The Recipe for Trusting Virtual Assistants
Consumer-facing AI is meant to make consumers’ lives easier and more efficient. But if they don’t trust these systems to help them, the value of this technology is moot. In an effort to create the perfect “recipe” for building trust in everyday virtual assistants, Phil Gray, EVP of Corporate Development at Interactions identifies four key ingredients
As AI solutions and virtual assistants make their mark on people and businesses, we’re learning there’s a hesitation to trust the technology with big decisions – like where to invest money, how much to pay employees, or what treatment to provide a sick patient. Reasonably so. However, when we look beyond these high-stakes applications, there’s good reason to trust many AI-powered virtual assistants that are in use by millions of people each day.
When the stakes aren’t so high, virtual assistants are consistently touching the lives of people in meaningful ways. Building trust in this realm means consumers can lean on virtual assistants to handle tedious daily tasks that otherwise consume time and attention. Consider the millions of Alexa devices in homes across the U.S., or the thousands of companies that have looked to intelligent virtual assistants to handle customer service inquiries, both online and over the phone.
Over the past few years, I’ve worked closely with brands, which have launched AI-powered intelligent virtual assistants to their customers. This consumer-facing technology is meant to enhance the brand’s customer experience, and make lives easier and more efficient. But, if consumers don’t trust these systems to help them, or they do not perform well, the value of these products and this technology is moot. In an effort to create the perfect “recipe” for building trust in everyday virtual assistants, I’ve identified four key ingredients below:
First, make clear what your virtual assistant can and cannot do for the end user, while also ensuring that the technology is up to the task without placing additional burden on the customer. People lose trust when the capabilities of an AI-powered system are “oversold”. They also lose trust when the AI system keeps them from reaching an actual person when necessary.
Ease of Use
The virtual assistant must be easy to use to gain consumers’ trust. Conversational virtual assistants need to move beyond short utterances and isolated phrases, and reach a level of understanding that allows them to lead open conversations that mimic the natural flow of human conversation. For instance, if someone calls a customer care virtual assistant and is forced to speak in “key words” and often has to repeat themselves because the system had difficulty understanding, then they will quickly lose trust and confidence in the system.
Perhaps most important, the system must be intelligent. This could mean that the virtual assistant is aware of who you are, and has context for your current relationship with a business. For instance, if you receive a notification to update your credit card via SMS, you should be able to contact that business directly back on SMS to fulfill the request. Alternatively, if you choose to call instead of using SMS, the system should still anticipate that you are calling to update your card.
Another example of demonstrating intelligence is for a virtual assistant to proactively recognize that your recent order was put on backorder. When you call to address the issue, the system immediately asks if that’s the reason you’re calling. Additionally, the system should leverage information it already has about you, such as a mobile number or email address, and use it to personalize the experience and minimize redundant requests. Finally, truly intelligent systems should realize when it can’t help, and quickly redirect you to a person who can.
Lastly, for consumers to rely on and trust virtual assistants, the system must be accessible on the channels they use. This could range from a mobile phone call, to Facebook Messenger, Amazon Echo, SMS, and beyond. Studies show that there are vast differences in channel preference based upon age demographics, so if you have a diverse customer base you may need to support all of these channels and more. Furthermore, the system should not force you to start over if you switch channels. If a customer starts on one channel such as a voice phone call, but gets interrupted and otherwise decides to connect with you on another channel such as SMS, the system should seamlessly link the conversation, so the customer does not have to start over.
As each of these ingredients find their way into everyday virtual assistants, trust and adoption in the technology will grow. In doing so, brands will reap the benefits of their investment, quickly contributing to improved customer experiences and increased customer satisfaction. Eventually, the consumer mindset will shift from skepticism to appreciation, and the brand will benefit in the long run.