Custom analytics
Create your personalized statistical reports to monitor the performance of your bot
If the predefined analytical data that we provide is insufficient, you can create your own personalized data reports through the Custom Analytics menu.
Create a rapport
To create your personalized analytical reports, go to Custom Analytics > Configuration > Report and click on Add a report. Give a name and click on Create.
As you see in the screen-shot above, you can also create a report via import. Click on Choose a file then Import.
Please note that once you have created a report, you can apply filters to it such as consultation space, language and time period. Data will be apdated automatically in the report.
Once the report is created, you can start to add graphics (we call them "components") inside. To do so, click on Add a new component.
Choose the one that suits you by clicking on it, which will take you to the configuration panel of the component.
Let's take a closer look at the Table and the Graphic component, as they are a bit more complexe.
They both have the 3 following sections: sources, loops and restrictions.
Sources: This section allows you to select and configure the data to be displayed.
Converter: a converter is a type of data. Our solution provides a wide variety of converters. You can refer to the Annex section for the complete list.
Inner loop: an inner loop allows for the distribution of data based on the selected criterion (e.g., by language, consultation space, solution used, etc.). If applied, the converter will be sorted according to the selected criterion.
Restrictions: a restriction is a limit that can be applied to the converter. Only data that meets the restriction criterion will be displayed.
Dimension: a dimension is a restriction criterion. You can only use 1 dimension per restrction
Note: a converter can only have one inner loop but can have several restrictions.
Loops: a loop has the same function as an inner loop aboved-mentioned. The main difference is that an inner loop will be applied to one converter inside the source, while a loop will be applied to all the converters in the component.
Restrictions: this section has the same function as the one in the Source section. The main difference is that, here, restrictions will be applied to all the converters in the component.
Once you have configured your report, if relevant data is available on your bot, you will see immediately the results in the Preview section below.
Export a report
Once you've created a report, you can export it in a few simple steps.
First, go to Custom Analytics > Reports and select the report you want to export at the top of the page.
You'll then see two export options - one to export the entire report and the other to export individual components of the report.
It's important to note that you can export the entire report in either Excel or XML format, while components can be exported in Excel, XML, or JSON format.
Receive a copy of your report by email
After creating a report, you can receive regular copies of it in your email inbox.
To do this:
Create an email export in Custom Analytics > Configuration > Exports:
Click Add an export.
Give the export a name.
Set the sending frequency (daily, weekly, on a specific day of the week, etc.).
The report format will always be Excel.
Finally, choose the report of which you want to receive the copy, and click Add.
Subscribe to the email export that you just created by going to your Account settings,
then select the email export:
Selected email exports will also appear in Custom Analytics > Configuration > Exports. The Unsubscribe option allows you to stop receiving the copy sent to your email inbox.
View a report
Once created, reports can be viewed in Custom analytics > Reports
Below are a few examples of personalized reports:
Create a personalized alert
Use alerts to track the evolution of your bot's data and receive notifications when necessary.
Note: alerts can only be applied to converters that are available in custom analytics. Refer to the Annex section to see all converters available.
First step: create a predefined source
Before creating an alert, you need to first create a predefined source, which corresponds to the type of data that your alert is supposed to track.
To acheive this, follow the steps below:
Go to Custom analytics > Configuration > Predefined sources and click on the + button to add new a new source.
Fill in the fileds:
Name (optional): if the filed is emptly, it will be automatically filled out with the name of the converter you choose.
Unity (optional)
Max item (optional): it defines the maximum amount of items (from 3 to infinit) that will be shown in your report.
Converter (mandatory): select from the drowdonw list the type of data you want to track
Internal loop (optional): it allows you to represent the data source (converter) in different categories such as language, consultation space or Livechat agent's competency.
Restriction (optional): it enables you to restrict the data source according a predefined dimension.
Note: a converter can only use one internal loop but have several restrictions.
Once done, your predefined source is saved automatically and it can now be used to create your alert.
Second step: configure your alert
Go to Custom analytics > Configuration > Alerts and click on the "Add an alert" button.
Fill in the fileds:
Label (mandatory): it corresponds to the name of your alerts.
Predefined source: choose the predefined source you just created from the dropdown list. You can already notice that an alert can only track one data source.
Trigger type: it corresponds to the format in which the data will be showed. You can choose between Absolute (data will be shown in numbers) and Relative (data will be shown in %).
Threshold:
Is less than: the altert will be triggered when the indicateur is less than the Min threshold.
Is greater than: the altert will be triggered when the indicateur is greater than the Max threshold.
For example, if you define an absolute minimum threshold of 5 on the number of conversations during a given period, you will receive an alert when the bot has only 4 conversations during that period.
Sampling period: it corresponds to a period of time during which the threshold will be compared. For example, if you set the sampling period to "week", the alert will be triggered if the total amount of conversations during the week is less/greater than the defined threshold.
Compared to (the option is available only when a relative threshold is used): this allows you to determine the time period that the sampling period will be compared to. For example, if you set the sampling period to "day" with a relative minimum threshold of 10%, using "compared to the day before" means the alert will be triggered if there are less than 10% of new conversations compared to the previous day. This feature is useful to track the evolution rate of a data source.
Save your alert by clicking on Update.
Final step: receive and interpret your alert
If an alert is triggered, you can view it in Custom analytics > Alerts.
Preferences
Dimensions
A dimension is used in custom analytic reports to filter data according to specific criteria.
For example, you can use the "consultation space" dimension to filter conversation data from a specific consultation space of your bot. Another example is "Livechat competency", which allows you to show data belonging to a specific domaine of expertise of your Livechat agents.
The BMS provides several default dimensions that can be used directly when creating your personalized reports, but you can also add custom dimensions.
NOTE
Please note that custom dimensions must be variables coming from your conversations. They can be found in Learning > Conversations
How to create custom dimensions ?
To create a custom dimension, go to Custom analytics > Configuration > Preferences, and click Add Dimension. Select DialogExternalValue.
In the "Optionl Configuration" field, enter the variable you want to use as your dimension. For example, use the variable name that we saw above.
Save it by clicking on the check icon.
Your new dimension can now be immediately used in your personalized analytical reports (Custom analytics > Configuration > Reports) as an internal loop or a restriction. Please note that custom dimensions only apply to statistics collected afterward and not backward.
Compute period
This section allows you to determine the number of days for which your analytics will be created.
Reset
By clicking Clear bot analytics, all your bot's analytics will be cleared.
Transfer the analytics to the endpoint URL
This feature allows you to generate automatic analytics on all dialog data associated with your bot. In this section, you will then be able to configure your automatic exports of dialog data in XML format.
Thus, for a configured bot, the system retrieves all the data from the completed dialogs every minute and then sends all the data in POST request on the listening URL (see diagram below).
More concretely, please indicate the endpoint URL of the bot for which you want to export the dialog data, in the Push Analytics To field Endpoint URL then click Save. A POST request is thus sent to each recorded dialog, which then makes it possible to obtain the data automatically without the need for new exports.
Note: you can press the Test button that appears below the field to test and verify that the export is functional.
Annex: list of converters
Converters represent all the data that you will be able to use in your analytics reports. This page presents all the available converters.
Dialog
Number of dialogs: this is the number of dialogs;
Average duration of auto chat dialogs without qualification: this is the average duration of a dialog with the bot (without qualification);
Average duration of auto chat dialogs with qualification: this is the average duration of a dialog with the bot (with qualification);
Number of dialogs with Livechat:this is the number of Livechat dialogs;
Average duration of Livechat dialogs (without survey completion): this is the average duration of a dialog with the bot (without taking into account the filling out of the survey);
Livechat dialogs duration (with completed survey): this is the average duration of a dialog with the bot (taking into account the completion of the survey);
Number of Livechat dialogs picked up: this is the number of successful Livechat escalations;
Number of picked up dialogs with at least one manual answer from the operator: this is the number of Livechat escalations where the operator has at least provided a manual answer (and has therefore not only used predefined answers);
Duration between escalation and first answer: this is the average time between escalation and the first answer from the operator. This includes the length of the queue as well as the number of time that the previous operator was waiting for a forwarded dialog. The automatic greeting answer on the Livechat is considered an answer;
Breakdown of time between escalation and first response: this is an indicator to divide the time needed for successful chat escalation and first answer from the operator;
Duration between escalation and dialog opening (the operator has logged off): this is the average time between escalation and opening of the Livechat dialog;
Number of dialogs passed by the queue: this is the number of dialogs that were preceded by a queue before they could be assigned;
Queue length: this is the average time spent in queues (in seconds);
Number of people who left the queue: this is the number of people who left the queue;
Satisfaction with automatic dialog: this is the satisfaction data on dialogs with the automatic bot;
Average number of Interactions by dialog: this is the average number of Interactions by dialog;
Distribution of dialogs by qualification: this is a distribution of dialogs according to their qualification (direct answer, Livechat, failure, etc.);
Number of completed surveys: this is the number of completed surveys at the end of a dialog;
Filling time of the operator survey: this is the average filling time of the survey sent by the operator at the end of a Livechat dialog;
Time between receiving the dialog and the first answer:this is the average time between the operator receiving the dialog and their first answer (an automatic greeting response Livechat is considered an answer);
Time between receiving the dialog and the first manual answer: this is the average time between the operator receiving the dialog and the first manual answer (the automatic greeting answer Livechat is ignored);
Average response time: this is the average time between each user's question and the answer provided by the operator (in seconds);
Number of abandoned dialogs: this is the number of dialogs left by the user within 90 seconds;
DialogsCallbackRequestsCount: this is the number of phone callback requests;
DialogsCallbackDurationBeforeOperatorAnswersThePhone: this is the average time before the operator supports the recall;
DialogsCallbackDurationEndCallAndRemoveDialog: this is the average time between the end of a call and the deletion of the dialog;
DialogsCallbackDurationBetweenConnectedAndCompeted: this is the average duration of a call;
DialogsCallbackRatioRequestsAnswer: this is the ratio of the supported callback requests to total callback requests;
Distribution of LiveChat opportunities: this is the number of Livechat opportunities triggered by knowledge.
Interaction
Number of operator answers made directly after the user's question: this is the number of answers given by the operator after a user's sentence;
Number of Interactions: this is the total number of interactions;
EndUserInteractionsCount: this is the number of user interactions;
Number of business interactions: this is the number of business interactions;
Total feedback count: this is the number of feedback sent by the user;
Average dialog engine runtime: this is the average dialog engine runtime (in milliseconds);
Average servlet calculation time: this is the average servlet computation time (in milliseconds);
Maximum dialog engine runtime: this is the maximum runtime of the dialog engine (in milliseconds);
Servlet maximum compute time: this is the maximum servlet response time;
Distribution of Interactions by answer type: this is the distribution of interactions by answer type.
Survey
Number of completed surveys: this is the number of completed surveys;
Number of completed survey fields: this is the number completed survey fields.
Volume
Number of supervisor help request: this is the total number of supervisor help requests;
Number of manual transfers: this is the number of manual transfers made during Livechat dialogs;
OperatorNotPickedDialogsCount: this is the number of dialogs automatically transferred if an operator did not retrieve them;
Number of dialogs using knowledge (associated with entry point): this is the number of dialogs using knowledge;
Number of knowledge feedback per dialogs: this is the number of knowledge reviews per dialog;
Number of dialogs using a tag: this is the number of dialogs using a tag;
Number of actions: this is the number of actions;
Number of clicked links: this is the number of clicked links;
Number of suggested links: this is the number of clicked links;
Waiting queue capacity: this is the total capacity of the waiting queue;
Queue occupancy rate: this is the queue occupancy rate;
Number of visitors: this is the number of visitors (a visitor is identified as such for 24 hours via a cookie);
TeaserClickCount: this is the number of clicks made on a teaser.
Satisfaction
Total feedback count: this is the total number of reviews provided by users;
Number of negative feedback on knowledge per dialog: this is the number of negative feedback of knowledge per dialog.
Duration
Average duration between escalation and opening of the dialog (the operator has opened the dialog): this is the average time between escalation and opening dialog (when the operator takes over the dialog)
Time in seconds with at least one operator available: this is the time during which there is at least one available operator (in seconds);
Operator statuses duration: this is the length of time an operator is in a particular status (Phone, Busy, etc.);
Connection time (available): this is the connection time of an operator.
Operator
Number of max connected operators: this is the maximum number of connected operators;
Occupancy rate of an operator: this is the occupancy rate of an operator (busy time with at least one dialog / time connected with available status);
Simultaneity rate: this is the average concurrency rate of the dialogs managed by the operator;
Breakdown of Concurrency: this is the split of the simultaneity rate of the number of dialogs;
Operator productivity rate: this is the productivity rate of an operator. It is calculated as follows: it retrieves the number of dialogs opened by the operator, then divides it by its connection time (in seconds) then is multiplied by 3600 to obtain a value per hour.
Cobrowsing
Number of dialogs with cobrowsing: this is the number of dialogs with cobrowsing initialized;
Number of cobrowsing errors: this is the number of errors encountered during initialization of a cobrowsing dialog;
Amount of sent cobrowsing requests: It is the number of sent cobrowsing requests;
Number of accepted cobrowsing requests: this is the number of accepted cobrowsing requests;
Number of rejected cobrowsing requests: this the number of cobrowsing requests that have been rejected;
Waiting Queue
Time spent in queue: this is the average time user spends in the queue (in seconds).
Codes corresponding to converters
Last updated