Is “chatbot” a misnomer? Let’s think about what creates successful chat experiences

[dangrady510]

A chatbot is a computer program that simulates interactive human conversations.  

I'm sure you've come across one on your favorite airline, retailer, or cable provider's website or mobile app.  The chatbot is just sitting there waiting and in some cases over anxiously popping up to "simulate an interactive human conversation" with you.

The problem is, you most likely didn't go to the site or app just to "chat", you most likely went there to get something done.

In fact, if you think about it, the term "chatbot" is a bit of a misnomer (a wrong or inaccurate name or designation).  As we start to see more and more chatbots surfacing in the enterprise, we may want to consider calling them "workbots" because this is more in alignment with why we'd interact with one of these digital entities - we want them to do something on our behalf.

The type of work we are typically asking the bot to help us with usually falls into one of three buckets: 

• Information retrieval - initiating a search across a knowledge base or checking the status of a request or case
• Information gathering - providing step by step guidance through a process while shielding users from the multiple application interfaces behind the scenes.
• Automated fulfillment - kicking off a workflow that automates the resolution or completion of a task.

Successful enterprise chatbot deployments can assist in employees with these types of tasks with as little friction as possible, eliminating mundane work and increasing overall productivity across the organization.  

However, with so much buzz about AI and machine learning, organizations can easily get overly distracted by Natural Language Understanding (NLU) and forget that while those capabilities play a role in the consumer or employees' experience, overall satisfaction will be predicated on the bot's ability to help them complete their work - not solely its ability to understand the nuances of the human language.

Organizations that are considering deploying a chatbot within the enterprise should also be focusing on the following in addition to the NLU component:

• Where are we going to make the chatbot available?

On a Service Portal? Via Mobile Devices? Do we want to embed it in Slack, MS Teams or Facebook Workplace? In order to maximize the adoption and usage and more effectively scale your services you want to make sure the chatbot is available where the employees already are.  The ServiceNow Virtual Agent is accessible via mobile devices, comes with prebuilt integrations to Slack, MS Teams, and Facebook Workplace, it can be embedded in third party portals as well as the Now Platform's native Service Portal.

• What domain specific data do we have access to that can improve the employees chatbot experience?

The more data the chatbot has access to, the more intelligence it can bring to the conversation and the less it will have to ask the employee for.  This will streamline the conversation and path to resolution.  Think of all the data the Now Platform already has about an employee - their location, organizational structure, IT assets, information about the status of services that impact them, tickets or cases they have open, etc.  And for data that doesn't natively reside on the platform, it offers a number of prebuilt integrations through Integration Hub.  The ability to seamlessly access this data provides an enhanced chatbot experience.

• Are we able to offer a Live Agent transfer option?

No chatbot you roll out is going to be perfect.  It'll never be able to handle every question or situation.  If we keep in mind that the ultimate goal is to get the employee to resolution we should be thinking about providing them a path to get to a human, quickly and seamlessly.  The ServiceNow Virtual Agent offers the ability to transfer to a live agent both dynamically(based on conditions within the conversation) or at the employee's request.  The full context of the conversation is transferred over so the Agent can easily grasp the situation and avoid the annoyance of re-asking questions.  This post talks more about the importance combine humans and chatbots.

• What services are we already providing?

You've spent time building up a Knowledge Base, feeding and caring for your CMDB, creating catalog offerings, automating processes via workflows and orchestrations.  Use the Virtual Agent to more effectively scale them out by making them available via this new modern conversational interface.

When we are having a real conversation or "chat" with another real human being we appreciate it if the other person is "actively listening".  When dealing with a chatbot that is simply "simulating an interactive human conversation" the action the chatbot is able to take based on what it has heard - the work it is able to accomplish on our behalf - is the piece that consumers and employees appreciate most.  This appreciation will manifest itself in increased productivity and improved employee and customer experiences.

Check out other Virtual Agent Blog Posts:

The Digital Ensemble that Helps Virtual Agent Create Better Experiences 

Now Intelligence & Virtual Agent - Better Together

How to Chat About Chatbots