GovLoop

How to Get Chatbots Up and Running

In today’s digital age, people are not only comfortable with online self-service options such as chatbots, they expect them. Gartner estimates that by 2022, 85% of customer service interactions will start with self-service, compared with 48% in 2019. This extends to government engagements, too, where they can be a boon to overburdened contact center agents and serve as a bridge between digital transformation and CX. Here are four steps to fielding chatbots.

1. Understand Chatbots’ Benefits

Agency contact centers have a finite amount of resources, including the number of people trained to handle and respond to questions. They can handle only so many requests at a time, which can leave employees and constituents alike feeling frustrated.

But chatbots allow for an unlimited number of interactions simultaneously, said Tony Pearson, Solution Consultant Leader for Public Sector at Genesys. Specifically, they can handle frequently asked questions, point users to needed websites or forms and assist with online transactions. With such self-service options, call volumes decrease and human agents have more time to focus on more complicated requests.

What’s more, chatbots are relatively easy to implement. “We can show you in an hour how to stand it up and soon after put it in production,” Pearson said. “I don’t see setup as a massive overhauling task, as long as you have the right technology in place.”

What to do: Identify contact center bottlenecks and bring in stakeholders to set a strategy for alleviating them with chatbots.

2. Lay a Technological Foundation

Pearson defines the “right technology” as forward-facing solutions, such as Genesys Cloud. Legacy technology, such as private branch exchange systems, which support voice environments, not digital ones, don’t work for a multimedia, omnichannel experience.

But don’t get rid of voice, either. Instead, apply the bot to interactive voice services, too.

“Some people may be very technology-averse and some people may be super technology-savvy,” Pearson said. “You want to make sure that you’re also working with a technology that is able to transition those channels.”

When you have a single pane of glass for how people consume your agency’s services, they will get the same answer no matter where they go. “That bot is really going to allow for that consistent experience because now you’re not getting five different human agents that are answering about the status of a complaint,” he said. “You have one bot that goes into the same system and tells the same exact information every time, no matter what channel you go in.”

What to do: Determine the multiple channels the public uses to engage with your agency.

3. Foster a Growth Mindset

Getting support for chatbots is more of a cultural issue than a technological one. Because workers often associate AI-driven technology with job replacement, it’s important to make sure your call center’s human agents see chatbots as a job enhancer.

“That easy stuff gets handled by the bot,” Pearson said. “The agents are now going to be, more likely than not, handling more complex problems on a regular basis versus handling simple things. Now you’re really challenging people in the contact center.”

What to do: Make sure human agents know the bots will enhance, not replace, their work.

4. Plan Ahead

Because chatbots are AI-driven, they never have to become static. As more information becomes available, agencies can train the bots to provide it proactively – before they even become the subjects of frequently asked questions. With predictive analytics, agencies can anticipate what constituents might start looking for, driving internal and external CX that much further.

“When you’re iterating the bot, a little bit of that is from the human process and understanding from an analytics standpoint how we can make improvements,” Pearson said.

“But also, technology and machine learning get better over time, so as a given agency is collecting data on their own contact centers, bot vendors, including us, are getting better – we’re learning lessons over time, and we’re able to tune our machine learning to be able to handle more and more and more,” he said.

What to do: Treat your first venture into chatbots as a starting point, not a finished product.

This article is an excerpt from GovLoop’s recent playbook, “Chatbots Take Constituent Engagement to New Levels.” Download the full playbook here.

Exit mobile version