Great question, Graham. I don’t pay much attention to industry benchmarks with chatbots because there are too many variables -- from traffic volume to the structure of the sales chatbot to how you are driving people to the chatbot.
You are going to have to benchmark against 1) the goals and purpose of the chatbot 2) the performance of that particular chatbot. For the first one, if the goal of the chatbot was to generate leads to nurture or for the sales team to follow up with, is that happening? If the goal is to get qualified, booked sales meetings, is that happening? The main ratios you are going to look at here are:
- View to chat rate
- Chat to lead rate
- Chat to booked sales meeting rate
When it comes to benchmarking against this particular chatbot, after the first 50 chats, you have enough data to set the benchmark and optimize the chatbot’s performance. In your analysis of the conversations, do you see areas where more people could be completing the chatbot, converting into leads, or booking sales meetings? If so, make those adjustments and re-benchmark. It is just as much a qualitative customer experience review as it is a quantitative metrics analysis.
I created a quick video recently about what you should expect immediately after you launch your chatbot. It doesn’t cover industry benchmarks but focuses on expectations, adjustments, and people using the chatbot the way it was intended.