#contactcenterworld, @Synthetix_Ltd

AI software can play a powerful role when engaging with customers online. What does this mean for the Contact Centre? Neldi Rautenbach, Head of Marketing with Synthetix explains more in the following article.

The role of Artificial Intelligence in Customer Engagement

AI software can play a powerful role when engaging with customers online.

Every few years we eagerly anticipate the next big thing in the realm of Artificial Intelligence. Like computing, it’s hard to distinguish between classifications for AI, sometimes referred to as machine intelligence, machine learning, deep learning, or cognitive learning.

Artificial intelligence, in fact, is a broad term. In other words, machine learning, natural language processing, robotics, and computational intelligence all refer to a wide variety of algorithms and methodologies under AI.


Virtual Agent, Chatbot, Intelligent Virtual Agent, Virtual Assistant, Personal Virtual Assistant, or Conversational bot?

Google, IBM, Facebook and Microsoft have all made announcements this year about Intelligent Virtual Assistance. Advances in text analysis, natural language processing and automation are enabling many queries and tasks to be resolved without any human interaction. Gartner states that this combination of speech and Virtual Assistant technologies provides a strong customer service proposition. Offering more than just search, an Intelligent Virtual Assistant should enrich the quality of the customer experience and assist the customer throughout their online interaction.

As we are one of the pioneers of Virtual Agents as a concept - building our first versions in 2001 - it’s important to clarify the differences between Personal Virtual Assistants, Virtual Agents and Chatbots to understand their value to a business as AI has led to erroneous expectations, especially in terms of customer engagement.

Personal Virtual Assistants

Google Now, Siri, Alexa, Cortana – There’s a new voice assistant in town and they are having a showdown. The hint is in the name.

Personal Virtual Assistants are friendly, helpful, personalised, can perform tasks and answer questions. This type of AI has been designed to be helpful and highly entertaining. They can have a voice, a face or Avatar.

In a personal capacity, there isn’t a big downside if the system misunderstands a command or responds with the wrong answer, but in a business sense, this would be critical.

Different accents, dialect, synonyms, or contradictory meanings can have speech recognition AI struggle to respond accurately.

Virtual Agents

Synthetix were creating virtual agents for companies like the BBC, Vauxhall, and Ford, using AI techniques comparable to Siri, Cortana and Google Now, before they became the norm. It’s true, our NLP (Natural Language Processing) technology has answered millions of customer questions by making sense of naturally (and sometimes, badly) worded questions.

The difference between a Virtual Agent and a Virtual Assistant however, is that a Virtual Agent is goal orientated, designed to perform detailed tasks or queries, much like a virtual employee with a specific job description. Like Virtual Assistants, Virtual Agents should be friendly, personalised, and conversational.

Although Virtual Agents are designed to mimic human learning, understand user intent, and to perform tasks that would normally require human intelligence such as speech recognition, NLP, decision-making or language translation, it should not be posed to be a real human. Automation can be a great customer service tool, however in a recent study by myclever, the ‘personality’ of a Virtual Agent barely registered with consumers – AI needs to be smart and effective first - witty and personable second.

Chatbots

A chatbot (also known as a talkbot, chatterbot, Bot, chatterbox, Artificial Conversational Entity) is a piece of software that conducts a dialog via auditory or textual methods. Chatbot programs are often designed to credibly simulate how a human would behave as a conversational partner, thereby passing the Turing test. However, simulating a human being doesn’t make for a more compelling customer service experience as Microsoft discovered earlier this year.

Chatbots live within in a social domain or messenger platform and are popular amongst millennials whom prefer these channels for customer support, shopping or to interact with a brand. Asking a chatbot a simple question will have it deliver a helpful response or carry out a simple administrative task such as making a reservation at a restaurant.

Chatbots are only as intelligent as they are programmed to be. Some conversational programs simply scan for keywords within the input, pulling a reply matched to the most similar wording pattern or matching keywords from a database. Other chatbots use AI techniques such as Natural Language Processing, rules, and decision trees to ‘understand’ intent and commands. It can be self-learning making it smarter through each conversation. In fact, the Facebook Messenger chatbot is learning increasingly more about the language patterns and questions asked by users globally.

Chatbots are also popular in smartphone app deployments. A popular category of smartphone app that relies on a chatbot is the dating sim or romancebot category. The 36 You Games app "Boyfriend Maker" and WET Productions Inc.'s "My Virtual Boyfriend" are popular examples.

Will Virtual Assistance replace the contact centre?

Imagine one affordable, intelligent point of contact that can efficiently respond to a limitless number of queries with relevant information, any day, any time. Why is it then that Virtual Agents haven’t replaced the contact centre?

None of the technologies mentioned is a ‘plug and play option. Consider the potential combinations of colloquialisms, questions and meanings that exists within everyday language. Although it advances every day, AI technology is not adequately advanced to understand all the questions put to them. In the land of customer service, every interaction matters and any investments in these types of technologies will falter if the user experience comes up short.

Programming the right features are key. Although these technologies can be powerful in engaging customers’ careful consideration should be taken for other methods such as an intelligent self-service tool powered by a knowledge-base might be quicker to deploy and prove effective.

Determine which technology is right for you – To bot or not?

#contactcenterworld, @Synthetix_Ltd