Finding the human touch in a sea of bots

In the digital age, a customer’s bad experience with a brand can easily go viral with a single social media post. We’ve heard countless stories of such unpleasant experiences, be it with an airline or telecom services provider, and it can truly tarnish a brand’s reputation when a lot of people relate to these stories. A common denominator to bad customer experiences often boils down to communication, or the lack of it, which is why at the very basic level, it’s important for organisations to make themselves accessible to customers through multiple communications channels. Right now, delivering great customer experiences and having a strong customer communications strategy have never been more important.

Contact centres, which offer a direct line of communication, are an important customer touchpoint. This is also the place where a company’s reputation can often be won or lost. Customers, when dialling into the contact centre, are almost always looking for immediate answers to questions and they want issues resolved quickly.

To address the large volume of requests coming in at speed, organisations need an always-on approach to customer engagement and service, so chatbots have become a popular solution.

How chatbots succeed, and where they fail

Empowered with artificial intelligence (AI) and machine learning (ML) technologies, chatbots are very effective at answering simple questions and resolving straightforward issues such as resetting passwords, updating addresses, or checking order and service status in near real time.

However, their ability to provide good customer service is often limited as soon as the complexities of human emotion come into play. In fact, many people are likely to have experienced frustrating conversations where chatbots fail to identify the nuances of requests or feedback, and direct the customer through a series of actions that ultimately do not resolve anything. Also missing is the ability to offer sympathy, support and promises of prompt action — something many frustrated customers seek.

With customer expectations now at an all-time high, such experiences are detrimental to the customer experience and the brand’s reputation. Beyond the basics like quality service and fair pricing, customers today expect personalised interactions and connected experiences across all platforms. They want to feel that they are important and worthy of attention from the brand they are engaging with.

In this sea of bots, brands will need the ‘human touch’ to stand out from competitors. The challenge is how organisations provide a real, human-like experience in their customer service delivery without compromising on organisational efficiency.

Where ChatGPT fits in this new era of customer engagement

The generative AI chatbot that has taken the world by storm, and its rivals such as Google Bard may be part of the solution. With natural language processing (NLP) capabilities, tools like ChatGPT are able to understand natural language spoken by humans, and respond with coherent, fluent and human-like responses in a fraction of a second. It is as if your queries are being answered by a person instead of a machine. With the updated GPT-4, OpenAI’s new large multi-modal model that can solve difficult problems with greater accuracy, and accept both image and text inputs, the AI-powered chatbot is now more powerful than ever.

Through the ability to understand users’ intent, and provide recommendations in an intelligent and conversational manner, ChatGPT is transforming the way businesses interact with customers. Used in tandem with automation solutions, organisations could potentially accelerate customer engagement workflows, while increasing personalisation in their customer service delivery. A couple of simple examples may be:

Use case #1: Analyse customer feedback

While many businesses may already rely on software robots to collect and collate customer feedback received through online chatbots, they can deepen their insights further by using ChatGPT to analyse that data. They can get software robots to send a set of customer feedback to ChatGPT, and enter the following prompt: “Please identify the sentiment of this feedback, assigning it as positive, negative, or mixed”. Once the robot receives the answer from ChatGPT, it can then collate the positive, negative and mixed answers for further review. Based on the sentiment, useful feedback can be processed downstream and shared directly with the team for consideration and action.

Use case #2: Create a customer response email

ChatGPT can also write appropriate responses to customers who have shared negative feedback. This whole process can even be automated from end to end. When negative feedback is received, a software robot shares a prompt and the text of the email with ChatGPT, which replies with an appropriate response message. The bot then validates with customer support before sharing it with the customer. This shortens the response and resolution time for customer feedback, which can go a long way to improving the overall customer experience.

#3: Evaluate customer service conversation

ChatGPT can rank the quality of support provided by customer service agents in online chats. ChatGPT does this by returning a satisfaction score, and if the score falls below a certain threshold, a software robot can automatically escalate it to a manager for review. This lets managers continually enhance the training of their support teams.

What’s next?

As much as we’re seeing rapid progress with AI/ML technologies, the world is not quite ready to see human customer service agents completely replaced by robots. Contact centre hubs in Australia and across Asia will not entirely be replaced just yet. There are still instances where we need a real human behind the screen. However, it is interesting to note that automation will likely play an increasingly critical support role in the near future.

Where automation is adding the most value is by streamlining contact centre processes, giving agents more bandwidth to address more complex problems. For example, to address a single customer request, an agent may need to switch back and forth between multiple systems, which can be inefficient and distracting. With the help of software robots that collect information from disparate systems and present it in a single, consolidated interface, agents can easily address the customers’ needs, while focusing on delivering great customer experiences. This results in a better employee experience for the contact centre agent and better relationships with customers, which ultimately supports a stronger positive brand sentiment for the company.

Image credit: iStock.com/PhonlamaiPhoto