IBM’s Watson computer has taken on a new job — that of customer service agent — as Big Blue puts its Jeopardy-playing computer into a new role. This will also be the first time IBM delivers Watson completely as a service, instead of as a highly customized software product for select customers in the medical and financial services field. But as Watson expands its role it may invite more comparisons to Siri, Apple’s natural language processing assistant.
Tuesday at the IBM Smarter Commerce Global Summit in Nashville, Tenn., IBM plans to launch the Watson Engagement Advisor, aimed at helping consumer brands better recommend products to customers and provide better customer service at scale. Yes, that’s right. This is a technology that can diagnose cancer will be used to help sell people more products.
Why we need Watson-level AI for customer service today.
Still, customer service is a legitimate and complicated problem, especially in an era where social media meets our desire for a personalized and instant response to any inquiry or service issue. Firms have to engage with customers via phone calls, tweets, Yelp, Facebook posts and for all I know, angry letters. And many of those customers using new media don’t want to wait for a response. Companies that can offer good service quickly in a variety of mediums have an advantage. And Watson would allow them to do this at scale. Imagine offering Ritz Carlton service at Holiday Inn prices.
Brands who buy the Engagement Advisor software will get access to a much smarter virtual agent that can sift through massive amounts of information to respond to users’ questions quickly. As someone who was totally schooled at Jeopardy by Watson, I cannot emphasize enough how fast it is.
The IBM release notes that the Engagement Advisor software is designed to help existing customer service personnel answer questions quickly or it can be deployed via the brand’s mobile site where customers can interact with Watson directly. As IBM’s release says, “In one simple click, the solution’s “Ask Watson” feature will quickly address customers’ questions, offer advice to guide their purchase decisions, and troubleshoot their problems.”
It’s possible this will remind users of Siri, Apple’s chatty personal assistant on the iPhone and iPad. However, instead of being deployed on a device, Watson is embedded on a brand web site.
It can greet customers by name, however and offer to help them via a chat window on the company site or via a mobile push alert, that will appeal to people who want to tweet or text their customer care questions without having to stay focused on a single web page. In the ideal case Watson will have access to customer records plus the data stores it was trained on, and will be able to use both in giving a customer a recommendation or help.
The business of Watson is a big one
Instead of naming customers directly IBM writes that brands including ANZ, Celcom, IHS, Nielsen and Royal Bank of Canada are, “exploring how the Watson Engagement Advisor can help them engage with their customers.” This may be phrased this way because the initial pilot projects involving Watson require a lot of training of the computer before it becomes valuable. During Watson’s “apprenticeship period,” IBM in some cases hasn’t charged clients, or charges them lower rates.
But it’s no secret that Watson is a big business bet for IBM. At last year’s Structure conference, Dan Cerutti, IBM’s VP of Watson Commercialization, explained IBM’s ambitions for Watson, including delivering the machine as a service over more and more devices. IBM sees Watson as a new type of computing and plans to build out new business models to support it, as Cerutti detailed in our chat almost a year ago.
Along the way Watson not only impresses with its ability to filter through reams of data to correctly answer a natural language question, it also has been able to do this as it shrinks in size. Since its television debut, Watson has seen a 240 percent improvement in system performance, and a reduction in physical requirements by 75 percent. The whole system can now be run on a server that takes up the size of four pizza boxes from a giant machine that took up an entire bedroom. Smart, svelte and delivered as a service. Get ready to meet Watson in more roles and in some surprising places.
Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
- The state of cross-platform media measurement
- The importance of putting the U and I in visualization
- A near-term outlook for big data