Not all AI-driven chatbots designed for customer support are equally well-developed.
Take, for instance, AirAsia’s introduction of AVA in 2019, an AI-enhanced user support application. AirAsia’s CEO, Tony Fernandes, recently acknowledged that AVA was labeled as Southeast Asia’s “most disliked AI chatbot” due to the high volume of customer complaints and proposed solutions it generated.
AVA is not alone in the realm of disfavored robotic entities. She has since been replaced by Bo, a second-generation AI bot. A mere 25% of consumers who interacted with AI bots between December 2022 and February 2023 expressed willingness to utilize them in the future for issue resolution, as per Gartner’s findings.
The rapid expansion of relational AI, which seeks to humanize bot interactions by understanding customer queries in a contextual manner and responding in a more natural language, holds the promise of altering this perception. McKinsey estimates that conceptual AI could save organizations globally up to $404 billion annually.
With a projected enterprise expenditure of $1.3 trillion on generative AI by 2032, many businesses are eyeing the Customer Experience (CX) jackpot. Capgemini reports that 63% of retail enterprises are exploring the integration of generative AI to enhance customer service, while Gartner’s research indicates that nearly 40% of companies across various sectors intend to prioritize CX in their investments.
To maximize returns with minimal risks, where should businesses truly invest in conceptual AI strategies? Customer service experts suggest that generative AI can enhance both agents’ and customers’ experiences in the following four key areas:
“Copilot” for real-time support
The repercussions of subpar complaint handling, characterized by ineffective chatbots, prolonged wait times for human assistance, and inexperienced agents, could potentially cost companies future business and jeopardize revenues amounting to $887 billion, according to the 2023 National Customer Rage Survey.
Maeve Condell, the customer success lead at Ultimate, an AI-powered virtual agent platform, believes that generative AI, in the form of a digital assistant or “AI copilot,” could alleviate this pain point for human call center workers.
A conceptually AI-trained assistant can facilitate immediate access to information for call center employees or suggest responses by tapping into the customer database, Condell explains. “We can provide auto-generated responses as a starting point for the solution, which the agent can quickly review, modify, and deliver.”
By reducing the risk, an AI copilot can enhance the efficiency of contact center staff by granting them swift access to practical solutions and offering language suggestions to effectively convey information. This is particularly beneficial for less experienced employees.
A recent Stanford study revealed that contact center agents equipped with copilots experienced a 14% boost in productivity, with novice or less skilled workers reaping the most significant benefits. The study’s authors assert that conceptual AI levels the playing field, mitigating output discrepancies and substantially aiding lower-skilled staff members.
Development and training of chatbots
Present-day bots often rely on keyword searches to extract information from a knowledge base and present it in a robotic manner, posing the sterile question, “Does this resolve your issue?” The communication style is overtly objective and often misses the mark.
The linguistic capabilities of conceptual AI bots can enhance existing customer service bots, even if direct customer interactions without human intervention may still be challenging. Think of it as a sophisticated “My Fair Chatbot.”
Benedikt Schönhense, the co-founder and head of data science at Springbok AI, views the integration of conceptual AI into the bot creation process as one of the most promising and least risky applications currently available.
Businesses can leverage conceptual AI to create sample dialogues, rephrase potential customer inquiries, and automate a significant portion of the training process, Schönhense explains. Moreover, they can enhance an existing chatbot using relational AI by simulating user inputs with varying levels of detail and human tester prompts.
The beauty of this approach lies in the ability to tailor the conversational style and tone to align with the company’s desired image, ranging from assertive to casual, depending on the user base.
Tracking interactions across help tickets
The frustration of repeating the same issue to multiple agents is a common ordeal. Personally, I’ve encountered scenarios where I had to rehash the identical problem with three or more representatives just to seek a resolution, as noted by Schönhense.
Relational AI’s capacity to generate and contextualize responses is a valuable asset that businesses can leverage in customer service, effectively eliminating this pain point for both customers and support staff members.
This feature proves particularly advantageous when interactions span various platforms such as phone calls, emails, web chats, apps, and social media engagements.
Vijay Vittal, a product development executive at LocoBuzz, an AI customer support platform, points out that a customer service agent might need to sift through extensive notes from multiple interactions across different channels to grasp the customer’s predicament fully. Rapid case summaries can be generated using conceptual AI, condensing complex scenarios into concise reports to expedite resolution.
Onboarding and training for new call center hires
The staggering 38% turnover rate for call center employees in 2022 underscores a critical challenge in the industry. A survey conducted by Amazon revealed that 55% of call center agents felt inadequately trained to deliver high-quality customer service.
Just as conceptual AI proves invaluable for chatbot training, it can also serve as a training tool for call center staff through simulated conversations, preparing them to act as copilots and handle various scenarios effectively.
This low-risk, high-reward approach to staff education can significantly accelerate the onboarding process for new or inexperienced agents, as highlighted by Condell. “Utilizing this tool online offers a stress-free method to train before engaging with customers directly,” resulting in a substantial reduction in the time required to educate fresh support agents, as exemplified by one of Ultimate’s clients with intricate internal processes.
In conclusion, human oversight remains paramount in maximizing the ROI and ensuring the safe deployment of this technology. Embracing relational AI as a guiding principle can help businesses extract maximum value while maintaining a human touch in customer interactions.
