set of user event data items representing user experiences from the perspective of a user device during a corresponding frame of the at least two frames of a user session. A frame score is calculated for each of the at least two frames in each of the at least two user sessions by analyzing the frame's set of user event data items. A user session score is calculated for each of the at least two user sessions by averaging the frame scores for the at least two frames in the user session of the at least two user sessions. A user experience score may then be calculated for the application by averaging the user session scores of each of the at least two user sessions.

Abstract: A plurality of completion times associated with an application request may be obtained. The plurality of completion times may include a first completion time and a second completion time. A plurality of response times associated with a first asynchronous operation triggered by the application request may be obtained. The plurality of completion times may include a first response time associated with the first completion time and a second response time associated with the second completion time. A first correlation score may be determined describing an effect of the first asynchronous operation on the application request based on the first completion time, the second completion time, the first response time, and the second response time. Visualization data may be generated representing the first correlation score.

Abstract: A computing device includes at least one processor and a machine-readable storage medium storing instructions. The instructions may be executable by the hardware processor to, in response to a detection of an application event during execution of an application, read a plurality of entries included in a call stack at a time of the application event. The instructions are also executable to determine a classification identifier for the application event using the plurality of entries included in the call stack, and store information about the application event in a database entry associated with the classification identifier.

Abstract: Example implementations relate to dynamic graphic entity determination. Some examples may include a data access engine to access data values. Each data value may be associated with a category and a subcategory of the category. Some examples may include a display entity determination engine to analyze the data values to determine whether an overlap of a proposed graphic representing the subcategory and a proposed graphic representing the category exceeds an overlap tolerance level. Some examples may include a visual representation generation engine to generate an actual visual representation based on the analysis. The actual visual representation may display: a single consolidated graphic representing at least the category and the subcategory if the overlap does not exceed the overlap tolerance level; or the subcategory as a separate graphic from a graphic representing at least a portion of the remainder of the category if the overlap exceeds the overlap tolerance level.

Abstract: Examples disclosed herein involve measuring application performance based on user engagement of the application. In examples herein, an activation of an action item may be detected and a degree of user engagement of the application in response to the activation of the action item may be determined. A threshold time for acceptable performance of the action item is adjusted based on the degree of user engagement.

Abstract: Example embodiments relate to generalized snapshots based on multiple partial snapshots. An example method may include accessing multiple partial snapshots, each from a different client. The method may include creating a generalized snapshot from the multiple partial snapshots. The generalized snapshot includes multiple target pixels, and the color of each of the multiple target pixels may be determined by considering colors of multiple source pixels, each from a different partial snapshot.

Abstract: Examples include a user segmentation system for staged application rollout. In an example, an application usage engine receives data for users including a number of screens used in an application version and a number of user actions used in the application version. A staged rollout engine determines, for each user, an early adopter probability based on the number of screens used in the application version and the number of user actions used in the application version, and in some examples a preconfigured factor or weighting. For a subsequent application version, staged rollout groupings are determined. Users are assigned to staged rollout groupings based on the early adopter probabilities.

Abstract: In one example of the disclosure, incident data indicative of a crash of an application is obtained, the crash occurring during an access of the application by a user at a computing device. A session depth at which the crash occurred is determined based upon the incident data. Time period data is obtained, the time period data indicative of a time period, following the crash, until the user next accesses the application at the computing device. A severity score for the crash is determined based upon the session depth and the time period.

Abstract: In one example of the disclosure, transaction data is accessed, the data transaction indicative of a user transaction with an application made via a computing device during a session. A first measurement of duration of the user transaction and a second measurement of duration of the session are determined based upon the transaction data. Expectation data indicative of a user expectation for duration of the transaction is accessed. A user emotional state penalty is determined based upon the first and second measurements and the user expectation.

Abstract: Example embodiments relate to partial snapshots for creating generalized snapshots. An example method may include, in response to an event, accessing a user interface screen or view associated with an application. The screen or view includes multiple pixels arranged over an area. The method may include sampling the multiple pixels by capturing a portion of the total pixels included in the screen or view. The sampling may include maintaining a dispersion of the captured pixels over the area of the screen or view and preventing clustering of captured pixels within sub-areas of the area. The method may include transmitting the captured portion as a partial snapshot to a system to create a generalized snapshot from the partial snapshot and other partial snapshots.

Abstract: Examples disclosed herein involve measuring application performance based on user engagement of the application. In examples herein, an activation of an action item may be detected and a degree of user engagement of the application in response to the activation of the action item may be determined. A threshold time for acceptable performance of the action item is adjusted based on the degree of user engagement.

Abstract: A computing device includes at least one processor and a machine-readable storage medium storing instructions. The instructions may be executable by the hardware processor to, in response to a detection of an application event during execution of an application, read a plurality of entries included in a call stack at a time of the application event. The instructions are also executable to determine a classification identifier for the application event using the plurality of entries included in the call stack, and store information about the application event in a database entry associated with the classification identifier.

Abstract: A plurality of application requests from an application may be monitored, with each of the application requests associated with an instance of a plurality of instances of a user interaction with the application. A request response time for each of the application requests may be determined. A user interaction time for each of the plurality of instances of the user interaction with the application may be obtained. From the plurality of instances, slow instances that have a user interaction time greater than a slow threshold may be selected, with each of the slow instances comprising a selected application request of the plurality of application requests. For the slow instances, the request response times of the selected application requests may be averaged to generate an average request response time. A comparison request response time for the selected application request may be determined.

Abstract: Examples relate to calculating user experience scores for applications. The examples disclosed herein collect, for each of at least two frames of each of at least two user sessions of an application, a set of user event data items representing user experiences from the perspective of a user device during a corresponding frame of the at least two frames of a user session. A frame score is calculated for each of the at least two frames in each of the at least two user sessions by analyzing the frame's set of user event data items. A user session score is calculated for each of the at least two user sessions by averaging the frame scores for the at least two frames in the user session of the at least two user sessions. A user experience score may then be calculated for the application by averaging the user session scores of each of the at least two user sessions.