When evaluating chatbot performance, key performance indicators (KPIs) help ensure it's effective, efficient, and delivering value. Here are some essential baseline KPIs for chatbots in customer service and hospitality, which you can evaluate performance on.
Most of these parameters you can monitor from the intuitive Myma.ai Dashboard.
1. Response Time
Measures how quickly the chatbot responds to user inquiries. Faster response times are crucial for a positive user experience, with a good goal being less than 5 seconds.
2. Resolution Rate
Tracks the percentage of conversations where the chatbot successfully resolves the user's query without human intervention. Aim for a 70-80% autonomous resolution rate. For anything that the chatbot is unable to answer, it should create a ticket and assign it to the right staff member.
3. Fallback Rate
This metric shows how often the chatbot fails to understand or handle a user’s request and escalates it to a human agent or falls back on a generic response. A lower fallback rate indicates better training and adaptability.
4. User Satisfaction
Myma.ai's Chatbot measures customer satisfaction during each conversation by allowing users to leave a thumbs up or thumbs down, along with optional comments. A satisfaction score of 90% or higher, based on positive feedback, is generally considered a strong indicator of effective service.
5. Engagement Rate
Represents the percentage of users actively interacting with the chatbot. High engagement can indicate the chatbot’s ability to attract and retain user attention. You can measure this as a %age against the website traffic. Anything greater than 1% is good. You should also check how many conversations are outside your normal business hours, which will usually go unanswered.
6. Conversation Completion Rate
Percentage of conversations where users complete the intended task, such as booking a reservation, finding information, or completing an inquiry.
7. Website Referrals & Leads
The Chatbot will refer customers to different pages of your website and capture group booking, events, and other leads. Measure how much value that is generating.
8. Average Conversation Length
Tracks the average time or number of messages exchanged per conversation. Shorter interactions may indicate efficiency, but very short conversations could also indicate unresolved issues.
9. Conversion Rate
Measures how many booking referrals the Chatbot can generate and the value of those referrals.
10. Training Data Utilization
An internal KPI measuring how effectively the chatbot utilizes training data to answer queries helps gauge the quality of its machine learning model. This is something that the Myma.ai team monitors closely.
Setting and tracking these KPIs helps you understand how well the chatbot is performing, where it can improve, and ultimately, how to provide better service.