AI in Customer Support: Separating Potential from Practicality

Pull Quote: “Generative AI is not meant to replace people. It’s meant to provide information that actually can help [customer service agents] meet the needs of that customer faster, better, and smarter.” - Jennifer Quinlan, global managing partner and customer transformation consulting services leader at IBM iX.

With innovations in artificial intelligence (AI) attracting considerable attention, AI has become a significant trend and selling point for businesses. Sixty-three percent of organizations are increasingly adopting AI over any other digital technology, integrating it into various workflows to stay ahead in the business landscape. One of the main adoptions of AI is in customer service, with 56% of business owners focusing on customer-centric AI. 

While there is an abundance of enthusiasm about the future of AI in customer service, there are also concerns about the realistic current benefits of AI in customer service. Despite advancements, AI in CX (customer experience), is still young and yet to mature. That’s where the question of what we can do with customer-centric AI today comes up.

Abinash Tripathy, founder and Chief Strategy Officer at Helpshift, spoke about AI helping with customer service phone calls and remarked, “AI is very good at doing classification. An intent extraction is based on just a few sentences or words. But it’s not good at maintaining a sustained conversation the way you and I are communicating right now, and that’s where it all falls apart where people are trying to apply it to replace a human and try to sustain a long, drawn-out conversation.” He explained how it will take time for these types of applications to be more dependable for complex situations. About 80% of customers have simple queries, and an AI customer service bot can handle them. The remaining 20% will still need a human to assist them for the time being.

Understanding what is hype and what is practically possible is crucial to using AI's competitive edge in customer service correctly. In this article, we’ll differentiate the two, discuss the best strategies for integrating customer-centric AI, and look at cases where AI in CX shines. 

What is Possible with Customer-Centric AI?

Businesses are thrilled about AI's possibilities, often hyping its transformative power in bold marketing statements. However, these claims sometimes stretch beyond today's technological achievements. Gartner's 2018 forecast brought some realism to the conversation, predicting that 85% of AI projects would face errors up through 2022. Despite progress, it's clear that AI still has hurdles to overcome, including achieving reliable accuracy and further refinement. 

This leads us to consider what businesses can realistically achieve with AI's current capabilities and how leaders can leverage these advancements. Whit Andrews, vice president analyst of Gartner, says, “Look at how you are using technology today during critical interactions with customers — business moments — and consider how the value of those moments could be increased. Then apply AI to those points for additional business value.” Andrew’s advice shifts the focus from the theoretical potential of AI to its practical applications. It highlights the need for businesses to identify areas where AI can make a real difference, suggesting a more measured and incremental approach to integrating AI into business practices.

Considering customer-focused AI tools like chatbots, text generation, and visual marketing tools, the goal isn’t to replace human roles and enhance them. These tools aim to speed up processes and boost human productivity. Businesses can more easily appreciate its benefits by setting realistic expectations for AI in CX.

‍

A great instance of this is how NØIE, a skincare company, used customer-centric AI to transform their approach to customer feedback. They turned to Lang.ai, an AI tool designed for natural language processing and machine learning, to automate the sorting and handling of customer inquiries. They slashed their response time by an impressive 89%. Now, every customer message is quickly categorized, assigned a level of importance, and forwarded to the appropriate human responder. This step alone massively boosted their team's efficiency. It's a solid example of how even straightforward AI in customer service applications can significantly improve business operations.

Challenges Enterprises Face When Adopting Customer-Centric AI 

Tripathy also highlighted the challenges enterprises face when adopting customer-centric AI solutions. He started by saying that developing robust AI chatbots is complex. In line with this, 80% of consumers who have interacted with a chatbot in the last 12 months have said that using chatbots increased their frustration.

Tripathy continued by emphasizing the necessity of large companies driving significant AI breakthroughs. Smaller businesses often need more resources, data, and computational power to push the boundaries of AI technology. Larger tech giants, such as Google, Amazon, Apple, and Microsoft, have the advantage of vast datasets and substantial research teams, which enable them to make significant advancements in core AI algorithms. 

In contrast, smaller businesses face limitations in innovating with AI due to data scarcity, cultural resistance to change, and a lack of talent density. These challenges can make it difficult for smaller organizations to utilize AI effectively and keep pace with the rapid developments in the field.

However, hurdles are meant to be crossed. Some companies have battled against the odds and come out on top. Let’s take a quick glimpse at a couple of Fortune 500 company case studies:

• Amazon Q - Amazon has developed a tool that uses generative AI in customer service to assist agents with recommended responses and actions for faster customer support. Steven Dickens, the vice president and practice leader at the Futurum Group, expressed, "AWS Q will be a game changer for AWS customers who have many service options, often overlapping to navigate. AWS has resisted the urge to make an AI assistant for each service in its portfolio and, as a result, I expect to see Q become widely adopted in the months ahead by both developers and cloud admins alike." 

Amazon Q is an excellent example of how Amazon remains resilient in facing challenges by following key principles. They believe in solving customer challenges rather than merely implementing exciting technology. Also, by building step by step, Amazon ensures that each innovation, like AWS Q, is grounded in practicality and geared towards improving specific aspects of the customer experience.

• Bank of America's Virtual Financial Assistant, Erica - Since launching in June 2018, Erica has surpassed 1.5 billion interactions with Bank of America clients, with more than 37 million clients engaging Erica to help manage their financial needs. David Tyrie, Chief Digital Officer and Chief Marketing Officer at Bank of America, explains, "It took four years to reach the first billion interactions, but client engagement with Erica is accelerating, and it will easily exceed 2 billion within a few months. Our clients appreciate how simple and intuitive Erica is, and they trust it's only using proprietary, vetted data."

One of the challenges related to Erica was getting the chatbot to understand the casual and varied ways customers talk about their finances. For example, people often use slang like "dough" to refer to money, which initially puzzled Erica, who was programmed to understand more traditional financial terminology. This challenge is being addressed with attempts to help Erica understand different casual expressions. Like any other AI model, regularly monitoring and fine-tuning Erica is crucial for keeping the chatbot up-to-date and accurate, ensuring it remains valuable and reliable for handling everyday financial tasks.

Integrate Customer-Centric AI To Improve Your Operations 

According to Manish Goyal, the vice president and senior partner of global AI and Analytics at IBM Consulting, current AI technologies like generative AI can significantly improve customer service if appropriately integrated. Let’s understand the three key areas where he says customer-centric AI could dramatically improve customer service.

The first area is self-service, where customers are given tools to serve themselves. Virtual agents or chatbots serve this purpose, and over the years, they have gotten good at directing customers down a predetermined journey.

The second area where AI can improve customer service is by augmenting the human agent, whether in the contact center or the field. Agents in contact centers spend a lot of time searching knowledge bases to resolve customers' queries. This can be optimized using generative AI. It can improve information retrieval from the knowledge bases and present it to agents in a summarized way. The time customers are on hold can be reduced, improving their experience and allowing agents to handle more calls.

The third area is contact center operations. For a call center with tens of thousands of agents, gaining insights into what's happening across all the conversations between agents and customers is difficult and expensive. Using AI, companies can review transcripts of every call and analyze how and why agents are taking a long time to handle specific customer queries. This process helps companies find and resolve a problem's root cause faster.

Research shows 81% of customer service executives invest in AI in CX. Gartner reports that 98% have already integrated or plan to integrate AI into their customer engagement strategy. Among the 45% who have already incorporated AI, the majority (80%) use conversational agents like chatbots to help transform their customer engagement.

‍

AI Agents and Customer Satisfaction: Success Stories from Klarna, ING Bank, Erricons and T-Mobile

The benefits of AI in customer service are clear, and companies like Klarna, ING Bank, T-Mobile, and Ericsson are reaping them. These companies have seamlessly integrated AI into their customer service, streamlining their processes and setting new standards for customer satisfaction. Let’s examine how these companies have adopted customer-centric AI in their respective industries. 

Klarna is a company that provides global payment and shopping services. Their AI assistants, powered by OpenAI, work equivalent to 700 full-time agents. These ‘AI agents’ are on par with human agents regarding customer satisfaction scores. They are good at tasks like errand resolutions, leading to a 25% drop in repeat inquiries. Customers can have errands solved in less than 2 minutes compared to 11 minutes previously. Another exciting feature of this chatbot is its ability to communicate in multiple languages. It can communicate with customers in more than 35 languages. Thanks to this language support, Klarna has seen a massive improvement in communication with local immigrant and expat communities across all markets. Having extensive language capability enhances customer satisfaction and significantly broadens Klarna's market reach by catering to a diverse customer base.

Another area where AI in CX can be used efficiently is in banking. ING Bank has been using conversational AI since 2017. They started with their first chatbot, Lionel, which, due to its popularity, was quickly followed by Marie and, finally, Inge. ‘Thanks to these bots, service agents are freed to do more interesting stuff,’ says the program manager of ING Bank, Tim Daniels. He also claims that bots like ‘Maire’ are integrated with Facebook Messenger to carry out communications there, underlining the flexibility and adaptability of AI solutions in meeting customer needs on platforms they frequently use.

Moving to the telecommunications sector, T-Mobile and Erricson used customer-centric AI to solve problems they had regarding their ordering process. Up to 17% of all orders were falling out, anything from upgrading a price plan to getting a new phone, which had a detrimental effect on the customer experience. This was a huge problem. And the results of using anomaly detection as the solution were outstanding. They managed to reduce order fallouts by 95%, the time to identify an issue by 90%, and the order-to-activation process was shortened to 5 minutes or less for almost every order.

The Benefits of AI in Customer Service Are Real When Expectations Are Realistic

The importance of customer-centric AI is on the rise. Many businesses are looking to invest in AI for better customer service management. AI-driven tools offer solutions by providing deeper insights and enabling more intelligent decision-making, leading to significant cost savings. However, investments must be made with a pragmatic approach to AI integration. Leaders must focus on achievable improvements and long-term strategic planning and have realistic expectations when adopting AI in business.

Take that first leap forward towards optimizing your businesses with AI-powered customer service tools and chatbots. As companies continue to explore and use AI technologies, the future of AI customer service looks promising.