It is fair to say that “conversational analytics” does not exist today. Even if some companies may already be using it via Twitter type feeds, it is still an untouched territory in the mass scale of the Internet business.
Conversational analytics is a gold mine of user feedback, and there is nothing better than seeing praises and complaints articulated openly in their most naked form.
Web Analytics Versus Conversational Analytics
If you have a business relying on the Internet presence, you are most likely familiar with these terms of Web analytics: Bounce Rate, Conversion Rate, Time Spent, etc. In the absence of a recorded conversation with visitors, these measures are the only tool to guess the visitors’ intent and feedback. Hence, the concept of growth hacking has emerged to launch a trial and error process to improve presentations one step at a time. If you had a chatbot recording conversations about your products and services, then you would experience a major advantage:
Web analytics only shows you how your existing presentation performs (in time). Conversational analytics shows you what you are missing (instantly).
Conversational analytics, even before forming a statistical result, would constantly indicate the weaknesses of your presentation through user feedback like this: I cannot find your address … why should I buy this… when will this be available… where is the information related to….
Human Assisted Chatlines Versus Chatbots
Recently, human-assisted chatlines have been increasing in numbers deployed for marketing and support operations. It can obviously be very expensive to maintain such operations and conversational analytics obtained from them can be too few in quantity. In a manner of self-fulfilling prophecy, the meaning of conversational analytics is often reduced to managing FAQs where the identified FAQs would still keep the same human labor busy, perhaps 90% of the time. Chatbot effect is shown in the diagram below. Note that a Chatbot is not an FAQ machine, however if designed properly it can handle most of the FAQs while reserving human labor for more in-depth conversations.
Not only the bulk of the human labor can be eliminated by chatbots, but also the conversational analytics would have a different information embedded in them. For example, people may be asking “what is the price?” repeatedly. If this information was handled, they would move on to ask “Do you have any volume discounts?” If that was handled, they could move on to ask another more detailed question. Before human operator is engaged a chatbot can pave the way for more in depth questions to arise from the visitors which would otherwise emerge while keeping human labor busy. Since human labor is expensive, and slow, some of such questions would never reach the business without a chatbot implementation.
Conversational Analytics Behind a Firewall
Remember the old way of maintaining a complaint box in the entrance of offices? For internal operations, a chatbot assisting employees in an anonymous manner could be a real value, sometimes priceless feedback to the management. Otherwise, for obvious reasons, many suggestions and complaints may not surface to maintain the employee-employer relationships, which would progressively result in blind operations.
Upper management can show its confidence by deploying an anonymous chatbot for suggestions and complaints.
Anonymous or not, and internal chatbot could provide valuable information to the upper management to include, but not limited to, these factors:
- Suggestions for operational and business matters to upper management
- Information needs to conduct mission critical operations
- Satisfaction levels with compensation and benefits
- Promotional expectations and ambitions
- Readiness for emergencies, safety, and security tutorials.
- Maintenance support, remembering training knowledge quickly on the spot
All these needs of the employees would become transparent in conversational analytics as a byproduct of using an internal chatbot.