How do chatbots and channel selection work?

According to Gartner Twenty-five percent of customer service and support operations will integrate virtual customer assistant (VCA) or chatbot technology across engagement channels by 2020, up from less than two percent in 2017. Chatbots can be used to automate and simplify a number of different areas including sales & lead generation, customer & IT support as well as internal to an organization to automate processes which today are done through emails or often complicated enterprise applications.

Chatbots are run on a platform separately from the channels they integrate with so in it possible run a single platform on a number of channels and should be offered in the channels which your target customers use whereas for organizations to effectively deploy internal chatbots those would be implemented on the platform used for internal messaging.

How does it work

A chatbot functions by integrating a channel (such as Facebook Messenger, WhatsApp, Web or other) with a Messaging Orchestration engine, which in turn process incoming messages to and calls an AI/NLU engine to extract the information from the message from the end user in to something the software can understand. For example the phrase “I forgot my password and my email ID is [email protected]” gets turned in to the Intent “password_reset” and an Entity called email with value of [email protected]. So in essence it translates a written (or spoken) phrase in to specific values which the machine can do something with. In this case it would have predefined message to respond back to the user with how to reset the password or proceed in actually triggering the reset process using the email ID collected. These rules, answers and integrations are maintained in a Knowledge Base.

Additionally a chatbot needs to be able to trigger a set of questions to capture the necessary information from the end user based on triggering a flow. For example if the chatbot is missing information to be able proceed the chatbot should ask the end user questions to obtain this information. Another element of a chatbot is Sentiment Detection, which detects the mood of the customer to potentially modify the answers, or trigger escalation to a Live Agent to resolve the issue.

Finally, chat analytics to extract all kinds of statistics and information of the chat sessions is used to be able to understand improvement areas of the chatbot design. There are currently several chat analytics solutions available including Dashbot, Chatbase from Google, Botalytics. We also offer a chat and AI analytics solution called BotExpress Analytics (botexpress.ai link) which is available if chatbot is running on BotExpress platform but can also be integrated using available API’s.

Facing Customers

Consumer messaging platforms such as Facebook Messenger, WhatsApp or WeChat have become the primary tool of communication between individuals and are now used more than traditional social networks. Consumers expect to be able to interact with companies in the same way. Those companies which are easy and quick to communicate with are those which customers are more likely to use. Whether your chatbot targets sales and bookings, customer support, gaming or is informational, being available on multiple platforms maximizes your customer reach.

Internal Enterprise Channels

As in the consumer world, people within organizations increasingly use chat- based applications as part of their daily work. Innovative companies use chatbots to get things done quickly and easily. From simple applications such as reminders, to those fully integrated with existing enterprise applications chat-based programs are helping employees easily obtain or update applications with the information needed.

Symprio’s Chatbot Design Agency helps organizations build chatbots. Click here to tell us about your chatbot requirements if you think this may be an option for your organization.

Powered by Froala Editor