Building a better listener: LivePerson’s Rob LoCascio wants to design our automated future
LivePerson Founder & CEO Rob LoCascio listens to people. It’s what enabled him to successfully lead his company through a number of disruptions over the last 26 years, from the dot-com bust and 9/11 to our current pandemic. His entire New York City-based company is built around the concept of listening to consumers and building software that knows how to understand what they want.
In the aftermath of 9/11, LoCascio witnessed other business leaders attempt to rally their troops with rousing speeches about triumph over adversity. However, in listening to his people, he realized they needed time to grieve and to take stock of what was really important in life.
“We have to lead with empathy,” he says, referring to lessons learned during that difficult time. “I learned there that if you try to make people feel good about the experience…it rings false. We’re people…I remember thinking, I’ve just got to listen to people, hear them out.” It’s a principle that drives not only his management style, but also LivePerson’s entire business model.
LoCascio joined Scarlet Fu, Bloomberg Television and Bloomberg QuickTake host, for an online conversation on Tuesday, November 30, as part of the Cornell Tech @ Bloomberg Speaker Series. He shared the story of LivePerson’s transformation from a web chat provider to an AI innovator, and his bold vision for a future of conversational experiences mediated by bots that can really listen.
Watch the full discussion:
From the kiosk to the cloud
After graduating, LoCascio spent six months working for someone else, an experience that left him in tears when he was “fired by fax.” He was laid off during a downsizing, and learned he’d be let go when he spotted the order for his termination coming through the fax machine it was his job to run.
“If I’m going to live my purpose, live my destiny — I can’t have someone else dictating that for me.”
He was devastated, but this negative experience drove LoCascio to found his first startup, which placed informational kiosks on college campuses that would give students information and serve them advertising. The startup never took off, but when a big customer asked him if his company “did websites,” he realized that the Internet was going to be a bigger deal than he and his peers had anticipated. The Internet would allow students to get all the information his kiosk offered directly on their personal computers — no need to schlep to the student union.
“I said, ‘Yeah, we do websites,’” replied LoCascio. It was a fib at the time, but making good on this statement led him to the understanding that the Internet was the future, and more broadly, that a good, scalable distribution model was more important than content. LoCascio wound down the kiosk company and founded LivePerson as an attempt to capitalize on the Internet as it began to penetrate everyday life.
“I moved to New York City. I started building websites just to make some money to not starve. And what I started to realize is that there was a big piece missing. I went to Dell’s website because I needed to buy a computer. I’m on the website and I couldn’t get [my questions] answered. So, I hung up my dial-up connection, got on a call, and I’m talking about the website to the agent. I’m like, ‘I’m on the website, I can’t figure the thing out,’ and the person’s like, ‘I don’t have anything to do with the website, re-explain everything you want.’”
It was an epiphany. If the Internet enabled the ability to message back and forth in real time, why should someone have to get offline and make a phone call to an agent who has no context around what they were trying to do? What if there was a way to insert that agent into the web experience, so that they could solve the shopper’s problem in the same convenient environment? Plus, this way, the agent would already have context around the goal the shopper was looking to achieve.
We’ve all seen the ubiquitous little chat windows that appear in the lower right corner of countless websites, encouraging us to “Talk to an expert,” or “Chat with a customer service agent.” But in the mid-90s, this was a novel application of brand-new tech.
“I invented web chat for brands,” says LoCascio proudly. “That’s what I founded, and that’s what I created for the early version of the Internet.”
Pivoting to bots
Live chat quickly became one of the foundational technologies of e-commerce. LoCascio filed the first patent on it in 1999, raised $3 million, and took LivePerson public a year later. This technology allowed brands to build customer service conversations into their web environments.
In 2015, the company executed a major shift away from facilitating human agents via the browser, toward building mobile messaging applications, including for social media platforms like Facebook and Twitter, and branded apps like the ones customers use to remotely order food at Chipotle or book travel with Delta.
Next LivePerson launched a conversational AI platform, now known as the Conversational Cloud, that enables brands to deploy AI-powered virtual assistants that handle simple customer queries and transfer customers with more complex issues to human agents. This platform can be integrated with customer relationship management (CRM) software, enterprise resource planning (ERP) systems, and other existing applications.
“We have an extraordinary data set as a company,” says LoCascio. “Every month, we generate about 80 million conversations on our platform. If you’ve got this type of data, you have the ability to do pretty amazing things in the technology world, especially with AI.”
Conversational AI needs tons of training data to continually teach chatbots how to better handle customer queries. By labeling which conversational patterns led to positive or negative outcomes, LivePerson is able to train an AI to pursue optimal conversational strategies that are the most likely to result in a happy customer.
This vault of conversational data going back some 26 years gives LivePerson an edge, argues LoCascio, even over AI behemoths like Amazon. Amazon Alexa, he says, is not true conversational AI, as it is limited to a set of predefined commands, like “Turn on the lights,” and “Play music.”
In contrast, the conversational AI that LivePerson is perfecting allows for complex, nested conversations that you might have with a real person. When speaking with the most advanced conversational AI, you can ask questions within questions. The AI can understand your intent and the emotional content of your language. Despite recent advances, the world’s most advanced conversational AI might struggle in keeping up with a conversation among toddlers, but the tech is getting better all the time. As LivePerson’s bots continue to ingest more training data, and learn from their own past conversations, they get even smarter.
The masters of conversational design
One of the most common concerns that tech ethicists have around the automation revolution is that many workers, like customer service representatives (CSRs), are going to lose their jobs. LoCascio paints a brighter vision of the future, where these displaced workers will have the opportunity to be trained as bot builders, or conversational designers.
“Four or five years ago, everyone thought bot builders should be tech people or data scientists,” says LoCascio. But according to his research, the world is going to need millions of these conversational designers to help us navigate our automated future. Many futurists predict that fundamental human experiences are going to soon be augmented by sensors, actuators, speakers, and other Internet of Things (IoT) components. These components will gather data and give us the ability to reshape our physical environments with seamless, voice-controlled interfaces. For humans to ever feel comfortable using them, these interfaces will need to be intuitive and able to understand the idiosyncrasies of human language, reason, and emotion.
“Think about this whole idea of the metaverse,” says LoCascio. “Let’s say that exists. There will be virtual worlds. There will be conversational AI going on inside there. Brands will be sitting there with their avatars talking to you about something. And that’s going to need a conversational designer. It’s not going to be a human typing or talking through the avatar. It’ll be an AI… Everything’s going to be powered by these conversational experiences.”
LoCascio’s betting that because CSRs have a high emotional and verbal intelligence forged by many years of listening to people’s stories and helping them solve their problems, they’ll be able to craft richer, more satisfying conversational experiences. LivePerson has built a tool set for non-technical people to learn this emerging craft.
“People working in contact centers are extraordinary people,” says LoCascio. “They actually love people and they want to help people… They understand how to converse in a way where they’re not looking at you. They’re not a friend of yours, but they’re trying to make you comfortable. They understand how to do these conversations at a very high quality.”
According to LoCascio, we need to start thinking of CSRs, not as interchangeable cogs in an industry where 40% turnover is the norm, but as valued masters of conversational experience. CSRs can excel as conversational designers because, at a fundamental level, they know how to listen. This will be an invaluable skill as our automated, augmented, and virtual worlds emerge. If LoCascio is right, we may well find that CSRs who build bots will also build the next iteration of the Internet.
This article was originally published by Tech at Bloomberg.