The age of conversational AI is here, altering the landscape of traditional Web and mobile applications. What factors are enabling the success of these systems and what benefits could they offer your enterprise?
The graphical user interface and mouse propelled the personal computer into the hands of nontechnical consumers with its user-friendly design. Smartphones and tablets flooded the market with easy-to-navigate apps, controlled not by a keyboard and mouse, but by the swipe of a finger, making intuitive interfaces commonplace. Whether it’s a mouse or a finger, each iteration succeeded by aligning technology to our more human instincts. And what’s more human than having a conversation?
Enter the era of conversational systems. These systems, more commonly referred to as conversational AI (artificial intelligence), combine natural language processing, AI, and machine learning to understand and respond to free-form text or voice—in an engaging and personalized manner. With increased access to cloud computing and sophisticated algorithms, more companies can implement conversational AI solutions that can turn technology adoption into a conversation rather than an exercise in mastering a user interface. Many expect conversational AI to therefore alter the landscape of traditional Web and mobile applications—first by augmenting them and then, at their finest, by displacing them.
The appeal of conversational AI isn’t going unnoticed. Across nearly every industry, conversational systems can be found in our homes, cars, call centers, banks, and hospitals—and the use cases are growing. By one estimate, the global conversational AI market is expected to grow from US$4.2 billion in 2019 to US$15.7 billion by 2024 (with a 30.2 percent compound annual growth rate).1
Two key technology trends seem to be making this growth possible:
- The proliferation of messaging platforms (for example, WeChat) create myriad opportunities for businesses to interact with people through well-designed chatbots. Messaging platforms are becoming so intertwined in our lives that the four largest messaging platforms have more active users than the four largest social media platforms (4.1 billion versus 3.4 billion).2
- The commoditization of speech-based assistants, such as Google Home and Amazon Alexa, paves the way for companies to develop their own speech-based assistants and/or affordably overlay others’ conversational systems onto their platforms.
In the first of our five-part series on conversational AI (see the sidebar, “A five-part series on conversational AI”), we explore how these systems can enhance customer engagement, workforce operations, and business partner integrations.
Three reasons to integrate conversational AI into your enterprise
Though conversational AI has existed for over a decade, the use cases and applications continue to become more sophisticated and gain traction in a variety of areas. Further, as computing costs continue to decrease while capabilities expand, personalization at scale has likely never been more attainable. three areas where conversational AI can flourish:
- Integrating the user experience: There are over 26 billion smart devices in circulation across the world.3 Embedded with Internet of Things (IoT) sensors, their power lies in their connected nature. However, toggling from one app to another to control any variety of devices can make the experience feel disjointed. In response, voice assistants are becoming the new interface for interacting across applications and devices—eliminating the need to ever directly access any individual enterprise’s application.4 Many of us already use home assistants to help us play music, listen to a weather report, and dim our smart lights, but this could be just the beginning. For instance, Pillo, an in-home health care technology provider, is deploying voice assistants to manage patient health. Powered by a voice interface, Pillo uses machine learning, facial recognition, and video conferencing to offer services compliant with the Health Insurance Portability and Accountability Act (HIPAA). These services can be paired with smartphones and wearables to administer medications, answer medical questions, and facilitate video conference appointments with health care professionals.5
- Optimizing the repetitive, with a personal touch: From e-commerce sites to self-service kiosks to call centers and help desks, people commonly engage with these resources to tackle any number of repetitive tasks. It can range from a regularly replenished micro sale of things such as detergents and lawn fertilizers, to seeking answers and support to frequently asked questions. While robotic process automation (RPA) and one-touch ordering buttons are already transforming many of these tasks, they don’t always provide the most customer or worker-centric experience. However, finely tuned conversational systems can. In fact, younger generations seem to be gravitating toward accessing information through chatbots, with 70 percent of millennials reporting a positive experience after using them.6
- Innovating through conversation: Conversational systems are, by nature, an unstructured way of communicating. This means every session provides an opportunity to gather more information for algorithmic improvement and creating relevant product solutions that better align to the user’s objectives. In addition, conversational systems are continually being redeployed to solve new business issues. For instance, the automotive industry is piloting the use of voice biometrics to unlock personal information such as navigation history and recent phone calls.7
Building conversational AI for your organization
- Align the goal to the need: Data sets, though expanding every day, are still finite. In this sense, consider starting with well-defined goals for your conversational systems and avoid one-size-fits-all solutions. In fact, many of these systems perform best under strict guidelines. Expecting a customer service chatbot for an online grocer probably isn’t best suited (or necessary) for providing stock portfolio advice.
- Remember the human: By nature, AI solutions are “smart” in a narrow sense.8 That is, they are very good at solving specific problems but still fail considerably at more human tasks such as critical thinking and empathy. Focusing on the need at hand, conversational AI can solve repetitive tasks in a more human-centric manner. However, a health care benefits chatbot, for instance, would probably not be the most qualified “person” to give a sensitive message denying benefits. In these cases, human understanding and intuition cannot be replicated.
- Put ethics at the forefront of your conversations: It’s hard to discuss AI without thinking of the ethical implications. And as mimicking human conversation only improves, the issue is only amplified. Designers of conversational systems can best serve their stakeholders by being transparent about the fact that they are interacting with a conversational system. Further, designers should consider workforce displacement and how they can implement conversational AI while still providing opportunities for their employees to grow in their careers.9
Indeed, these systems are at the cusp of potentially changing the way humans interact with machines.
As we continue on the conversational AI journey, more advanced applications, use cases, and paradigms will likely evolve, resulting in enhanced productivity for businesses and more personalized and accessible experiences for end-users.