How to Create Custom Reports

Index:

1. How to Use Custom Reports to Improve Your Chatbot's Experience
2. Generate Event Log with NPS Data

 

How to Use Custom Reports to Improve Your Chatbot's Experience

When building a bot, it is important to monitor its performance to ensure it is fulfilling its purpose as a service and communication channel. 

Additionally, it is through data analysis that we can gain insights to improve our customers' experience.

For this, Blip offers the possibility of event logging (also known as event-tracks or trackings). 

Event logs must be configured in the Builder so that every time a user passes through a point in the flow, this event is saved for analysis. They are composed of:

• Category: the identification of the event log. In general, it refers to the point in the flow where the log will be made.
• Action: the details of what happens in a category. For example, the quantity or choice of the user.

In this article, you will find X tips for reports you can create to monitor your chatbot usage in more detail.

 

1. User Choices
During the conversational flow, it is common for your users to need to make decisions and choose between a series of listed options. 

In these cases, it is important to know which options are the most chosen and which are the least accessed.

To create this type of report, go to the Builder and identify the block where the options are listed, and the blocks to which the user can be redirected depending on the chosen option:

 

 

In the blocks to which the user is redirected, go to actions > entry actions > create event log.

 

 

Create the event log by choosing a name for the category, which should be written the same way in all the blocks to which the user can be redirected. Name the action with the name of the block.

 

 

In the analysis module, go to the custom reports option > create report. Name your report, choose the add chart option, and select the chart format you prefer. In this example, we will use a pie chart.

 

 

In the pop-up that opens, give a title to the chart, select the dimension “custom events”, and choose the category you created in the previous step. Then just click add.

 

 

Then you will have a chart indicating which options are the most and least chosen by your users:

 

 

2. Qualitative Information

Exception flows are a great source of ideas to improve your chatbot. In general, they indicate user inputs and desires that have not yet been mapped as functionalities of your chatbot.

In the Builder, choose the exception block > actions > entry actions > event log:

 

 

Create the event by choosing a name for the category and in the action field use the variable that saves the text entered by your user: {{input.content}}

 

 

In the analysis module, create a custom report with the list-type chart:

 

 

In the pop-up, choose the custom events dimension and the category you created in the previous step:

 

 

In the generated list, you will see the messages sent by users:

 

 

Actions that are repeated most often are good candidates to enter the backlog of new functionalities. Additionally, check if there are opportunities to improve the words and phrases understood by your chatbot to validate the redirection condition.

 

3. Satisfaction Surveys

Satisfaction surveys are great sources of information about the quality of service provided by the chatbot, whether it is fully automated or with human assistance.

To collect the scores from your satisfaction survey, choose the block where the survey question is in your flow and create the event log in the exit actions.

They will be executed after the user sends the message with their score.

 

 

Give a name to the category, and in the action field, enter the first score of your scale.

 

 

At the bottom of the event log configuration, you will find the option to set conditions for the event log. Set the condition as the user's response being equal to the score being tracked with this event. Repeat the process for all the scores accepted by your scale.

Repeat the process for all the scores accepted by your scale.

 

 

In the analysis module, create a custom report with a column-type chart.

In the pop-up, choose the custom events dimension and the category of the event you created in the previous step.

 

 

In the pop-up, choose the custom events dimension and the category of the event you created in the previous step:

 

 

Now you can monitor how satisfied your customers are when they talk to your chatbot.

 

 

Consider creating a list of messages from users who gave bad scores. Just follow the procedure described in item 2, but applying it to the block where you ask for user feedback.

 

Generate Event Log with NPS Data

Right after the human service block, in an entry action, save the values of the Ticket information in a variable (the available fields are in our Ticket documentation):

Example:

And/or:

 

Then, in an exit action of the same block or the next block, create an event log where you will put the variables or created variable and the score variable.

Example:

 

When creating the custom report, select the type “custom report” and the Category equal to the event log category, in this example it would be “evaluation”. 

The variables can only go in the “Action” field of the event log.

 
For more information, join the discussion about this topic in our community or watch the videos on our channel. 😃