Considerations For Deploying Enterprise Chatbots



Esther Shein
05/22/2018

Chances are, you’ve interacted with a chatbot already without knowing it – perhaps when you’ve gone online to inquire about how to return something you bought or to find out why it didn’t arrive on time

There’s been a lot of buzz about chatbots and their growing prevalence in the enterprise. Regardless of whether you think these AI-enabled bots are the “it” app, the critical question to ask is whether they are effectively solving the core pain points your organization has. 

The drive to digitally transform themselves has meant organizations are going the extra mile to connect with customers – and especially millennials – in a more personal way. Also, customers today want self-service and the information they seek…now. No longer are chatbots a “nice to have.” This has created a significant business need and momentum for chatbots in the enterprise.  

In particular, demand is huge in contact centers, where there is high turnover, increased need for training and a need to reduce operating costs, according to a recent Deloitte report. Fueling the rise are advancements in natural language processing, processing speed, machine learning models and data availability, the report notes. Chatbot platforms are also maturing and enable users to more easily build and manage chatbots with intuitive drag-and-drop interfaces.

“One of the key challenges to designing good solutions for users is how to make things easy to understand and intuitive,’’ agrees Kevin McMahon, director of Emerging Technologies at digital consultancy SPR. “Chatbots that can contextualize the motivations of its users can significantly ease the learning curve of a system and improve the quality of interaction with it. This can be a key driver to successful interactions and outcomes.”

A majority of SPR’s enterprise customers are looking for ways to leverage chatbots for both internal- and external-facing user interactions, he says.

Chatbots save time, are easy to interact with, reduce error rates in tasks involving humans and improve customer engagement, according to Gartner.

See related Chatbots: AI’s First True Use Case For Enterprise Mobility

Chatbots are also becoming, well, chatty. They are more conversational and able to better understand both written and spoken text. The Deloitte report also finds they have different degrees of intelligence ranging from being able to answer questions to having the full capabilities of a customer service employee. 

For example, there are FAQ chatbots, which are the simplest form of chatbot. They can understand questions and provide relevant answers. At a slightly higher intelligence level are virtual assistants, which have more integration with enterprise systems and consequently, can perform basic functions like looking up personal information. At the high end of the intelligence spectrum are virtual agents, which the Deloitte report says can replace a human employee. These bots have the ability to handle complex dialogs, processes and security protocols.

What’s the Right Bot For You?

Like with any software platform, if you’re looking to deploy a chatbot in your organization, do your research first. The growing number of vendors and product features can make the process overwhelming. There are several considerations:

  • Assess the impact on resources. Do you want to empower an existing process or replace it and build something from scratch?
  • If the bot is for a customer-facing application, make sure it represents your organization in an appropriate way.
  • Make sure to use a bot that fits your needs. You don’t want a simple chatbot to do complex tasks. Conversely, you don’t need a virtual agent to something simple.
  • When integrated with back-end systems, a chatbot increases in value.
  • Start with a proof of concept to prove value and then add in complexity incrementally.
  • Find out whether your customers are really interested in using bots and if so, what their expectations are.

Enterprise Use Cases

Besides call centers, chatbots are already being used to deliver an enterprise message request to an employee when there is a change in a back-end record, triggering an event. The request asks the employee to respond with “approve,” “deny” or “defer” in the app, according to Gartner.

They are also being used in tandem with Internet of Things (IoT) technologies for inventory management in warehouses. Workers are notified when a product is out of stock or if a shipment has arrived containing those out-of-stock items. 

When companies need to handle a large volume of customer – and employee requests, chatbots can be a game changer. And you can expect to interact with more of them in the years ahead.

Gartner estimates over half of medium to large enterprises will have deployed product chatbots by 2020. The firm also predicts that 25 percent of customer service and support operations will be using virtual customer assistants by 2020, compared with less than two percent in 2017.