Artificial Intelligence

AI in customer service: The past, the present, and the future

Nuno Brito

By Nuno Brito

0 min read

Customer Service Ai

Understand the basic types, provided insights, and future-looking applications of artificial intelligence.

Think of the role of artificial intelligence (AI) in customer service and most will immediately think of a chatbot. Indeed, chatbots are a common way to employ AI in a customer service setting. There are, however, many other ways that AI-enabled technologies can improve customer service.


What is AI?

AI is the ability of a computer—or a robot controlled by a computer—to do tasks that are usually completed by humans because they require human intelligence and the ability to judge a situation or event. There are four basic types of AI:

  1. Reactive machines. This is the most basic level of AI. These machines cannot create memories or use learned information to influence future decisions; they are only able to react to situations that are happening in real-time. IBM’s chess-playing machine, Deep Blue, is an example of a reactive machine, as it observes chess moves and reacts based on programmed chess strategy. In the contact center, simple chatbots that can answer pre-programmed questions would be reactive machines.
  2. Limited memory. As its name implies, a limited memory machine can retain some information earned from previous events or data input. Limited memory machines build their knowledge base using that memory in tandem with programmed data. Autonomous vehicles fall into this category, comparing stored data, such as routes, with real-time information, such as the speed of surrounding vehicles. In the contact center, a system that has access to a customer’s historical interactions with a brand, and uses that information to make decisions to react to the current interaction, would be an example of a limited memory machine.
  3. Theory of mind. Theory of mind (ToM) machines are designed to emulate human thinking and understand that humans have feelings and thoughts that influence their actions. In the contact center, these AI machines would be programmed with natural language processing (NLP). NLP allows ToM machines to detect tone of voice and other subjective characteristics of the interaction, such as detecting when a caller is becoming frustrated or irritated.
  4. Self-awareness. This is the ultimate goal of AI researchers, to develop a machine that has human-like awareness and is conscious of its existence and effect on its surroundings. This level of artificial intelligence is still the work of science fiction writers.

The history of AI in customer service.

Technology has long driven advancements in the contact center. From the earliest days of private branch exchange (PBX) systems to automatic call distributors (ACD) to interactive voice response (IVR) systems, technology has been central to innovation, efficiency, and productivity. Today, AI is continuing that tradition. However, its arrival in the contact center is not a novelty.

As in all other industries, the rise of the internet and text messaging in the 1990s brought tremendous change to contact centers, empowering customers with new channels with which to engage with brands.

These technologies enabled the development of the first commercial bots—computer programs that simulate human conversation through voice commands, text, or both. The internet made it possible for consumers to contact a brand at any time of the day; chatbots filled the need for 24/7 baseline customer service that humans could not provide.

Happy Agents Happy Customers

EBOOK

How AI improves agent experiences in retail

The first “chatterbot,” ELIZA was developed at the Massachusetts Institute of Technology in the mid-1960s. Originally developed for a psychological setting, ELIZA would aid doctors treating patients.

First created to demonstrate the theory that communication between humans and machines would be superficial, chatbots have turned out to be key to customer service in today’s modern contact center.


How AI is being used now.

AI currently supplements the human workforce through the deployment of virtual agents, essentially chatbots that can handle frontline activities such as account or balance lookup, password resets, and other common, yet low-complexity activities.

Aided customer self-service is another current use case for AI in the contact center. This type of assistance quickly provides relevant information to customers, helping to increase customer satisfaction (CSAT).

As time progressed, artificial intelligence solutions also aided human contact center agents, prompting them with next-best actions, reducing data entry, and handling other mundane aspects of the role of an agent.

Solutions like Talkdesk Agent Assist provide agents answers or support to progress the conversation and simplify tasks such as searching product information. Agent assist technology provides human agents upsell and cross-sell opportunities based on access to database information on the products and services purchased by customers.

A third way that AI is being deployed in today’s contact centers is for data collection and analysis. The volume of customer data generated through the contact center is vast and can provide a wide array of insights, including:

  • Customer intelligence. Allows brands to analyze the customer data stored in its CRM and gain a more thorough view of the customer.
  • Conversational intelligence. Provides subjective insights into the customer’s mood, how empathetic the agent was, and other aspects such as sarcasm. This data is typically exclusive to voice interactions.
  • Interaction intelligence. Comprises an analysis of every aspect of the customer experience from the minute the customer dials the phone or engages a chatbot. This includes data like estimated wait time, compared with the actual time the customer waited to engage with an agent.
  • Contact center intelligence. Analyzes every customer interaction—conversation, email thread, and text. Contact center intelligence provides a 360 view of contact center operations and can provide recommendations for improving the customer experience.

What the future looks like for AI.

To be clear, artificial intelligence and AI-enabled virtual agents will not replace human agents in the call center any time soon. That said, the 2020-21 pandemic did prompt many contact centers to accelerate the adoption of AI-enhanced solutions.

Future contact centers will increasingly employ a hybrid model that combines AI and human interaction. With chatbots and guided self-service rising in efficiency in handling low-level customer inquiries, human agents will continue to address more complex issues and problem-solving—as well as providing the empathetic and personalized customer service that consumers demand. Here are a few future-looking applications of artificial intelligence to keep an eye on:

  • Pre-emptive customer service. With today’s connected smart devices, it’s possible that a future customer service engagement is not initiated by the consumer, but by a device. For example, if your connected refrigerator detects a fault, it can contact customer service without consumer input. Customer support can then research the potential error and proactively get in touch with the consumer, alerting them to the issue and scheduling a technician to come service the device. These connected devices can continuously collect data that informs the brand’s contact center for additional analysis and insight.
  • Specialized human agents. As AI-enabled technologies become smarter and address more complex issues, human agents will require additional training on the most complex customer interactions, perhaps specializing in a specific product or service area. This could also include analytics training that would provide the skills needed to make sense of all the data captured in the CRM.
  • New workforce makeup. When chatbots and other AI-assisted technologies become mainstream, different skills will be needed from human agents. Agent work will become more empathetic and relationship-oriented and will involve a higher level of upselling and cross-selling to existing customers. Empathetic agents will need analytical thinking, communications skills, and problem-solving acumen.

While AI-enabled technologies have made considerable inroads into the contact center, there is still much advancement that can be made in using these solutions to perfect customer service.

SHARE

Nuno Brito

Nuno Brito

Nuno started writing for blogs in 2009. He has extensive experience researching and writing about contact center best practices and customer experience. Nuno loves emerging tech trends such as AI and machine learning. When he's not having fun exploring content writing, you can find him at the beach.