group of chatter teeth toys, with one in front of the others
Editorial

Who Should Be in Charge of Enterprise Chatbots?

8 minute read
David Roe avatar
As chatbots become ubiquitous in the enterprise, it's important to determine what role they're playing and what value they're creating.

Love them or hate them, chatbots and digital assistants are now a confirmed technology in the workplace. As such they need to be supervised and integrated into the wider enterprise. So who should be responsible for them and how should they be used?

Chabots Focused on Customer Support

According to Rupert Pople, who founded and develops Your Smart Home Guide, it is widely accepted that artificial intelligence (AI) will dominate just about every aspect of operational, strategic and financial processes in the enterprise. That said, chatbots and digital assistants more closely align with customer support.

Messaging apps as well as business specific online chats are becoming the ‘go-to’ place for customer support, in preference to calls or email. Research firm Gartner, for example, estimates that by 2020 chatbots will power 85 percent of customer support interactions.

Research showed that chatbots reduced the average resolution time of most inquiries by huge amounts. Replying by email can take in excess of 24 hours while chatbots typically complete most basic tasks in under an hour. Additionally, customer service calls incur huge additional costs such as call centers, which rely on wages and overheads. Chatbot automation will require some computing power and server storage, but the costs associated with this are small in comparison.

"In the long run, the quality of the support has the potential to be much more accurate and consistent than it is from humans," Pople said. "Initially there will be more problems, but as the quality of AI (and digital assistants) improve via machine learning and other technological enhancements this will change."

Related Article: Why I Hate Customer Service Chatbots

Use of Chatbots Expanding

However, Joshua Kail, a tech PR professional who works with a number of different AI and chatbot vendors, said that with the diversification of chatbots, their roles and value for an enterprise are constantly expanding. If chatbots are used consistently to improve customer experience, they are now also widely used in communications, work management and project management customer support. With this being the case the role of who is responsible for them also varies. In other words, chatbots should be the responsibility of the chief line-manager.

If the company has a Chief Digital Office (CDO) in place, as many organizations now do, they are the ideal person to oversee chatbots. The CDO, by nature, has one foot in IT and another foot in the business side of things. This gives them the uniquely needed perspective of duality between technical and functional.

For those companies without a CDO the oversight should fall equally upon IT and the department head. This means that for a chatbot that runs customer support, the head of marketing or head of sales/support (depending on the organization of the company) should oversee how the chatbot is being used, the quality of its interactions and, ultimately, customer satisfaction with the experience. IT should oversee chatbot implementation from the nuts and bolts of how it is integrated within the mobile or web experience as well as ongoing oversight for any bugs or other issues that may arise.

“For a chatbot to run effectively it needs oversight from both of these perspectives. Again a CDO is in the best position to bridge the differences, but open internal communication between departments is another option — there are even bots for that!” Kail said.

Without this supervision, enterprises face the same risks they face when cell phones and smartphones became everyday products. At the time, workers would bring these phones and apps into the organization creating massive headaches for security. Shadow bots are now a growing concern as workers bring them into the organization and use them for work without the authorization of IT.

"This, like BYOD [Bring-You-Own-Device], falls on IT to put the technical protocols in place to stop employees from uploading their own software into the system. It also falls on HR to create, enact and educate the needed behavioral policies and protocols with employees," Kail added.

Related Article: Is Your Enterprise Ready to Fight Off 'Rogue' Chatbots?

Chatbots Move Into Marketing, Sales

The vice president of marketing at HubSpot, a developer of inbound marketing and sales software, Jon Dick pointed out that Chatbots have been used in customer service for some time, but that enterprises are starting to bring chat into their marketing and sales processes as well.

He said at HubSpot they use chat staffed by humans and bots, their experience is that people engage more with bots (as much as two times more on some pages) even when live chat is available. In the wider industry, HubSpot research surveyed a few hundred B2B companies on their use of chat, and found companies that are growing have invested far more in chat and monetization than companies with stagnant or declining growth rates.

“Chat generally is a massive friction reducer, because it allows people to connect in real-time,” he said. “Imagine being a buyer who wants to ask a simple question or make a purchase on your website. Relying on an entirely human-powered process adds a huge amount of unnecessary friction to this process — why make a buyer who's ready to purchase now wait to speak with a human?"

Overall, HubSpot views chatbots as an investment in customer experience rather than a strategy to lower costs or increase employee productivity. Consumers are asking for live chat at every part of the customer lifecycle — 20 percent want to use chat in the sales process, 33 percent in the marketing process, and 42 percent want it as an option for support.

Learning Opportunities

In terms of ownership, Dick points out that chat is a full flywheel software, meaning that lots of departments need to work together to successfully deploy it. HubSpot sees marketing as the right owner of the software and bot building, and that sales and services are the right teams to provide the human interface.

“That said, we cannot build great bots without understanding the conversations that humans are having and the goals of the users,” he added.

Related Article: New Data Indicates People Don't Use Chatbots Like We Think

Chatbots: Homes to Enterprise

What people experience in their personal lives will eventually begin to permeate their work lives as well. With digital assistants beginning to do more and more for us personally — from reminding us about appointments, to confirming delivery information or booking dinner and hotel reservations — it’s a logical leap to expect to see that same technology moving into our work lives, said Splice CEO Tara Kelly.

Imagine a smart office where a digital worker can just ask the digital assistant to find and book an available conference room, send a lunch order to the cafeteria, or broadcast a message to the rest of the team asking them to push back a meeting by 10 minutes. Companies are always looking for ways to improve efficiencies — these seem like natural extensions of that approach.

Deployment will naturally fall to IT, but it will be interesting to see how different departments will be involved in the build, she said. The power of an ever-evolving AI-based assistant able to handle employee requests across departments shouldn’t be discounted. HR would be using it to answer questions on employee policies and benefits; accounting would be using it for inquires on expense reporting and invoice payments; IT for help desk requests … the possibilities are endless.

“I think we’ll see an entire ecosystem develop for enterprise-level skills that can be customized to be on-brand and meet evolving needs,” she said. This will be similar to the way that people never imagined that they would want to use a mobile phone to summon a car, level a picture on the wall, act as a flashlight and magnifier, or be an all-in-one remote control.

Finally, marketing will want to be deeply involved with the development to ensure that messaging and voice are both on-brand. That’s where we’re helping several companies today with their external-facing skills, and I’m sure we’ll soon be doing the same on the internal-facing side.

Kelly said the same risks exist with employees bringing their own assistants in as they do with employees doing anything in a rogue manner. However, she said the better question is probably “what percentage of people are already using a shadow assistant at work. Mobile phone apps can already do everything from transcribing the audio of conference calls and meetings to accessing company calendars and emails.

Related Article: Chatbot or Chatbaby? Why Chat Technology Needs Time to Mature

No Progress After Siri?

Despite progress, Sean Kendall, director of customer experience at TetraVX, said that today there is little change in the digital assistants from Siri, created eight years ago, to today. While there is little argument that enterprises are embracing digital assistants, an overwhelming majority of enterprises use it as a dictation tool; voice recognition software has been around long before Siri and Alexa.

The role will expand to help with scheduling, management of daily activities and collaboration, but it still feels limited in scope. Soon they should, and will, play an entrenched role in conducting training to team collaboration management by coordinating hundreds of schedules for meetings and events. The overall virtual assistant is ripe for disrupting the enterprise collaboration market, but because it has to span across so many diverse platforms, programs and security hoops it is just now starting to hit a stride.