What are chatbots and how are they used?
Past decade’s rise in use of artificial intelligence (AI) has revolutionized computer science and gave birth to many new related technologies.
Many business owners are just beginning to understand what benefits AI can deliver, as more organizations are turning their eye towards a particular manifestation of AI — the chatbot.
Chatbots are programs that use conversational AI designed to simplify human interaction with computers, who can then interpret user’s intent
and respond accordingly (usually by providing information or performing some operation).
“80% of business to customer communication is going to be done through bot messengers within next three to five years.”
Mikael Yang, Manychat
How do chatbots work?
Chatbots process user input via text parsing or speech recognition, interpret user’s intention and respond by performing some relevant operation. Chatbots functionality varies and they may be classified according to several factors including input type, programming type and purpose.
Text and speech
Input type classification indicates what chatbots can use as input to interpret user’s intent. Text input type usually involves users submitting text queries via chat or messenger type interface, where chatbots apply text parsing to interpret user’s intent. With the number of users using messaging apps like WhatsApp, Slack, Skype, Facebook Messenger skyrocketing, humans easily adapt to this type of interaction as it resembles any text-based conversation. Text parsing bots are frequently encountered as your counterpart in a messaging app, or via chat window pop-up on a website.
Another input type is speech, where users speak their inquiry or instruction, and chatbot uses speech recognition and natural language processing to interpret intent. Natural language processing (NLP) is a branch of artificial intelligence that teaches machines to read, analyze and interpret human language. In essence, this allows the computer to understand what you are asking and how to appropriately respond. Speech recognition also may also allow a computer to verify the identify of a human they are speaking to based on their voice. NLP chatbots have been popularized by Siri, Alexa and Cortana, and even if you haven’t used the
Simple and smart
Chatbots can be further classified based on how they were programmed:
Simple chatbots work based on pre-written keywords that they understand and respond according to defined case logic. This type of chatbot often will provide users with a multiple choice menu of responses to help keep the input relevant. If the user has submitted input without using a recognizable keyword, chatbot will commonly respond by saying they don’t understand the input, re-iterating valid input options or forwarding the inquiry to a live agent. In the latter case, a human will take over the conversation until inquiry is resolved or conversation ends. Simple chatbots that are actively managed, will likely be modified over time as more conversation options and outcomes are identified.
Smart chatbots rely on AI when they interpret user input. Instead of matching user input to a matching scripted answer, AI is connected to various databases and resources which are used to qualify the input and determine a relevant response.
Service Type and purpose
Chatbots can be used for various business tasks like collecting information about users, answering questions, booking meetings or even processing orders. The general idea for chatbots is to simplify tasks, improve or streamline customer service and e-commerce. Here are some common types of chatbots according to their purpose or the service they provide:
Customer service chatbots: Many businesses are using chatbots as a first contact when customers need help. Companies often employ chatbots to help customers easily navigate their websites, find information, answers to FAQs and to complete simple tasks like appointment bookings. This type of chatbot is widely used in contact centers where they form IVR systems that try to identify the caller and qualify the call intent. These are usually simple bots that operate strictly using expected input and pre-defined responses.
E-commerce chatbots: Various e-commerce companies use chatbots as an additional interaction channel with their customers. This type of bot is often attached to company’s online store or social media account, and is designed to lead customers through a linear process flow to initiate orders or complete transactions. Similar to customer service bots, these are usually simple bots that rely on text parsing.
Virtual assistant: Personal assistants like Siri, Cortana and Alexa have risen in popularity as their benefits have become well known accepted by everyday consumers. People use them to quickly retentive information, or perform simple tasks such as appointment booking, placing calls, navigation or controlling some of the smart devices within one’s home. Despite performing seemingly simple tasks, virtual assistants are quite sophisticated compared to most text parsing chatbots.
While it’s still early for this technology in terms of it’s maturity, it can definitely free up a lot resources for your business by absorbing some of the functions currently performed by your staff.
Some popular platforms for building chatbots are:
- Dialog flow
- Oracle Virtual Assistant
Quayside Digital Consultants are experienced in building multi-purpose chatbots for use with social media, messengers and websites. We can review your existing processes, operations and identify opportunities where chatbots can automate processes, save cost and deliver value to your clients.