Rebecca Gibson |
Almost everyone has experienced the frustration of contacting a customer service department for help, only to be forced to wait in a long queue, answer the same basic questions multiple times and then discover that the person they are speaking with doesn’t have the information to be able to assist them anyway.
“The biggest challenge for any contact centre agent is being able to quickly and confidently provide the most helpful answer to the customer the first time they ask the question,” says Enrico Karsten, CEO of cloud-based contact centre and dialogue management platform provider Anywhere365. “Often, agents don’t have the in-depth knowledge to resolve a specific query – and they don’t know where to find the contextual information they need to help them, so the process becomes unnecessarily long and arduous for both the agent and the customer. This can be detrimental to organisations as just one bad experience can prompt customers to go elsewhere and never return.”
According to Karsten, there are two key technologies that can help businesses to improve and expedite the customer service experience: artificial intelligence and machine learning.
“AI can be used to analyse interactions and pinpoint optimal responses, as well as to deliver detailed contextual information to agents in real time so they can complete their interactions as quickly as possible with the highest success rate,” he says.
In addition, AI can be used to perform sentiment analysis to help agents understand how their behaviour might be impacting the customer experience.
“Agents can usually recognise when a customer is becoming frustrated, but they don’t always know which actions to take or what to say to quickly de-escalate and resolve the situation,” says Karsten. “But if an AI engine is continually analysing customer sentiment and sharing Net Promoter Scores (NPS) with the agent in real time, they can easily gauge how well they’re performing and whether their answers are satisfying the customer. Plus, they can follow the AI-generated suggestions to improve the interaction and achieve a better outcome.”
By implementing machine learning technology, enterprises can eventually create recommended dialogue flows for agents to follow in specific scenarios.
“If, for example, 20 agents have all answered the same question from different customers, we can use AI to analyse their responses and determine which one had the highest NPS score and success rate,” says Karsten. “We can look at what the agent said, what information they referred to when looking for the answer and where they found it within the organisation. From there, we can use machine learning to create standardised scripts for agents to follow the next time a similar question arises in future.”
Once organisations have these scripts in place, they can offer human agents the opportunity to let an AI-enabled bot take over during certain parts of the conversation. “It might be that the bot can autofill some of the responses in a text chat, for example,” says Karsten.
Crucially, though, the agent will always retain full control throughout the entire conversation.
“If the bot is leading the conversation and an issue arises – such as the customer entering their address incorrectly multiple times – it will notify the human agent so they can resume control and offer additional assistance,” says Karsten. “The AI and machine learning tools will enable the bot and the human agent to work together in harmony, ensuring the conversation is so seamless that the customer will never feel like they’re interacting with a bot at any point. They will simply be satisfied that their query was resolved in a quick and efficient manner.”
AI and machine learning tools allows human agents and bots to work in harmony (Image: iStock/RossHelen)
Although AI-powered agent assist services are still in the nascent phase, Karsten predicts they will rapidly become a mainstream solution for many businesses.
“Many enterprises are already experimenting with using AI-powered chat or voice bots to help them manage basic customer queries automatically, but the real opportunity lies in developing AI bots that help the agents,” he says. “Anywhere365 is already using this technology for text chats, but the next step is to develop similar capabilities for voice chats over platforms like Microsoft Teams – and it won’t be long until we reach that stage.
“In this scenario, the agent would be able to see their real-time NPS scores, suggested dialogue flows, useful information and more on their screen while talking to the customer. This will really show the power that AI and machine learning have to transform the query resolution process and both the agent and employee experience.”
This article was originally published in the Winter 2022 issue of Technology Record. To get future issues delivered directly to your inbox, sign up for a free subscription.