How AI is Changing Customer Service


How AI is Changing Customer Service

By Paula Bernier, Executive Editor, TMC  |  March 07, 2017

Artificial intelligence, which uses algorithms and rules-based logic to automate a wide variety of tasks within many verticals, is hot. It occupies the top spot in 2017 tech trend lists from such notable companies as Ericsson and Gartner.

Cognitive systems and AI adoption across a broad range of industries are likely to drive worldwide revenues from nearly $8 billion in 2016 to more than $47 billion in 2020, according to IDC. Automated customer service agents is one of the key AI areas that attracted investment last year, the research firm noted.

AI’s Role in Personalization

“From better purchase recommendations, to smarter customer service that predicts what a consumer is actually trying to do, AI promises to fundamentally transform entire businesses and industries,” said Scott Horn, CMO at customer engagement solution provider [24]7. “Investments in AI are paving the way for a personalized world where every interaction matters, and every company can benefit."

IBM Watson famously competed on and won Jeopardy! several years ago, and now this technology is used in many commercial applications. AI is also already present in such well-known applications and solutions as Amazon Alexa, Apple Siri, and Google Assistant, which leverage natural language processing to recognize voice commands; Facebook’s (News - Alert) deep learning facial recognition algorithm, which can instantly identify a person with nearly 98 percent accuracy; and Amazon, Netflix, and Spotify, all of which use machine learning to understand how each item in their catalogs relate to the other and each customer’s preferences.

The Democratization of AI

However, until recently, AI solutions were beyond the reach of most companies. In introducing Einstein a few months ago, Salesforce explained that’s because implementing AI requires bringing together massive and diverse data sets, creating specialized predictive models to extract value from the data and continuously learn from it, having infrastructure and DevOps support to keep all of these processes up and running, and doing complex integrations.

But now, new tools like Salesforce Einstein, Amazon AI, and the Facebook Messenger Platform make AI more accessible to more organizations.

“Companies like Salesforce, which have committed more than $4 billion specifically to this effort with its latest product Einstein, will lead the way,” said Justin Thompson, vice president of strategy for customer experience software and research firm MaritzCX. “Providing intelligent insights and recommended actions against customer sentiment, engagement, and future behavior will be first. Predefined outcomes related to conversion and retention will be the initial goal, and AI will move beyond telling what drives it to recommend how and when to drive it.”

Salesforce Einstein, which the company calls the world’s smartest CRM, was designed to allow any company to deliver more predictive and personalized customer experiences across commerce, sales, service, marketing, and more. Einstein was introduced in September of 2016. Two months later Amazon Web Services (News - Alert) announced three AI services that it said will make it easy for developers to build apps that can understand natural language; turn text into lifelike speech; have conversations using voice or text; analyze images; and recognize faces, objects, and scenes.

These two announcements followed the introduction several months ago of the Facebook Messenger Platform. This platform already has given rise to 30,000 chatbots from various companies that Facebook’s 900 million monthly users can employ to pay for meals and shop online, noted Rajeev Shrivastava, chief strategy officer at cloud contact center software provider inContact, a NICE company. 

“Customer service from humans is inherently unpredictable,” said Anthony Pappas, president of brand marketing and customer experience at customer experience strategy and planning company DMI. “Not every company can significantly invest in training and customer service personnel. However, AI provides companies a more scalable and affordable option for customer service – especially in the digital age.”

A Bit More About Bots

Pappas said the hospitality and travel industries have been doing trials of AI, and both Expedia (News - Alert) and airline KLM have implemented AI-powered chatbots to help with customer service. Meanwhile, DMI has been working with travel distribution company Zumata to integrate natural processing into a chatbot.

“The app uses IBM Watson’s artificial intelligence technology to process customer hotel search requests, resulting in a new level of personalization,” Pappas said of the Zumata project.

Meanwhile, Google Ventures-based Signpost has introduced a CRM bot called Mia. When someone calls or emails, Mia collects their information and sends them special offers. She can also request that satisfied customers post reviews and unhappy customers are flagged.

Vodafone in Qatar also recently gave life to a bot. Named Hani, and powered by [24]7 technology, it invites website visitors to type in their questions, and it maps those questions to the correct answers, regardless of how the questions are phrased.

In December, a company called Workbot introduced Workbot for Slack, which has the ability to process and interpret URLs from supported apps like Salesforce and GitHub. When a link from a supported app is sent in Slack, Workbot can automatically display relevant information from the app directly in chat and allow users to take immediate action. Additionally, customers will be able to manage their Workato recipes directly from Workbot using a new app called Workato Admin.

And Chyme and Kore are both developing bot technology specifically for the workplace, explained Eugenia Corrales, senior vice president of product at business phone system, contact center, and UC solution provider ShoreTel. Chyme’s bots help workers in contact centers quickly determine the customer’s problem by identifying patterns of previous customer behavior, while Kore’s Smart Bots can help retailers use a simple messaging platform to automatically figure out if certain merchandise is in stock.

Forty percent of consumers are open to interacting with chatbots, and nearly a third of them actually prefer it over working with a person over the phone or via email, according to a new customer experience study by [24]7. Meanwhile, 44 percent of the respondents to the 2016 Aspect Consumer Experience Index said that if a company could get it right, they would prefer to use a chatbot or an automated assistant for customer service interactions. That’s up 4 points from Aspect’s 2015 survey.

“The accuracy and interaction quality of intelligent assistants such as Apple’s (News - Alert) Siri and Amazon’s Echo are driving consumer interest for similar experiences with the brands they do business with,” said Joe Gagnon, Aspect’s chief customer strategy officer. “This is no longer just a way for companies to reduce costs by handing simple and repetitive queries over to automated assistants. This is an opportunity for companies to satisfy a growing customer demand.”

In addition to waiting for customers to call on them for tasks, AI bots can be used to help brands identify and respond in the appropriate tone to customer comments sent via email or posted on social media, noted Frank Pettinato, senior vice president of multichannel customer engagement at customer service outsourcing firm Telerx, a subsidiary of pharmaceutical company Merck. This kind of thing, he added, can enable businesses to respond to comments a whole lot faster and more accurately than they could otherwise.

 “The goal here is to cast a wide net,” he said, but without letting on to consumers that they are receiving an automated response.

The technology to do that already exists today, he added, and Telerx has been working with its clients to use Facebook Messenger technology to deliver automated responses. But that’s not very sophisticated today, he said, adding that Telerx will be expanding its AI bot pilots in 2017 to enable higher-level functionality.

In 2016 AI was applied to solve known problems, said Abdul Razack, senior vice president and head of platforms, big data, and analytics at technology services and consulting company Infosys.

“As we move forward,” Razack said, “we will start leveraging AI to gain greater insights into ongoing problems that we didn’t even know existed.”

That could relate to customer service or, really, any area.

For example, Gartner has noted that AI and machine learning could be used in banking to model current real-time transactions as well as predictive models of transactions based on their likelihood of being fraudulent.

Analyst Rob Enderle recently blogged about how AI could in the future help prevent people from sharing content that could make them look bad or even lead to their dismissal at work. It could do that by recognizing potentially problematic communications and advising people not to share them, he said. An AI-powered solution could even help formulate something more appropriate.

And IBM has multiple AI initiatives, one of which involves working with MIT (News - Alert) to develop systems that emulate the human ability to understand and integrate inputs from multiple sources of audio and visual information into a detailed computer representation of the world. This could be applied to various industry verticals including education, entertainment, and health care.

“AI is currently doing a great job of solving bounded problems like chess. The rules are well defined, and there is a large, but set number of permutations or moves available,” said Colin Earl, CEO of business software company Agiloft. “A less bounded problem is a self-driving car near a primary school in traffic. AI will continue to expand incrementally into more unbounded applications, but it will take time for deep learning to catch up to human quality in these complex scenarios. In the next year, we will see AI creep more closely into our daily lives to automate more – online and real-life purchases, and more.”

Edited by Alicia Young
Get stories like this delivered straight to your inbox. [Free eNews Subscription]
blog comments powered by Disqus