Viewing and generating chatbot reports in BMC Helix Dashboards
Out-of-the-box chatbot dashboards
Dashboard name | Role | Description | Reference |
|---|---|---|---|
Chatbot dashboard | Administrator | View the metrics for all chatbots in your system:
| |
Chatbot Overview | Chatbot owner | View the following metrics to assess the performance of each chatbot in your system:
| |
Chatbot User | Chatbot owner | View the following metrics to assess the performance of each chatbot in your system:
| |
BMC HelixGPT conversation details | Administrator | View the following metrics for all the conversations between BMC Helix Virtual Chat and BMC HelixGPT.
|
|
Example: Chatbot dashboard
A company has implemented multiple chatbots across three channels—Slack, Microsoft Teams, and the chatbot UI. The administrator wants to know the distribution of users across these channels, the users' feedback, and the most asked queries in these chatbots.
The administrator observes the following trends and can take actions to improve the effectiveness of the chatbots in the system:
Observed trend from the example report | Possible action |
|---|---|
29 percent of users have rated the chatbot as Dislike, Terrible, or Ok against 48 percent of Good and 22 percent of Excellent rating. | Investigate why 29 percent users have given a negative feedback. |
80 percent of users are using the default chatbot UI followed by 13 percent users in Microsoft teams and 7 percent in Slack. | Promote the adoption of other channels in the organization. |
In General Chatbot, the number of live agent transfer request are the highest. | Investigate how this number can be reduced by modifying the intents. |
Example: Chatbot Overview dashboard
A company has recently implemented a chatbot for HR-related services. The HR chatbot owner filters the dashboard based on HR chatbot to view the metrics and determine the performance of the HR chatbot.
The chatbot owner observes the following trends and can take actions to improve the effectiveness of the HR chatbot:
Observed trend from the example | Possible action |
|---|---|
There were 18 unsuccessful knowledge searches. | Investigate which search queries did not return any knowledge articles and work enhance the knowledge base for these search queries. |
There are 83% returning users, that is, end users are using the chatbot for self-service. | Monitor this metrics to see whether this number further increases indicating 100% adoption of chatbot. |
The request-status is one of the least recognized intents. | Investigate whether you want to promote this service among the end users. |
Example: Chatbot User dashboard
A company has recently implemented the default BMC Chatbot. After six months of its implementation, the chatbot owner filters the dashboard based on BMC Chatbot to view the metrics and determine the users' response to the chatbot:
The chatbot owner observes the following trends and can take actions to improve the user engagement with HR chatbot:
Observed trend from the example | Possible action |
|---|---|
There are five active users. | Increase user adoption by performing all or some of the following tasks:
|
A user has given the Dislike rating to the chatbot. | Investigate the reason and take remedial measures to avoid a negative rating. |
The percentage of returning users dropped considerably from October 2020 to November 2020. | Investigate reasons for this reduced user retention. For example, poor intent matching or negative feedback from users. |
The chatbot UI is the most popular channel. | Investigate ways to promote other channels such as Microsoft Teams and Slack or invest in the commonly used channel. |
Example: BMC HelixGPT conversation details dashboard
A company has recently implemented the BMC HelixGPT in BMC Helix Virtual Agent. After six months of its implementation, the administrator filters the dashboard based on BMC HelixGPT conversations details to view the metrics and determine the users' response to the chatbot:
The administrator observes the following trends and can take actions to improve the conversation between BMC HelixGPT and the BMC Helix Virtual Agent:
Observed trend from the example report | Possible action |
|---|---|
Service requests submitted (GPT bots) | Understand the trends of the service requests raised and see if a self-help knowledge article can be created to quickly help the users and reduce the query resolution time. |
Live Agent sessions in GPT bots | Investigate why the virtual conversations got transferred to live agent and determine if there are ways to improve the virtual chat help and reduce the deflection rate. |
Conversations by skill name | Investigate the distribution of conversations based on skills and improve the skills as required or add new required skills. |