AI, Chatbots, Live Chat

Automation First: 60% Customer Interactions Automated by Putting a Chatbot in Front of Live Chat

By 5 Minute Read

Automation First Where Appropriate…Then Live Chat

Live chat as a customer support channel has grown in importance as online consumers seek more immediate responses to their queries and the resolution of any problems they are experiencing. For e-commerce businesses, it has become a must-have as a means to provide the appropriate assistance to a customer in order to convert an online sale. 

By hitting the live chat pop-up button on a vendor’s website or mobile app, a customer effectively can interact with a live agent via messages without having to call a support number and wait on hold, or send an email and wait a day or more for a response. The more immediate and pressing a consumer’s issue is the more suitable a live chat interaction is in meeting their needs.

Take the example of someone making an online purchase of sporting equipment who may need help with sizing or accessories, or a customer having difficulties getting their credit card to process a purchase. Having a live agent at the ready to help them through this may avoid them abandoning their purchase and help expedite the sale. Of course, this depends on the quality of the interaction and the ability of the agent to resolve the issue and satisfy the customer.

However, despite the popularity of live chat as a customer service channel, there is no mistaking the fact that it relies on the availability of trained human agents, incurring the labor costs, scheduling issues, and problems associated with scaling a 24×7 digital web or mobile front-end. Considering the significant operational costs incurred by live chat, it is worth identifying where its benefits will have the biggest impact rather than using it for every customer and every potential query or issue they may have.

Why Your Live Chat May Need the Help of an AI Bot to “Automate First”

In this respect, serving customers with priority issues and ones that are best resolved immediately and through the intervention of a live agent are the best targets for your live chat investments. Other less urgent issues or those not requiring human intervention can be automated using a conversational AI solution i.e. a chatbot, that can escalate or hand over to a live agent when needed or when requested by a customer. 

So why use your human customer service agents to respond to repetitive and routine customer queries when you could use AI to automate customer service and increase your containment rate?

 

Automation-first Fro LiveChat

 

This is what our client, WorldRemit, discovered as the Covid-19 pandemic hit, generating increased customer request volumes to their live agents who often were spending valuable and costly time answering the same queries.

Here’s their story.

How WorldRemit Responded to the Impact of Covid-19 on their Contact Center

WorldRemit Logo

As background, WorldRemit is a leading provider of international remittance transfers,  enabling their 5 million customers to transfer funds quickly and securely to friends and family in over 150 countries worldwide. Before the pandemic, many of WorldRemit’s customers performed these transfers in-person at designated money transfer locations.

However, as COVID-19 hit and lockdowns went into effect, these premises closed forcing customers to turn to a digital means to transfer money home. Because many customers had never used a digital money transfer service before, the demand on WorldRemit’s customer support team spiked. This happened in conjunction with agents having to work remotely.

The company knew that they would have to find a solution so that they could respond and resolve common customer queries at scale and alleviate the pressure on live chat agents.

 

As Justin Sebok, Senior Product Manager at WorldRemit put it:

“Because of the sheer volume of calls we were receiving during the lockdown, we needed to move away from having an agent on the phone with a customer talking them through the process, especially when they were repeatedly answering the same questions. We needed to find a different method.”

An “Automation First” Digital Assistant Solution Handles the Top Customer Service Queries

An AI bot that could automate 60% of incoming requests and route the remainder to live chat was the solution. With time being of the essence, this needed to be rapidly implemented and serve a large customer base. WorldRemit selected ServisBOT to help them get a digital customer service assistant in place as quickly as possible and have it focus on resolving simple queries on both the website and its mobile app.

The purpose of the digital assistant was to help people to resolve their queries satisfactorily, rather than simply diverting them from contacting customer service agents. When customers are directed to upload a document or complete a specific task, they report a higher level of satisfaction than if they are simply directed to the FAQ page, where they have to search through lots of information to answer their query. 

WR Assistant and Quote

The bot was designed to address the top most common queries being fielded by the customer service team. If none of these met the requirements of the customer they had the option to choose a different question where they could self-serve with the FAQs from the website or choose to be transferred to a live chat agent. Once common questions were successfully being answered via the bot, the team added a feature that allowed customers to select self-service options providing real-time information about queries. These include questions such as tracking refunds, checking transfer status, and options for resolving related issues, such as uploading a driving license required for identity verification. 

Fast Rollout of the Digital Assistant Followed By Continuous Improvement

The first version of the AI digital assistant was up and running within 2 weeks. But the team continued to work on a number of improvements as soon as it was launched. Weekly meetings were conducted with cross-functional teams to capture fresh business requirements and key learnings from customer interactions with the bot. Changes were implemented throughout the product development process to address any recurring issues that were driving customers to seek live chat support. Chat transcripts were analyzed in order to discover what customers asked. These processes helped to provide insight into how queries could be automated on the front end. 

The limited screen space on mobile phones presented a challenge when delivering customer service through a mobile app. WorldRemit found that the in-app digital assistant made better use of the screen space because customers self-selected the relevant questions and were automatically presented with brief responses. 

Justin Sebok, the Senior Product Manager at WorldRemit summarized this process as follows:

“ServisBOT provided a lot of suggestions about what was possible. They were very responsive when we requested changes and catered to our needs to ensure that we built something that really worked and got the results we needed.”

An Automation-First Approach Yields 60% of Chats Handled by the Digital Assistant, 40% by Live Chat

To measure how well the digital assistant was working, WorldRemit worked with ServisBOT to develop a daily reporting system that tracks how many customers resolve queries with the bot, versus how many are ultimately handed over to a live agent by it.

Automation First: 60% Customer Interactions Automated by Putting a Chatbot in Front of Live Chat 1

 

 

60% of those who engage with the digital assistant don’t need to speak with an agent as they are now being pointed towards useful information and self-service features on World Remit’s site.

 

This has dramatically reduced service costs for WorldRemit, improved customer satisfaction, and reduced the pressure on customer service staff since only 40% of customers escalate to a live chat agent and they do so using a transaction ID.

Bot Interactions

 

30% of all visitors to WorldRemit’s customer support pages engaging with it as

It handles 140,000 interactions per month. 

 

 

“We would have had to hire a lot more people. By using the digital assistant, we’ve been able to divert calls and greatly reduce the number of conversations initiated via live chat.  We’ve removed a lot of repetitive manual effort that didn’t add much customer value, and we’re now continuing to add more automated options for our customers.”

Following the success of this digital assistant in automating live chat, the company plans to further integrate the digital assistant with their business systems to provide additional transfer features and status information to customers and payees.

Learn more about this case study on automating live chat.

Close this Window