The bot builder's robust natural language feature empowers you to configure responses to customer queries with information from the knowledge base/FAQs. This helps you make engaging conversations with customers and streamline issue resolution.
The Natural Language Performance (NLP) report provides essential metrics to evaluate the efficacy of natural language in bots.
- By default, the report presents data from the last 30 days, but you can customize the date range or apply various group by filters.
- You can export the report to your mail or download the report as PDF.
- As this is a curated report, you cannot edit the report or add new metrics/filters. However, you can clone the report or create a custom report and add your preferred widgets/metrics.
To access the report:
- Go to Support Analytics and click Curated Reports from the left panel.
- Choose Natural Language Performance Report from the list of curated reports.
The report comprises the following widgets:
- Volume Analysis trends: Lets you analyze the performance trends of answers in your NLP module, providing data on user queries, answered queries, small talk interactions, and Q&A responses. Use this data to assess the effectiveness of your responses and how they benefit your customers. Leverage this information to improve your answers.
- Q&A performance analysis: Offers details on the number of responses, feedback, and helpfulness percentage received for each question and answer combination. Utilize this information to identify questions with lower scores and take measures to improve them.
- Answer and Q&A response rate: Lets you monitor trends of answer rate and Q&A response rate to identify periods when these KPIs may decline and take corrective actions.
- Q&A response performance trend: View trends in Q&A responses, feedback received, and helpfulness.
List of attributes (metrics, filters, and group by)
Each widget in this report comprises metrics/attributes from different modules on Chat and Bot analytics. For example, the Volume Analysis trends widget is created using metrics from the Answers and Feedback modules.