How conversational AI has evolved, and why it is rapidly redefining how customers interact with businesses.
Imagine if customers could talk to a business whenever they need something, not just during working hours, not just when a human is available, and not just through protracted processes.
getty
People have conversations in person or via text, phone, video chats—the obstacles have never been lower. But when it comes to interacting with a business, points of friction are more numerous, with customers often forced to peruse bespoke websites and apps, deal with phone trees, wait for email responses, or accommodate the limited availability of human staff during business hours.
That needn’t be the case: it’s time for organizations to reevaluate the role of conversational AI in customer experiences.
In many cases, it’s already become obvious that simple speech can replace high-friction processes. Some readers may recall when programming a VCR to record a TV show was a complicated technical feat that inspired comparisons to brain surgery, for example. Today, streaming content and cloud storage have rendered physical home media mostly obsolete, and recording something is as simple as a spoken command to Google Assistant.
But “conversational AI” encompasses much more than voice interfaces that offer only a few functions or work only if the user rigidly adheres to specific spoken prompts. Take the 2021 Toyota Sienna. When drivers need assistance, they don’t have to rifle through a physical manual or even phone a representative—they can simply speak to a voice-activated digital manual.
Customers are always going to demand less friction, greater convenience, and faster resolutions—and now is the time for businesses to stay ahead of that curve and reimagine their customer experience by exploring what conversational AI can do.
Likewise, when customers call into a contact center, they don’t need to dial in passwords to authenticate—their voice can do the job, securely bringing up all their account information with just a few words. And as unemployment claims spiked throughout COVID-19, state governments like Illinois did not force citizens to wait until limited human resources could field all inquiries; rather, the state supplemented its human workforce with virtual agents that could understand and respond to citizens’ questions, contributing to a 11.5x year-over-year increase in claims processed during the early months of the pandemic.
Just as people can talk to one another as required, imagine if customers could talk to a business whenever they need something, not just during working hours, not just when a human is available, and not just through protracted processes. Conversational AI is making this kind of experience a reality, not just an ambition. In this article, we’ll explore how conversational AI has evolved and why it is rapidly redefining how customers interact with businesses.
How conversational AI technologies have matured
Conversational AI is making better customer experiences a reality, not just an ambition.
Google Cloud
Compared to just a few years ago, conversational AI technologies have advanced in three key ways that make them significantly more accessible for businesses:
1. Interactions are natural.
Until recently, voice interfaces relied on specific spoken commands that often led to stilted, artificial interactions. That’s no longer the case because of technologies like BERT, a natural language processing model developed by Google, deployed across Search and other services, and open-sourced for anyone to use.
BERT can interpret words in context, discerning, for example, whether a statement with the word “raptor” refers to a dinosaur, a bird such as a hawk, or an aircraft. Models like this comprehend human conversation much more precisely than their predecessors and serve as a foundation for more fluid and interactive conversations between humans and AI. BERT can be easily trained for domain-specific terms that exceed its out-of-box capabilities, making natural conversational experiences more accessible than ever for any business. This enables businesses to provide frictionless customer experience for any purpose a customer may have in mind.
2. Touch points for conversation are ubiquitous.
Voice interfaces were initially limited to specific services from single companies or a handful of partners. Over time, those experiences have expanded, largely as the companies behind those digital assistants opened application programming interfaces (APIs) that give third party developers access to voice technologies. These APIs are a big reason why assistants that started on smartphones from a few brands are now available on many devices from many brands.
Since then, basic APIs have given way to more advanced APIs and even more flexible and user-friendly services that make conversational AI accessible to businesses without long-running expertise or investments in machine learning. This democratization means businesses can create experiences across a range of touch points, from mobile apps to smart home devices to in-store kiosks, setting the stage for conversations between customers and businesses to be as ubiquitous as conversations between one person and another.
3. Conversational AI is scalable.
Thanks to the emergence of cloud services that make AI more accessible, conversational AI experiences are more scalable than ever, able to reach billions of people and devices. They’re also increasingly sustainable, because when businesses tap into high-performing cloud models instead of developing their own technologies from scratch, they also tap into vetting processes that have been tested in production, such as refinements to avoid bias.
But building conversational experiences requires specialized knowledge of a business, its use cases, its customers, and its previous interactions with customers. The same underlying technologies that make natural interaction between people and AI possible can also automatically parse documents and databases for this kind of information, all while preserving privacy and security. This can make it easier and faster to fine tune or build on top of models such as BERT, move these refined models into production, and scale conversational experiences to more customers.
Related video: Learn how call centers at Optus, Telus, and Verizon improved customers’ experiences and agent capabilities by gathering insights with the power of AI to delight their customers.
Start the conversation
With points of entry to match wherever businesses are in their AI adoption, conversational AI is primed to become a major means of customer interaction. Building the technology backbone for conversational AI has been prohibitively difficult and expensive, but organizations can now select from a wide range of services to accelerate their progress, from out-of-the-box virtual agents to APIs that serve as building blocks for new experiences.
Far from replacing human-to-human interaction, these technologies help extend business-customer interactions in new directions. They let customers reach a business whenever they need, using the same conversational approach regardless of whether they’re speaking to a human agent, an AI, or a combination in which employees and AI agents hand off the service call at different points. Customers are always going to demand less friction, greater convenience, and faster resolutions—and now is the time for businesses to stay ahead of that curve and reimagine their customer experience by exploring what conversational AI can do.
Read this next: Get the free report from Forrester to learn how businesses have saved costs and improved customer experiences with Contact Center AI.
