A chatbot, chat bot, or chatterbot, is an application that conducts automated conversation through an online messaging platform. The application uses text or text-to-speech (TTS) to simulate and process human conversation, allowing people to interact with digital devices as if they were real people and potentially replacing direct contact with a live human agent.
In this definition...
What purposes do chatbots serve?
Common examples of chatbots include e-commerce website pop-ups and those built into computer and smartphone operating systems, such as Siri, Alexa, Google Assistant, and Cortana. Chatbots automate operations such as customer service, request routing, and information gathering.
When used on websites, chatbots often use a traditional popup window to ask questions, then provide basic answers, point to other resources on the website that might meet the customer’s needs without the involvement of human support representatives.
Chatbots give organizations that rely on their websites as a primary point of customer service an additional communication option beyond traditional forms, telephone calls, or sending an email.
Typically, they are able to answer common questions, access and route users to knowledge bases and support articles, provide certain automated technical support functions, carry out tasks by taking and carrying out commands, locate businesses, schedule appointments, serve up promotions, and more. These functions are especially useful on B2C websites to provide customer service, sales, and marketing functions.
Chatbots can also support employees or users on internal platforms or inside of applications. Besides business functions, chatbots can be used for play, recreation, and entertainment, such as in talking toys.
According to a recent survey, what people like most about chatbots is their quick responses, the ability to, when needed, connect with a live agents, and the availability of help outside normal customer service hours.
Small Business Computing asks. “Are chatbots right for your business?”
Malicious use
Chatbots can also be used to commit cyber crimes or for other purposes considered unethical or troublesome. In such cases, they are generally known as “bots.”
Bots may be used to issue spam, phishing bait, advertisements, enticements to reveal personal information, promote fake news, false claims, and propaganda, or establish artificial social proof. Dating sites and apps or games may also use bots to create the appearance of platforms populated by other human users and to increase user engagement.
How do chatbots work?
Most chatbot applications are accessed via website pop-ups, but can also send messages through SMS or email. They may prompt users with open-ended questions or selections, and then respond to input by routing the dialog, following up with questions or selections for clarification, or possibly serving up other relevant information.
Some chatbots may provide only simple functionality, such as answering a query with a single-line response. Basic chatbots scan text for keywords and generate responses using common phrases from a library or database. These are often task-oriented or declarative chatbots that focus on responding to inquiries. These chatbots use rules, natural language processing (NLP), and some machine learning (ML). Task-oriented chatbots often behave like interactive FAQs.
More sophisticated chatbots are data-driven, predictive, and conversational. They are often known as virtual assistants or digital assistants and can deliver more complex interactions, deeper personalization, and may simultaneously analyze or process input. These chatbots use natural language understanding (NLU), NLP, and ML to provide predictive intelligence and analytics. They understand more of the context of conversations and can learn user preferences, anticipate needs, and provide recommendations.
Chatbot limitations
While the sophistication of chatbots has increased, they still have trouble dealing with complex queries and topics, irregularities, accents, input mistakes, multiple simultaneous questions, and non-linear conversations that require extensive back-and-forth conversation with the user.
Chatbots are limited by the generative models that guide their interactions and learning. Learning requires large conversational datasets for processing by machine learning methods, and the quality of a chatbot’s conversation is determined in part by the dataset used to train it.
The effectiveness of a chatbot is also limited by the human side of interactions. Many people are uncomfortable or mistrustful when they know they are dealing with a machine rather than a live agent, especially when discussing health, finances, or other other sensitive information.
Chatbot trends
Tools for the creation of chatbots, including low-code development platforms (LCDPs) and no-code development platforms (NCDPs), are becoming more widespread. Many chatbots are now possible to set up out of the box in less than an hour with little or no technical expertise.
Chatbots are also becoming more widespread and accessible due to the expansion of 5G networks. All of this makes chatbots increasingly easy and low-cost to develop and adopt to replace or enhance live human agents for business functions. And the rise of the ‘bots isn’t just happening in customer service. The user of chatbots is increasing in complex fields like finance as well.
eWeek.com reported on the growing consumer acceptance of chatbots in January 2021. At the time, Juergen Lindner, Oracle’s senior vice-president of global marketing said:
“Digital is the new normal, and technologies such as artificial intelligence and chatbots play a vital role in managing finance. The research indicates that consumers trust these technologies to accelerate their financial well-being over personal financial advisors, and business leaders see this trend reshaping the role of corporate finance professionals.”
— “Sign of the times: Bots earning more credibility than people”, eWeek, 2/10/21, accessed 3/25/22
What are the key benefits of a chatbot?
Chatbots offer operational efficiencies and cost savings to businesses. They are able to serve more customers faster, to scale services flexibly, to personalize services, and to engage with customers more proactively than by relying on live human agents. Chatbots can handle simple inquiries and can route more complex issues to live service representatives to allocate resources more efficiently.
In some scenarios, services provided by chatbots can be better and faster than their human counterparts. Moreover, chatbots are available 24/7 to engage and solve problems for internal employees and external customers.
Want to know more?
This definition is just the start. Read deeper on how to put chatbots to work for your organization.
Explore top chatbot platforms for your organization | Enterprise Networking Planet
How do you assure success when you add chatbots to your helpdesk? | TechRepublic
Bots are earning more credibility than people? eWeek’s Michael Hickens makes the case. | eWeek
