robot suit running away
PHOTO: Roy Chan

Not all AI is created equal. Siri can find, book and calendar local movie tickets in minutes. Alexa or Google Assistant can turn up the thermostat automatically and order you a sweater when the weather gets cold. Virtual assistants are ever-present and increasingly capable and people have quickly adjusted to these types of interactions.

By contrast, most enterprise chatbots essentially consist of a hovering box on a webpage — an experience that pales in comparison to the norm consumers are used to. Floating boxes have their place, but they’re quickly becoming only a small part of a bigger picture.

Consumer-facing artificial intelligence (AI) gets its power from its connections. Siri’s movie reminders come from having calendar access, and Alexa’s temperature regulation relies on integration with a separate smart device. Currently, enterprise corporations using bots for interactions with current or potential customers can’t bring the same breadth of influence to the table, because they can’t access the same information without crossing major privacy boundaries. However, they are able to expand bots’ influence in their own digital knowledge centers.

Forward-thinking companies are building bots that treat their enterprise communications networks the way consumer bots treat their parent devices. Rather than being siloed within a single messaging channel, these pieces of proprietary AI shift from IM to voice call to Facebook, from placing an order to registering a query and back without interruption — like an independent entity. 

Just as Siri and Alexa use data from their respective devices to improve their services, connected enterprise AI uses information from different channels and internal databases to make smarter, more flexible bots.

Related Article: A Good Chatbot Is Hard to Find

Bringing More Chat to Bots and More Bots to Business

The first challenge for understanding how chatbots will change the enterprise is the ambiguity of the word “chatbot.” According to an Oracle report (registration required), 80 percent of businesses plan to use chatbots by 2020, but the types of systems that customers actually interact with will vary wildly. Some chatbots will answer common questions with canned responses, others will shift between messaging platforms, automate services, and even enable complex transactions. As chatbots become the norm, businesses with more advanced systems will stand clearly apart from their competitors in the eyes of consumers.

One differentiator will be bots that hop between platforms, freeing customers from the need to visit the webpage, use an unfamiliar service, or engage with a brand on a platform they consider private. Certain platforms will be indispensable for comprehensive service. Three billion people now use the “Big 4” messaging systems: Facebook Messenger, WhatsApp, Line and WeChat, according to statistics compiled by Accenture. SMS, too, remains a critical way to interact with large audiences, securely verify users and conduct instant transactions. Just as marketers have come to take their engagement cues from individual customer preferences, service-oriented chatbots will do the same, meeting each customer on his or her own platform of choice, giving customers the same flexibility they’ve come to expect in personal communications.

More strikingly, some of these chatbots will also be linked to broader systems that allow them to either access more information (through customer databases) or take more action (through integrations with other communications systems). Like Siri can summon a personal contact at will, enterprise chatbots will be able to give customers exact and relevant details without asking, or forcing them through a series of menus. Customers will be working with chatbots that know their past orders, their preferences, and possibly information about their location, such as the local weather.

Bots connected to communications systems will be able do things like monitor customers’ language and tone and connect them with a human agent if there are signs of frustration, complex questions, or opportunities to help customers further. They’ll also be able to pass on accumulated knowledge so that everything the bot learns, its human counterparts learn as well.

Related Article: 4 Questions to Ask Before You Send in the Chatbots

Merging Chatbots and Virtual Assistants

Some integrations bring chatbots more places, others simply give them more functionality. As more evolved bots start to resemble virtual assistants like Siri, Google Home and Alexa, some will also simply become virtual assistants themselves. 

In response to rapidly changing consumer preferences, some companies have started employing consumer products for enterprise purposes. Amazon recently announced a new iteration of Alexa for businesses, and the freelance marketplace Upwork already has a section of full-time professional Alexa experts ready to apply their skills for enterprise purposes. Microsoft has the business-focused AI Cortana, and Google and Apple will likely be close behind. Bringing this type of AI directly into pre-existing communications channels is an easy way to make chatbots immediately more sophisticated.

Virtual assistant-powered bots are able to recognize problems and opportunities through natural language processing and make judgments about next actions. With these type of technologies embedded into chatbots, as well as integrations with CRM tools, internal data systems, and other business apps to give bots more context, the result will be a generation of hyper-capable, multi-functional enterprise AI.

According to Gartner, by 2020, daily interactions with chatbots will be the norm, and 30 percent of web browsing will be done through a bot, rather than a screen. It doesn’t seem like an extraordinary claim, based on the ways enterprise chatbots are already evolving.

AI is becoming a part of everyday life. It’s inevitable that chatbots will play a major role in defining how businesses function in the near future, but it won’t be chatbots as we see them today. Like humans, bots perform best when exposed to a myriad of different tools and types of information.  The more connected they become, the more impressive their output will be.

Related Article: Intelligent Virtual Assistants: Improving, But Still Not AI