Artificial Intelligence and the Customer Journey

Article Featured Image

Virtual assistants—like Alexa and Siri—have gained traction in the consumer environment and are likely to continue to spur an uptick in voice-activated search and other customer activities, from purchases to providing feedback to seeking service and support. In short, voice technology can aid consumers at many points along the customer journey.

“Digital customer care is on a major upswing with the expectation that 85% of customer relationships with an enterprise will be managed without human contact by 2020, according to Gartner,” says Chris Bergh, CEO of DataKitchen Inc., a DataOps consultancy and platform provider. “Conversational AI is critical and plays an integral role in improving customer service, reducing agent-assisted service costs, and cutting down on fraud.”

Unfortunately, consumer expectations may outpace technological reality at this point, challenging both brands and tech vendors to meet those expectations.

How Customer Expectations Are Changing

Research by Adobe Analytics, based on a survey of 1,000 U.S. consumers, found that the most common voice activities (in 2018) included asking for music (70%) and updates on the weather (64%). Consumers were also using voice search for “asking fun questions” (53%), conducting an online search (47%), getting news reports (46%), doing basic research (35%) and seeking directions (34%). But emerging uses of voice technology were also identified:

  • Shopping for, or ordering, items: 30%
  • Food delivery or takeout: 17%
  • Flight/hotel research: 16%

A 2019 report from Microsoft suggests growth in the use of voice and digital assistants, with 72% of respondents indicating that they had used a digital assistant in the past six months and more than half saying that they believe digital assistants will help them make retail purchases in the next five years.

These consumer purchase applications of voice technology represent opportunities for retailers and others to capture and maintain market share, improve engagement with existing customers, and streamline operations. Artificial intelligence (AI) is transforming the customer journey. That’s especially true when it comes to speech technologies (specifically conversational AI). Savvy brands and service providers are recognizing this and seeking ways to automate and voice-optimize their processes. Their efforts, though, may not be keeping pace with consumer expectations.

Arle Lommel is senior analyst at CSA Research and a former researcher at the German Research Center for Artificial Intelligence (DFKI) in Berlin, where he focused on machine translation. Lommel points to three primary areas that stand out in terms of customer expectations, based on CSA Research’s perspective:

  1. An expectation that conversational agents will speak the languages and dialects of customers and honor their local customs and requirements.This becomes more complex than it may initially appear. For instance, he says: “English-only agents may work in parts of the United States, but may vary regionally—so apps trained in the Northeast may not meet the dialectal or regional requirements of the South. Globally, you cannot expect to release these outside of the U.S. and find acceptance.”
  2. An expectation that virtual agents will be able to recall and respond based on prior conversations—the notion of “state,” a term used by developers to refer to context.Today’s voice technology largely applies to conversations taking place now, without the benefit of previous conversations to enhance meaning, Lommel notes. “If you ask Siri for directions to Chattanooga and five minutes later ask, ‘What can I eat when I get there?’, it will have no idea of where ‘there’ is. This lack of state keeps agents from being truly useful in many of the areas we might want them to help us. To some extent, technologies such as IFTTT [if this, then that] try to work around this, but until agents have a robust way to model and store state securely and confidentially, they will remain fundamentally limited.” Add to this the reasonable expectations that customers may have that all of their previous interactions with the company will be accessible to, and understood by, these agents—verbal, written, in-person, digital, etc.
  3. An expectation that virtual agents can deal with extemporaneous speech in a variety of domains. “The ability to deal with a wide variety of unexpected behavior in a robust fashion will be critical if developers are to prevent agents from being tricked, hacked, or used for nefarious purposes,” says Lommel. He provides the example of a virtual casino agent not programmed to reject bets with a negative number as the value, allowing users to game it to make big winnings—fortunately, not in real money.

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues
Related Articles

Transformation in the Contact Center

Customer service has moved from call centers to a multichannel and now omnichannel customer experience approach, with cloud-based communications, well-designed IVR, and powerful reporting capabilities thanks to artificial intelligence (AI), chatbots, natural language processing (NLP), and more.

Higher Learning: AI, ML, and Speech Tech in Academia

Academics are at the forefront of the biggest AI changes in the industry. Learn more about the latest breakthroughs, the questions researchers are trying to answer, and the challenges they face.

The Key to Business Growth: Using AI to Break Barriers in Sales Conversations

Businesses let hundreds of millions of dollars slip through their fingertips because they fail to recognize opportunities embedded in conversational nuance. But AI can help.

Improvements in Agent Performance and Customer Insights Drive Expanded Partnership Between CallMiner and Sitel Group

Sitel Insights leverages CallMiner's AI-fueled interaction analytics to provide customer experience analytics for major brands and Sitel's workforce.