What is the role of generative AI in modernising the customer experience?

Artificial Intelligence (AI) is revolutionising many aspects of business operations, particularly customer service. The use of bots to handle high-frequency, low-complexity tasks is well established.

Most customer interactions now start digitally, with AI-powered bots or intelligent virtual assistants (IVAs) accounting for around 85% of these interactions, according to ZK Research. With the advent of generative AI, bots can handle even more complex transactions.

Despite these advances, there are situations where generative AI falls short, particularly when customers are unable to resolve their queries effectively. This underscores the importance of human involvement in customer service, especially when it comes to complex issues and problem solving.

THE IMPACT OF AI ON CUSTOMER EXPERIENCE

The customer experience is now the most important brand differentiator, surpassing other factors such as price and product quality. According to data from ZK Research, two-thirds of millennials have changed their brand loyalty based on a single bad experience. Today, 90% of companies compete on the basis of CX, a significant increase from 28% five years ago. Improving the customer experience is a top priority for organisations with digital transformation initiatives.

In short, AI is likely to impact most business or customer interactions. 78% of organisations plan to invest in AI to improve the customer experience. Companies that provide a superior experience continue to be patronised by customers, who remain loyal to the brand. On the other hand, those that don’t provide a good experience struggle to retain customers.

Generative AI makes it easy to quickly personalise customer interactions thanks to its ability to analyse vast amounts of data. Natural language processing (NLP) tools powered by generative AI, such as OpenAI’s ChatGPT, are democratising AI beyond the realm of specialised companies. Such tools are now available to everyone.

LINGUISTIC MODELS AND EVOLUTION OF INTELLIGENT VIRTUAL ASSISTANTS

Intelligent Virtual Assistants (IVAs) have had a mixed reception, ranging from scepticism to outright dislike, mainly due to their limited intelligence without AI. An IVA is another form of Interactive Voice Response (IVR) which, despite its role in customer service, has been more frustrating than helpful to users. One of the main problems was the lack of integration between systems, with agents failing to capture the information entered into an IVR. The same was happening with IVA.

With the implementation of large linguistic models, IVAs have become more conversational. Instead of using specific phrases, users can now communicate using natural language. Some IVAs have improved to the point where human agents are preferred in certain scenarios due to their speed and lack of waiting time.

Applying AI to IVAs has made them smarter, more accurate and more capable, resulting in a significantly improved customer experience. People can speak in their own voice and be understood, without having to adapt their speech patterns to the needs of the software.

CHANGING METRICS FOR CALL CENTRE PERFORMANCE

In the age of AI, traditional performance metrics such as average handle time and first call resolution are becoming increasingly irrelevant. Artificial intelligence and advanced routing systems have enabled contact centres to handle complex transactions such as mortgage processing in a single call, improving customer satisfaction and potential revenue generation.

In addition, the customer journey is now a key focus for call centres. The new performance metric ensures that customers who need human intervention reach an agent quickly, while others are served efficiently by AI. This represents a significant shift from high frequency, low complexity interactions to high quality, high engagement interactions.

THE STRATEGIC APPROACH TO AI ADOPTION

Organisations looking to adopt AI should take a strategic approach, which includes identifying key business use cases. For certain use cases, such as understanding the nuances of customer queries, generic AI may not be sufficient. This is where ‘custom AI’ comes in, designed and trained to handle more complex queries. Therefore, companies may need to build AI models using unique data sets that meet specific needs.

Human oversight remains essential in the use of artificial intelligence, especially in critical applications such as healthcare and finance. All companies should have policies in place to implement technologies such as ChatGPT effectively and ethically.

Companies should ask their vendors how they plan to integrate AI, and assess their own data to determine if they are ready to take advantage of it. Although it’s still early days, there will be many new opportunities and challenges ahead as we move into an exciting future with artificial intelligence at the forefront.