Human or AI? Realistic Chatbots on the Rise

What is a chatbot?

Chatbots are computer programs designed to provide a realistic conversational experience for humans. Chatbots can process human language (written or spoken) and provide responses of varying complexity. At one extreme are simple text-based chatbots that may only answer simple, one-line questions, such as providing business hours or store locations. 

At the other end of the chatbot spectrum are proprietary virtual assistants, like Alexa, Siri, Google Assistant and Cortana. These chatbots can provide a significantly wider variety of functionality than text-based chatbots can. Each of these chatbots can understand conversational language and are not reliant on text-based input. 

How are chatbots used today?

Chatbot usage varies greatly based on the complexity of the software and how it is deployed. Chatbots such as Alexa or Siri are used routinely by individuals for a wide variety of routine tasks, such as asking for the weather forecast, creating calendar events, or writing and sending text messages. These types of personal AI chatbots are virtual assistants and are unlikely to be used by businesses beyond employee personal use. 

Simpler AI chatbots, though, are being increasingly deployed by businesses across the e-commerce and online spaces. These chatbots typically appear as window pop-ups in a web browser to ask if a visitor needs help. These simple chatbots are already common: A recent survey found that 22% of micro businesses, 20% of small businesses, 11% of medium-size businesses and 12% of large businesses use these chatbots. Over the coming years, this deployment will significantly increase: 43% of micro businesses, 60% of small businesses, 80% of medium-size businesses and 71% of big businesses are planning to deploy chatbots, the survey found. 

In a business setting, chatbots are widely used to help customers find answers quickly without requiring human intervention. Typically, businesses deploy chatbots to answer common questions or to provide support outside typical business hours. 

According to chatbot and customer service company Tidio, business owners’ top three reasons for using chatbots are to facilitate faster replies to customer messages (26%), offer round-the-clock customer support (20%) and provide automatic replies to repetitive or common questions (18%). Essentially, business owners view chatbots as a means to improve productivity and provide more efficient service to customers. 

How do customers respond to chatbots? 

Customers’ responses to chatbots vary greatly. The reason the customer is interacting with the chatbot in the first place, along with the other means of assistance available, greatly affects their overall opinion of it. For example, Tidio found that 62% of customers would rather use a chatbot than wait 15 minutes or more to speak to a human representative. [Read related article: Small Businesses Provide the Best Customer Service.]

Similarly, regardless of wait time, customers would rather use a chatbot than speak to a human representative for a range of simple activities. Consider these findings from Tidio:

• 71% of customers would prefer to use a chatbot to check an order status.
• 67% would prefer to use a chatbot for help searching for products.
• 62% would rather use a chatbot to get information and deals.

A survey from the chatbot company Botco.ai found similar responses within the marketing industry. According to Botco.ai, 70% of survey respondents said chatbots answer all or most questions satisfactorily. According to this survey, customers likewise cited using chatbots most frequently for resolving simple issues. Consider these findings:

• 18% of respondents used chatbots to find business hours.
• 17% used them to request product information.
• 16% used them to find nearby store locations.
• 16% used them for customer service requests. 

Although chatbots are effective for simple tasks, customers do not like using them for complicated requests. According to a report from Verint, most customers found chatbots ineffective for detailed requests. Here are some more findings from the Verint report: 

• 32% of respondents said chatbots rarely or never understood them, while 28% said chatbots always or often understood them. 
• 30.5% of respondents said chatbots rarely or never fully answered their questions. 
• 54.5% of respondents said they always or often had to speak to a human representative after using a chatbot. 

Are chatbots really AI?

Depending on the complexity of the chatbot, these programs may or may not employ AI technologies. For example, simpler question-answer chatbots do not employ AI. Instead, these chatbots provide certain answers based on preprogrammed rules determined by the software developer.  

More advanced task-orientated, or declarative, chatbots use some amount of natural-language processing (NLP). NLP allows the chatbot to process the input from the customer, as well as to formulate their responses. In this way, NLP is used in combination with preprogrammed rules to create a chatbot that can answer simple questions or conduct simple transactions in a conversational and natural way. These chatbots also can use algorithms to sort through past recorded conversations to arrive at the best answer. These chatbots are the ones most commonly employed by businesses. 

The most advanced chatbots are the digital assistants created by large tech companies. These chatbots are data-driven and predictive. In addition to using NLP and preprogrammed rules, these chatbots leverage a wider range of AI-associated technologies, such as machine learning and natural-language understanding technologies, which allow the chatbots to understand human speech, provide recommendations to customers and anticipate user needs. 

What is the future of chatbots? 

Chatbot adoption is likely to increase rapidly. Consulting company Mordor Intelligence found that the North American chatbot industry is expected to have a compound annual growth rate of 24.6% between 2021 and 2026. During this time, the cost of chatbot adoption is likely to fall as additional companies enter the market. Additionally, the capability of chatbots is likely to increase as NLP and machine learning technologies are further refined. 

The further development and implementation of natural-language understanding also will further expand chatbot use cases and overall adoption rates. Globally, consulting firm Juniper Research predicts that the chatbot industry will be worth $142 billion by 2024. This change will be further influenced by demographic shifts. Tidio found, for example, that Generation Z survey respondents would rather use chatbots than interact with service representatives. Additionally, 56% of Gen Z respondents thought more companies should use chatbots in general. [Read related article: How to Empower Your Customer Service Team.]

Overall, increased customer demands for chatbots, customer expectations of what chatbots should do, and an overall increase in chatbots’ technological capacity will cause large shifts in this market. Adoption rates are likely to increase significantly, and chatbots will be better equipped to handle a range of human interactions, making them useful beyond predetermined scenarios. 

Chatbot adoption is on the rise, and so is sophistication

Chatbot adoption increased 426% in April 2020, following the first round of lockdowns due to the COVID-19 pandemic. Further adoption of chatbots by small and midsize businesses is likely to be driven by three factors: lower costs, improved technology and growing demand. Already, though, the chatbot market has undergone a round of normalization, as large numbers of customers and business owners believe chatbots fill certain business needs. 

How far chatbot adoption ultimately goes depends on how much the technology improves. If chatbots offer a seamless customer experience across a range of functions, their continued adoption is almost certain. And, with increasing numbers of chatbot service providers, it seems very likely that more businesses will continue to adopt this technology and that greater numbers of customers will come to expect it.