For modern companies, one of the key challenges is figuring out how to become “omnichannels” -- or operations where all services (online, phone, brick and mortar) are seamlessly integrated together.
An important step toward that integration is using chatbots, says Joseph Falcone, medical digital solutions lead at pharmaceutical company Bristol-Myers Squibb.
Falcone -- who was in Austin this week for a seminar on artificial intelligence -- says businesses are trying to improve the chatbot experience and make them integral to their operations. The seminar was part of the Growth, Innovation and Leadership: North America event, hosted by technology research and consulting firm Frost & Sullivan.
Chatbots, programs mostly used online that simulate human conversations, have existed for years with mixed results. But as businesses look to improve artificial intelligence capabilities, company leaders at the seminar said chatbots could play a significant role.
“What AI and digital are really about is becoming a low-effort organization,” Falcone said. “The most important word in AI is intelligence - doing things smarter and augmenting our (companies), so they can do things better and faster.”
Falcone said his company is trying to do that by improving its website interface and better integrating chatbots into it.
Many companies, from small to large, have been using chatbot technology in their websites as search engines and other customer support. Some, such as Verizon, even list a chatbot option on their website’s main page or “Help” page.
Companies around the world spent $26 billion to $39 billion on artificial intelligence in 2016, with most being spent by larger companies, according to a recent report by the research firm Mckinsey Global Institute.
A recent survey by Oracle, first reported by Business Insider, found that of roughly 800 decision makers from companies throughout the globe, 80 percent said they currently use or plan to use chatbots by 2020.
The rush for chatbots is tied to the savings they could potentially create for corporations. McKinsey’s estimates show, for example, that about $23 billion in salary could be saved in just the customer service representative sector by using chatbots to speed up customer responses.
Companies are also striving toward a future where more of their services are connected not only to their own products, but also to physical chatbots like Amazon’s Echo home systems, according to Brandon Mcgee, the AI and marketing director for Harland Clarke, a finance, marketing and communications firm based in San Antonio.
McGee, who previously led the mobile growth department at Facebook, helped deploy Harland Clarke’s first chatbot about about a year ago. The company is now concentrating on the improvement of the chatbot conversation, McGee said at the seminar.
There are companies that have been created solely to concentrate on building chatbot technology. Convertible, for example, which has offices in Austin and Dallas, began in 2014 to design and build chatbots for tech giants such as Facebook, Amazon and Google.
Other chatbot firms are also using AI to augment human workers’ performance.
At Creative Virtual USA, a provider of virtual assistance for companies like Chase Bank and Verizon, the company is testing AI that creates automatic responses to online customer questions which customer service agents on the other side can choose to send customers if the answer is accurate or alter and then send, according to CEO Mike Murphy.
Beyond that, companies are thinking about the ability for chatbots to do more than just answer questions, Murphy said. If chatbots could not only answer questions but also effectively sell customers products at the same time, companies’ operations could be optimized.
“Success within these customer service solutions really needs to combine two really important elements of business: a good appreciation for the fundamentals of customer service, and its technological capabilities in the back end systems to make things intelligent,” Murphy said. “These things have to be easy to access.”
Murphy and others at the seminar made a reference to the ride-hailing company Uber, which ballooned in popularity mainly because of its easy usability. For chatbots to have similar success, Murphy said, they need to be easier to access and use.
Right now, the process is not as seamless as it can be, Murphy also noted.
Many chatbots remain clunky and incomparable to human interaction.
At companies like Bristol-Myers, the challenge can be complicated because customers ask complex questions regarding medical treatment that chatbots can’t handle with ease, Falcone said.
“The marketing is better than the production right now,” Falcone said of chatbots.
Then there’s the concern about job losses.
With industries such as manufacturing having lost a significant number of once-human controlled jobs to automation, there are worries that the same is happening to customer service roles and other jobs at companies due to AI.
Analysts from major business auditor PricewaterhouseCoopers, for example, used various U.S. workforce and Census databases to recently predict that 38 percent of U.S. jobs could “potentially be at high risk of automation by the early 2030s.”
But that fear is inflated, said Michael Jude, a program manager at Frost & Sullivan.
Jude said corporations are looking to use chatbots and other AI products to improve the performance of their employees, not to replace them. He argued that the money gained from boosting employees’ productivity could actually create jobs or bring some foreign positions back to the states.
“People think AI is human replacement, but it is really human augmentation and amplification,” Jude said. “Instead of being on a phone call or chat with a customer for two hours, it can last two minutes (using AI).”
Companies must weigh the costs and benefits of investing in chatbots and other AI, Jude said. Estimates for chatbot costs are wide-ranging but can be priced in the thousands. Other AI such as machines for supply chains can cost a lot more.
Many corporations have already made that investment. More are sure to follow.
“Automation outsourcing alternatives to labor are adopted when labor becomes too expensive for the task at hand. That's neither good nor bad. That's the way businesses work,” Jude said. “So, the adoption of AI is not theoretical. It's certain.
“AI is an enabling technology. You don't set out to do AI. You set out to solve a problem, and AI is part of the solution. We’re now starting to understand what (that) really means.”