Analytics

In order for the VoiceTrends analytics platform to track callers progressing through your call-flow, it is necessary to configure the VoiceXML code correctly for each application. There are a few basic rules you need to follow when writing your code to get the most out of the Analytics platform:

1. All forms that you want to track must have an “id” attribute (e.g. <form id=“get_account”>).
2. All form items (<field>, <block>, <record>, <transfer>) that you want to track must have a “name” attribute (e.g. <field name=“account_id”>)
3. VoiceTrends enables logging by default. You can set logging to private and call-path tracking will continue to work. However, you must not explicitly disable logging in the sections you want to track.

For more information about configuring and enabling VoiceTrends in Plum DEV, please visit the VoiceTrends page.

The Analytics feature of Plum DEV can be used to perform analytics on any hosted VoiceXML application. The Analytics tool provides global details about data for an entire hosting account, such as minutes used, number of calls, and transfer rate, all the way down to the granular details of the recorded responses for each form input within an application.

Analytics can be used to track trends in data for such cases as (among many others):

• Captured form inputs
• Verifying code changes and affected behaviors
• Call Transfer Rate
• Call Disconnect Rate
• Most common form input
• Minute usage per application
• Average Call duration
• Page visits for each page within an application
• Form item visits
• No match, No input, Transfers, and Disconnects totals and rates on a per-page and per-form item basis

There are three levels of specificity when using Analytics. Different levels might be more useful based on what information you are trying to extract from your application. These three levels will be described individually to give an idea of what kind of data can be extracted from each.

Application reports consist of data for all applications assigned to a hosting account. The table and corresponding graph in this view display such data as total calls, number of minutes used, average call time, and rate of transfers, among others. This information can be useful for tracking general usage information for each application such as the total number of calls made to the application, the total minutes used by each application, the average duration of each call, and the count/rate of transfers that occur in an application. As this shows, application report analytics provide all the high level details about all applications associated with your hosting account.

Within an application report, there are a number of filters which can be applied to alter which data is shown. When viewing an application report, the following filters can be applied:

Firstly, as for all types of Analytics reports, the time frame can be set to one of the following three settings:

• Last 30 Days
• Last 7 Days
• Custom - allows for user-specified ranges

Changing this time filter will alter the time frame for the displayed data. This can be particularly useful if, for instance, you are running an application which expects a high volume of traffic around a certain holiday. Using this time frame filter, you could tune in to this time to quickly view your usage statistics for this time frame.

Aside from the date filters, the following filters are specific to the application report view.

Phone Number: toggles which application's data is visible on the graph

Data: The following values can be toggled to change which statistics are displayed on the graph. Checking any of these check boxes will instantly add the line to the graph. Note that the scale of these filters is not always the same. For instance, rate values are on a scale of 0 - 100 percent, whereas values representing a total are represented as a number, which can sometimes be very large. This should be taken into consideration when specifying which filters should be used together in order to produce a comprehensive graph.

• Calls - number of calls made to the application for the specified interval
• Minutes - number of minutes used by the application for the specified interval
• Duration - average duration of calls made to the application for the specified interval
• Transfers - number of transfers made by the application for the specified interval
• Transfer Rate - average percentage of transfers made by the application for the specified interval

Data is displayed in Analytics via table and graph. For the application reports, the following data is displayed in the table:

• Phone Number - Phone number to which an application is assigned
• Total Calls - Number of calls made to this application for the specified interval
• Total Minutes - Number of minutes used by this application for the specified interval
• Avg Duration - Average duration of each call made to this application for the specified interval
• Transfers
  • Total - Total transfers made by this application for the specifed interval
  • Rate - Percentage of calls in which a transfer occurred for the specified interval

The graph provides a visual representation of all of the correlating data being shown in the table, depending on the various filters currently applied to the graph (calls, minutes, duration, transfers, and transfer rate). Multiple items may be applied to the graph at the same time, although certain combinations might not necessarily produce a desirable visual representation, as certain values are in percentages where as others are not.

Page report analytics consists of all data for the pages within a hosted application. Running Analytics on a per-page basis can be useful for analyzing applications on a granular level. In the pages report view, you can view data about the following categories for each page of an application: visits, filled rate and total, no match rate and total, no input match and total, transfer rate and total, and disconnect rate and total. This data can be used to determine the usage statistics of your application. Viewing such fields as filled, no match, and no input can help to give an idea of how easy to understand and follow the forms within your application are. For instance, if you are getting a high percentage of no match or no input on a page, this likely hints at the fact that one of your forms might not necessarily be as user-friendly as you'd like. These details can help narrow down unpredictable behavior in your applications which can then further be analyzed in greater detail using form item analytics.

Within a page report, there are a number of filters which can be applied to alter which data is shown. When viewing a page report, the following filters can be applied:

Firstly, as for all types of Analytics reports, the time frame can be set to one of the following three settings:

• Last 30 Days
• Last 7 Days
• Custom - allows for user-specified ranges

Using the time filter can help to pinpoint behavior for a specific interval of time. For instance, if you're expecting heavy traffic around a holiday, you can filter by this range to view usage for this period.

Aside from the date filters, the following filters are specific to a page report.

Phone Number: toggles which application's data is visible in the table and graph

Page: toggles which page's data is visible in the graph

Data: The following values can be toggled to change which statistics are displayed on the page report graph. Checking any of these check boxes will instantly add the line to the graph. Note that the scale of these filters is not always the same. For instance, rate values are on a scale of 0 - 100 percent, whereas values representing a total are represented as a whole number. This should be taken into consideration when specifying which filters should be used together in order to produce a comprehensive graph.

• Visits - Number of visits to the page for the specified interval
• Filled - Percentage of filled forms on the page for the specified interval
• No Match - Percentage of no match occurrences for forms on the page for the specified interval
• No Input - Percentage of no input occurrences for forms on the page for the specified interval
• Transfer Rate - Percentage of calls in which a transfer occurred on the page for the specified interval
• Disconnect Rate - Percentage of calls in which a disconnect occurs on a page for the specified interval

Data is displayed in Analytics via table and graph format. For page report analytics, the following data is displayed in the table:

• Page - Page of the application
• Visits - Number of visits to the page of the application
• Filled
  • Total - Total number of filled forms for the page
  • Rate - Percentage of filled forms for the page
• No Match
  • Total - Total number of form no match occurrences for the page
  • Rate - Percentage of form no match occurrences for the page
• No Input
  • Total - Total number of form no input occurrences for the page
  • Rate - Percentage of form no input occurrences for the page
• Transfers
  • Total - Total number of transfers for the page
  • Rate - Percentage of transfers for the page
• Disconnects
  • Total - Total number of disconnects for the page
  • Rate - Percentage of disconnects for the page

The graph provides a visual representation of all of the correlating data being shown in the table, depending on the various filters currently applied. Multiple items may be applied to the graph at the same time, although certain combinations might not necessarily produce a desirable visual, as certain values are in percentages where as others are not.

Form Item reports consist of statistics for all of the form items for a specific page within an application. Note* In order for Analytics to be generated for form items, such as form, field, and block elements, a name and id must be specified for the form item. If these items are not specified, then no data will be captured for the item! Form Item analytics are especially insightful when it comes to tracking what actually occurs on each page. In this report you can view such details about each form item such as: number of visits, rate and total for filled response, rate and total for no match and no input responses, total and rate for transfers, and total and rate for disconnects. An example of when this data would be particularly useful is when viewing disconnects. If a form item has a high disconnect rate, then it's likely that the form is not easy to follow and users are getting stuck attempting to fill in this form. Additionally, you can affirm expected behavior, for instance if you have a high disconnect rate on a feedback-type form at the end of a survey which prompts for optional user feedback.

When viewing a form item report, there are a number of filters which can be applied to alter which data is shown. When viewing a form item report, the following filters can be applied:

Firstly, as for all types of Analytics reports, the time frame can be set to one of the following three settings:

• Last 30 Days
• Last 7 Days
• Custom - allows for user-specified ranges

Aside from the date filters, the following filters are specific to the form item report.

Application: toggles the number/application which is currently being viewed.

Page: toggles which page's data is visible in the chart and table.

Form Item: toggles which form item's data is visible in the graph.

Data: The following values can be toggled to change which statistics are displayed on the graph. Checking any of these check boxes will instantly add the line to the graph. Note that the scale of these filters is not always the same. For instance, rate values are on a scale of 0 - 100 percent, whereas values representing a total are represented as a number, which can sometimes be very large. This should be taken into consideration when specifying which filters should be used together in order to produce a comprehensive graph.

• Visits - Number of visits to the form item
• Filled - Percentage of filled form responses
• No Match - Percentage of no match occurrences for the form item
• No Input - Percentage of no input occurrences for the form item
• Transfer Rate - Percentage of calls in which a transfer occurred for the form item
• Disconnect Rate - Percentage of calls in which a disconnect occurs for the form item

Data is displayed in Analytics via table and graph format. For the form items report, the following data is displayed in the table:

• Captured Form Input - Clicking this '?' icon will display the captured values for input for this form item. See below for more details.
• Form Item - name of the form item
• Visits - number of visits to the form item
• Filled
  • Total - Total number of filled reponses for the form item
  • Rate - Percentage of filled reponses for the form item
• No Match
  • Total - Total number of no match occurrences for the form item
  • Rate - Percentage of no match occurrences for the form item
• No Input
  • Total - Total number of no input occurrences for the form item
  • Rate - Percentage of no input occurrences for the form item
• Transfers
  • Total - Total number of transfers for the form item
  • Rate - Percentage of transfers for the form item
• Disconnects
  • Total - Total number of disconnects for the form item
  • Rate - Percentage of disconnects for the form item

The '?' icon can be clicked to provide additional information about captured inputs for any given form item. This data can be useful when determining the frequency of occurrences for form input. Clicking this '?' icon will display a bar graph menu of the form inputs, which should look similar to the following:

Using this feature can give insight as to which form inputs are most frequently selected. This can be particularly useful when prompting for a menu selection. Using Analytics for form items can help determine the most frequent form input selections, which would allow you to optimize your prompts to play the most frequently selected options first. This would reduce the amount of prompts the user would have to listen to while simultaneously reducing the duration of the call.

The graph provides a visual representation of all of the correlating data being shown in the table, depending on the various filters currently applied. Multiple items may be applied to the graph at the same time, although certain combinations might not necessarily produce a desirable visual, as certain values are in percentages where as others are not.