Bots for Business, Chatbots
A Match Made in Heaven: Chatbots and the Digital Employee
Remote work has been on the rise for years now, but the COVID-19 pandemic has rapidly accelerated this trend in 2020. By using chatbots for remote workers, companies are capitalizing on two of the biggest workplace trends—artificial intelligence and remote work— and accelerating the digital transformation of their business.
January 20, 2021
Remote work has been on the rise for years now, but the COVID-19 pandemic has rapidly accelerated this trend in 2020. An article from the Brookings Institution estimates that roughly half of U.S. workers were telecommuting in the immediate wake of the pandemic. To facilitate remote work, organizations need to investigate ways of maintaining and even increasing their employees’ productivity and efficiency.
One way of improving the “new normal” of remote work: incorporating chatbots into employees’ routines. IT research and advisory firm Gartner have already predicted that by 2022, 70 percent of white-collar workers will interact with chatbots on a daily basis. The rise of the digital employee working remotely will no doubt help hasten this development: Chatbots empower users to solve their own problems while alleviating pressure on other human employees.
In the following text, we investigate how chatbots can enable and improve remote work, including a few case studies of companies that have successfully brought chatbots into their telecommuting workflows.
What’s new in chatbot technology?
Chatbots have been around for decades, beginning with the ELIZA program in the 1960s, but advances in the past year or two have radically altered what’s possible at the forefront of this technology.
Developing a realistic, human-like chatbot requires two complementary achievements: first, that the bot can understand what users say to it and what they need; and second, that it can respond appropriately in fluent, natural language. Programs such as Blender and GPT-3, both released in 2020, leverage cutting-edge artificial intelligence (AI), machine learning, and natural language processing (NLP) techniques to accomplish both of these objectives.
Blender is a state-of-the-art conversational artificial intelligence developed by Facebook, which the company claims is “the single largest open-source chatbot created to date.” The model's 9.4 billion parameters—more than three times as many as the previous record-holder—enable it to perform a wide range of tasks, including developing a unique personality, demonstrating empathy, and discussing a wide range of topics by training on articles from Wikipedia.
OpenAI’s GPT-3 is a highly advanced language generation model that, provided a short prompt of a few words or a sentence, returns a new text that it believes is most relevant to the given context. A blog post by ZeroCater founder Arram Sabeti illustrates how GPT-3 can be used to generate everything from fake screenplays and rap songs to interviews and user manuals.
Although GPT-3 is still in a controlled beta, companies are already experimenting with it for customer service purposes and the technology holds tremendous promise for building advanced chatbots. The customer service start-up Sapling, for example, is using GPT-3 to rapidly and automatically generate suggested responses for customer service representatives who are handling users’ help requests, while the start-up OthersideAI uses GPT-3 to write emails on your behalf using a simple summary of what you want to say.
How can chatbots help your remote employees?
The technology behind chatbots is constantly advancing—but how can you take advantage of these developments and put them into practice in the remote workforce? Following are just a few ways that your remote employees can benefit from using chatbots:
Onboarding and training. Your new employees need to get up to speed fast, especially when they don’t have an office to go to. Chatbots can help walk recent hires through the onboarding and training process, helping them fill out the right forms and answering common questions.
Support. Telecommuting means that it’s harder for employees to get technical support—after all, you can’t bring in your faulty hardware or software for an in-person inspection. This, in turn, makes the work of remote IT teams harder than ever. By offloading simple IT help requests to an automated chatbot, employees can get answers faster and help desk workers can breathe a little easier.
Productivity and efficiency. Chatbots can help employees accomplish their tasks faster in their day-to-day activities—and these small enhancements add up over time. If you are searching for the location of a recently accessed file, for example, you can ask a chatbot to open the containing folder (or simply open the file directly).
Workflow automation. In addition to simpler tasks, chatbots can also perform a great deal of heavy lifting behind the scenes. For example, chatbots can remind sales representatives to follow up with a promising lead. Once the conversation is over, users can provide the relevant information to the chatbot, which then creates a new entry in the organization’s sales CRM software.
Other interactions. As one of the most visible “faces” of the company for remote workers, chatbots can serve as the connective tissue that binds your employees together while working remotely. For example, chatbots can inform users of upcoming events or social activities or suggest new opportunities for training and professional development.
Three companies using chatbots for remote work
With so many advantages of using chatbots for remote work, let’s look at how three different companies have put them into practice in three different domains: IT, customer service, and HR.
With the pandemic’s arrival in March 2020, software company Adobe was left struggling how to shift its global workforce of 22,000 people to working remotely. As part of its efforts, the company created an organization-wide Slack channel for its IT help desk, where employees could get assistance with their questions and issues.
Because many of these questions and issues were cropping up over and over again, Adobe realized that it could build an automated chatbot to answer them. Using AI, machine learning, and natural language processing (NLP), the chatbot can point users to a relevant FAQ question or knowledge base article or walk them through common tasks, such as setting up a VPN.
According to Cynthia Stoddard, Adobe senior vice president and CIO, “This AI enhancement has reduced the average time required to dispatch and route email tickets from about 10 hours to less than 20 minutes. Continuous supervised training on the routing bot has helped us reach approximately 97 percent accuracy—nearly on par with a human expert. As a result, call volumes for internal support have dropped by 35 percent.”
The Adobe chatbot has been successful largely for two reasons. First, the company has sought to maximize its return on investment by focusing on the most frequently mentioned issues and questions. Second, the bot is able to learn from experience: Adobe engineers continually review its previous conversations to help train it on new scenarios and avoid errors.
Not only can chatbots help address the concerns of internal employees, but they are also increasingly used as a savvy marketing technique. When potential customers first arrive on your website, a bot can send them a predefined greeting and offer to help with any questions they have.
Of course, chatbots can be used at any stage throughout the buyer’s journey, both improving the customer experience and helping your human agents work more efficiently. Talla is an AI-powered software start-up that offers three interconnected solutions:
Customer Assist, a chatbot for automating 24/7 chat and web support. The company claims that it can help resolve 90 percent of customer issues, handle inquiries 50 percent faster, and make human agents 20 percent more productive.
Agent Assist, a collective repository that can answer a wide variety of common questions to help agents address more customers more quickly. If the tool can’t automatically provide an answer, it asks a human who can and then stores the result in its knowledge banks.
Automations for integrating with existing systems and streamlining workflows.
3. Artificial Solutions
Most chatbots work best when they are designed to handle a specific topic or a narrow area of inquiry, and human resources has become one of the most popular domains for deploying chatbots. HR chatbots can be used for purposes such as:
Updating employee contact information.
Asking for vacation time or sick leave.
Looking up information on healthcare, retirement, or other benefits.
Accessing company policies and regulations.
Making a new request for equipment.
If the chatbot faces a complex or multifaceted question, these issues can be redirected to a human HR employee.
In April 2020, the conversational AI technology company Artificial Solutions launched the Tiva chatbot for handling HR inquiries. The bot comes preloaded with existing dialogues for common questions (that you can modify, as necessary) and can also pull information from external websites.
By default, Tiva integrates with web chat platforms, interactive voice response (IVR) systems, and Microsoft Teams, but connecting with other solutions is easy. Tiva also includes analytics capabilities so that human HR professionals can see how employees are interacting with the bot and make proactive improvements. According to Artificial Solutions’ announcement: “Tiva solves an immediate pain point for organizations having to handle sudden changes as a result of COVID-19, but also looks to the future.”
The rapid and dramatic shift toward working from home has upended the workplace this year, but savvy organizations have been able to turn these challenges into opportunities. By using chatbots for remote workers, companies are capitalizing on two of the biggest workplace trends—artificial intelligence and remote work—and accelerating the digital transformation of their business.
Author: Dr. Mark van Rijmenam
Dr. Mark van Rijmenam is the founder of Datafloq and Mavin.org. He is a highly sought-after international public speaker, a big data and blockchain strategist, and author of the three management books, including The Organisation of Tomorrow, which discusses how AI, blockchain, and analytics turn your business into a data organization. He holds a Ph.D. in management from the University of Technology Sydney on how organizations should deal with big data, blockchain, and (responsible) AI, and he is the publisher of the f(x) = ex newsletter, read by thousands of executives.