As we lean into 2019, chatbots and digital assistants will continue to play a prominent role in discussions about artificial intelligence (AI) in the workplace.
For many employees, chatbots and other conversational AI tools will be a user-friendly introduction to enterprise search, something businesses have struggled with for decades. However, flawed learning models can inadvertently produce search results that are bogged down by content that doesn’t hold much business value. For example, rather than unearthing critical company documents, these search engines may instead highlight content that has a high engagement level and holds a certain amount of “viral” value — like a shared photo of the holiday party.
The potential for chatbots is in their ability to remove “Where is it?” from the search conversation, and transform it into a “What do you need? I’ll find it” experience.
When users show interest in particular topics, digital assistants can learn their tendencies and alert them when relevant content is being posted. While the prospect of automation is an exciting one, we must keep a realistic perspective on what we can and cannot reasonably expect from digital assistants in the near future. Like anything that sounds really good, the devil is in the details.
Related Article: Want to Use Chatbots and Smart Speakers in Your Workplace? Think Big
Conversational Shortcomings: Where Chat Technology Must Grow Up
In their current state, chatbots are used heavily in customer service and retail operations, helping shoppers get to where they need to be and answering straightforward questions. If you visit a retail site, chances are you’ll be met with a popup dialog box with a message from a friendly assistant asking what it can do for you. But while chatbots offer a number of benefits, they also have some widely known downsides: They get confused easily, they often can’t answer simple questions, and they often provide no more help than what you’d get if you just typed a question into a search box.
Additionally, these automated interactions lack a human element that people so desperately want. A recent study (registration required) from software company Acquia found that 45 percent of consumers in North America, Australia and Europe describe chatbots a “annoying.” Further, a recent PwC study on customer experience stated that 64 percent of U.S. consumers and 59 percent of all consumers feel companies have lost touch with the human element of customer experience. And 75 percent of global consumers say they would rather deal with a human than a chatbot or other automated systems.
Not being conversational enough is and has been a huge problem for chatbots. The whole purpose of a chatbot is to be able to provide around-the-clock support in a way that makes people feel as though they’re chatting with live human beings. That has yet to be achieved.
If chatbots are still struggling in 2019 to be conversational and effective at small tasks at the customer service and retail levels, imagine the problems they may face at the enterprise level. It’s entirely more complex to map out the detail-oriented conversational language that will help an employee navigate through the weeds of internal systems.
So before you roll out a chat-based system in the digital workplace, it is imperative to master the complex machine learning and language obstacles first.
Plus, even if conversational technology could be deployed easily, tech giants like Google would have made even bigger investments into the digital workplace and collated the output of various applications and their application programming interfaces (API) to showcase it. The fact remains that the amount of data required for a truly conversational and intuitive chatbot experience is far too vast. There is a greater level of machine learning required around language programming and intent that we have not seen, and likely won’t see in 2019.
Related Article: Chatbots Belong in the Workplace (Provided They're Well-Designed)
An Alternative to Chatbots: Technology That's Mature Enough for Today’s Digital Workplace
As noted earlier, a goal for these assistants is to be able to proactively help employees streamline the way they connect with the people, tools and information they need. They can already assist with small-scale tasks, but there is little proof that they can be effective at handling more complex tasks.
The solution to this is to build upon and make the most of the tools you likely have on hand already. Rather than having to ask a clunky chatbot for help, workers should be able to easily access what they need from a centralized location. Hubs like employee intranets allow companies to have information readily available at employees’ fingertips — all they have to do is search. Beyond their ability to store information in an easily-accessible way, intranets also allow you to instantly connect with the people, tools and other systems that you need. They are true collaboration and support systems.
This goes hand in hand with companies needing to expand their mobile-ready approaches. As the workforce gets younger with the infusion of tech-savvy millennials and Gen Z workers, companies need to expand their mobile strategies. Remote working is growing at the fastest rate we’ve seen, so it’s important for employees to be able to access what they need on the go.
Implementing social media features into your communication channels will also help connect remote employees to one another by enabling a more human element when they go looking for a resource. Younger workers grew up on social media, so this functionality will feel familiar to them, and the tools will be easy for them to adopt. Features like the ability to comment and share allow employees to exchange information and resources freely — and between actual humans instead of via bots!
Related Article: The Enterprise Chatbot: Your Future Coworker?
Chatbots will someday be put to meaningful use in enterprise settings, likely helping with search and assisting on day-to-day projects. But they currently face too many hurdles to be effective. As is the case with any evolving technology, there will be growing pains that need to be worked out before chatbots become widely adopted.
As you look to invest more in this technology in the coming months and years, be sure to roll out chatbots in environments where content, tools and resources are up to date, relevant and easy to find, so that the chatbots can better aid employees in finding what they need. Chatbots will never solve enterprise search problems if they’re built on cracked foundations.
Rather than viewing 2019 as the year chatbots will become fully developed and functional, it’s better to think of 2019 as the year they continue to improve so that they can one day become an integral part of any digital workplace.
Learn how you can join our contributor community.