Conversational artificial intelligence (AI) is on the rise, and both Gartner and Accenture believe the integration of conversational AI will emerge as a top priority for the c-suite by 2020.
Conversational AI is a voice assistant that can engage in human-like dialogue, capture context and provide intelligent responses. Examples include Apple Siri, Amazon Alexa and Google Virtual Assistant. Developments in AI in the past few years, including machine learning, natural language processing and image and speech recognition have promoted its use, according to Kenneth Research.
SEE: Tech Predictions For 2020: More must-read coverage (TechRepublic on Flipboard)
A recent Gartner report finds that by 2022, 70% of white-collar workers will interact with conversational platforms on a daily basis.
“There has been a more than 160% increase in client interest around implementing chatbots and associated technologies in 2018 from previous years,” said Van Baker, a vice president at Gartner, in a statement. “This increase has been driven by customer service, knowledge management, and user support.”
This means that consumers are expecting interactions with brands to be on their terms and full of personalized experiences, added Laetitia Cailleteau, global managing director for Conversational AI at Accenture.
Already, Accenture has seen a demand from high-level executive clients who are interested in incorporating conversational AI technologies into their business models, said Cailleteau, who is also a founder of Accenture’s London Liquid Studios.
“So far, we’ve had quite a lot of successes in the telecom industry, and big telco providers have been using [conversational AI] at scale,” Cailleteau continued. For example, at one client, Accenture is running the equivalent of five call centers with robots. The energy and banking industries are also using this technology.
Voicing a human experience
One of Accenture’s “golden rules,” which is critical for successful implementation of conversational AI is building a good design experience for dialog. Right now, apps and websites are flat and not promoting conversation by default because they aren’t meant to; they are more about displaying information, Cailleteau explained.
There has to be a paradigm shift in conversational design, and companies “need people who are more focused on the words they’re using; the crispness of language that is typically used in the film industry,” Cailleteau said.
This is especially important in the voice AI experience, she continued, because “if you don’t spend the right amount of time to design that level of experience, people will drop off. Then you have to escalate to a human and that is defeating the purpose” of a chatbot.
Call centers, for example, have typically been outsourced, and scripts were written years ago and are very inward focused, rather than human focused, she observed. “When we actually shift those scripts to a human focus … we tend to have a lot more uptake. People find it more natural to talk to a bot,” Cailleteau said.
Another shift in the use of conversational bots is that they can now be used on many different channels, including websites and text messaging, and if people have a bad experience they tend not to go back, she said.
Gartner agrees that demand for voice-enabled chatbots is increasing “as they provide a natural way to interact with conversational technologies,” the report said. “Be prepared to meet this demand by specifying voice support in your solutions.”
SEE: Managing AI and ML in the enterprise 2019: Tech leaders expect more difficulty than previous IT projects (TechRepublic Premium)
A new population of users
In 2020, expect conversational AI to move beyond call centers and into other channels, and outside of traditional business settings, said Cailleteau.
“Over the past few years at Accenture we’ve been conducting pilots with the elderly population to see how they work with conversational AI tools and have seen that voice AI is a very powerful medium to connect with the elderly,” she said.
This is because the elderly have been successful at using devices like Apple’s Siri, Amazon’s Alexa, and Google Assistant, and with an aging population globally, this can open up a new market segment for a lot of brands, Cailleteau explained.
What to think about
There are more than 2,000 enterprise chatbot vendors, according to Gartner, and organizations should implement governance policies and apply best practices to their conversational platform strategy.
Companies should also screen vendors carefully and make sure they can deliver an enterprise-grade offering, Gartner recommends.
Additionally, Garnter continued, only attempt to create chatbots internally if you have the right data science and machine learning assets. Otherwise, look to third-party providers that specialize in data preparation or those that build and host chatbots.
SEE: Microsoft HoloLens 2: An insider’s guide (TechRepublic download)
Speak human, but don’t pretend to be one
Companies will start prioritizing and promoting transparency of their AI technologies next year, Accenture believes. “It is becoming increasingly clear that to be successful, companies must be transparent about their business’ ‘internal kitchen’ when it comes to AI,” Cailleteau said.
In other words, “Speak human but don’t pretend to be one — that’s another golden rule of conversational AI.” Brands must be upfront that they are using bots “so humans know what they’re dealing with.”
Both Gartner and Accenture advise companies to incorporate tone, emotion, and personality into their bot technologies. “A chatbot is a representation of your brand and it’s definitely important to have the right tone of voice” and focus on being human-centered and address people by name, Cailleteau said.
But she cautions that it can be difficult for a bot to ascertain how someone is feeling based on a small snippet of information. This is why voice AI is becoming important, because a bot can get a lot more information through intonation than in a text.
Cailleteau also recommended that companies “be aware of the state of your knowledge.” Accenture worked with a large aviation company to create a bot to help people navigate complex situations when flying — such as a doctor carrying organs from one country to another with a stopover and how to figure out different airlines’ rules.
“So when you start asking a question of a bot you need to have curated knowledge,” to answer those questions, Cailleteau said, “and a lot of organizations haven’t invested in their knowledge, maintenance, and curation. That’s something that can trip a project.”
You don’t have to start a chatbot project with curated knowledge, she added, but you have to have people curate it during the implementation.
Besides the trend toward using voice in conversational AI, Cailleteau said companies must start offering asynchronous interactions so a customer does not have to be logged in to a website to wait for a bot to reply to a question but come back when they want and take the next step.
“For me, it’s an important investment to provide new means to interact with a customer,” she said. “Conversational AI is providing new ways for customers to interact with a brand on their own terms.”