The fun part was creating your chatbot’s persona and writing the content. Now it’s time to measure how well your chatbot is performing. For big organisations with call centres a great metric is the drop in calls the call centre receives that can be handled by a chatbot. For smaller businesses, you need to rely on analytics to measure your chatbot’s performance.
5 metrics to measure success
For smaller organisations the following are 5 KPIs (Key Performance Indicators) can help you assess your chatbot’s performance. The good news is that all of this information should be available in your chatbot software’s analytics dashboard.
- Response time: how long it takes for the chatbot to respond to a user’s query. It’s important to keep response times low to ensure a positive user experience.
- Completion rate: tracks the percentage of conversations that the chatbot successfully completed without needing to escalate to a human agent. A high completion rate indicates that the chatbot is effectively answering user queries and resolving their issues.
- User satisfaction: this measures user satisfaction with the chatbot’s performance, typically through post-chat surveys or ratings. It’s important to monitor this metric to ensure that users are happy with the chatbot’s responses and overall experience.
- Escalation rate: you can track the percentage of conversations that the chatbot was unable to resolve and had to escalate to a human agent. A high escalation rate can indicate issues with the chatbot’s capabilities or design.
- Conversation length: This measures the length of time it takes for a user to complete a conversation with the chatbot. Shorter conversations are typically better, as they indicate that the chatbot is able to quickly and efficiently resolve user queries.
Measuring answered rates
It’s also a good idea to measure the percentage of answered and unanswered user queries. A chatbot is never going to be able to answer 100% of user queries, but and answer rate of over 80% is a great goal.
If you need help working on your chatbot’s metric, contact us to see how we can help you improve your chatbot’s performance.