Future-Proof Your Customer Experience by Embracing the AI Wave

Key takeaways:

• What happens when 2024’s fastest-growing software category (AI) and the hottest business function (customer experience) combine?
• Despite the reality versus expectations drift, businesses struggle to accept that AI is only as good as its execution
• Enjoy a front seat as we explore the realities, learn from use cases of AI in customer experience, and derive an actionable blueprint

What visual do you imagine when you hear “customer experience”? 

Most of you probably thought of the last bad one you had rather than the tenth good one from two weeks ago. You may relate to this unconventional analogy: picture the perfect customer experience as a shape-shifting unicorn. Everyone wants it, but no one has it (not even the tech giants).

However, in the quest for this elusive unicorn, one negative touchpoint has the power to overshadow the entire customer experience. It disproportionately polarizes the experience from delightful to disappointing, leading to negative brand perception and the risk of losing future customers, especially if an unhappy customer shares bad reviews.

This risk is why AI has become mission critical in capturing the unicorn of customer experience: driving efficiency and scaling success. Businesses deploy AI to facilitate recommendations, customer education, support, and sales - either by complementing their existing customer experience tech stack or completely replacing it. 

But what they have failed to do is build AI into the fabric of customer experiences.

The result? A paradoxical landscape where brands wanted utopia, but what they created was dystopia.

The present reality is skewed. While one brand got trolled for having 1,000 people manually support the “AI” in customer experience, recent news headlines show other brands are doubling down:

• Google Cloud, Mercedes-Benz expand partnership to enable AI-driven customer experiences
• The Home Depot taps Google Cloud AI capabilities to enhance customer experience and optimize operations
• Best Buy boosts customer support with generative AI

So, is the sentiment toward AI investments still positive? Or is seamless customer experience using AI just a distant dream?

Let’s scan the industry outlook and take you to G2’s AI City and further down the conversational intelligence alley to distill the complexity. Here's how to get ready and not miss this pivotal opportunity to improve customer experience.

Business leaders are losing sleep over the AI opportunity

According to IBM’s latest report, customer experience strategy tops the list as a change agent for the next two years. However, companies are unprepared for AI. Even though three-fourths of business leaders anticipate AI will transform customer experience, many see it as a huge challenge that adds to the complexity of a changing world.

Source: IBM’s AI-enhanced customer experience

A closer look shows that the top three challenges executives face are integrating AI with existing tech, finding the right AI software to meet business needs, and regulatory concerns.

Source: IBM’s AI-enhanced customer experience

Recognizing these challenges, it’s clear that navigating the complexities of AI adoption is no simple feat for executives. Every once in a while, everyone needs a little help, don’t they? Let us help you get that roadmap right because AI doesn’t have to be scary.

Identifying specific hallmarks of AI’s role in customer experience

G2 has 21 AI categories featuring 1,078 AI vendors, 1,232 AI products, and 37,823 buyer reviews. To correlate this data, our VP of Market Research, Chris Voce, and his team shared a lucid visualization called the G2 AI City. 

Before we proceed and connect the dots, imagine this is NYC (but with AI), and you’re the customer touring this city.

The colors represent AI software categories, a.k.a the neighborhoods. The building height represents the number of software products stacked in that respective category.

The heatmap above shows AI chatbots as one of the hottest neighborhoods marked in yellow, orange, and red. Because chatbots connect with CRM, automation, customer analytics, insights tools, and social media management tools — it’s safe to call it the Times Square of AI City.

If you’re a customer visiting AI City, chatbots are the agents that will convert you from a visitor into a long-term neighbor by creating a valuable customer experience. And we know happy neighbors advocate for their neighborhood, leave good reviews, and bring more friends to the block party!

Tying in the G2 AI City and NYC reference, how can we not mention skyscrapers? 

AI sales assistant is the tallest skyscraper on the block but, ironically, isn’t a hotspot in AI City. Meaning that AI sales assistant software has the most vendors but not a lot of takers. Surprised? Let’s analyze why this is the case.

Why AI chatbots are winning MarTech budgets and customer attention

While AI sales assistants and AI chatbots both aim to enhance customer experience with AI-powered help, they differ in their approach. Sales assistants optimize sales tasks, whereas chatbots engage customers directly.

What gives chatbots an edge is:

1. Its direct customer-facing presence
2. Ability to break language barriers
3. Ability to enable smooth, swift, and effortless global support

The table below gives a comparison snapshot of AI sales assistants versus AI chatbots to help you draw your own observations.

G2's Vice President of Market Research, Chris Voce, shared his insight on the matter.

He further elaborated the value add that AI chatbots bring to customer experience through people empowerment.

Talk is cheap, and chatbots are expensive. To help you see the value AI chatbots bring to customer experience and leap into its strategic planning and execution, here are some use cases.

3 use cases on the evolution of AI in customer experience

While industries, such as healthcare, plan to deploy chatbots in the next two years, IBM identifies financial services and banking as the top industries that use chatbots to enhance customer experience. 

1. Bank of America’s Erica

Bank of America sought to enhance its digital banking convenience by providing over 37 million clients with a powerful financial management tool. Erica, an AI-driven virtual assistant, was developed to personalize, simplify, and respond to a growing demand for accessible and intelligent customer service.

Introduced in 2018, Erica leveraged controlled AI to provide real-time, personalized financial guidance and support. The AI chatbot was meticulously designed to meet a plethora of customer needs — from conducting transactions and tracking spending to intelligent call routing for intricate financial queries. 

The advanced tool harnesses proprietary, vetted data to furnish a secure, intuitive interface. According to David Tyrie, chief digital officer and chief marketing officer at Bank of America, what sets Erica apart is its controlled AI, which is a primary gateway to scaling personalization.

Source: Bank Of America

As of April 2024, Bank of America has executed over 50,000 upgrades to Erica's capabilities.

Over five years, Erica has processed over two billion interactions and has served over 42 million clients, from tracking spending to monitoring subscriptions. It has cemented itself as an invaluable tool for simple and complex financial tasks, empowering customers on numerous platforms. Erica consistently delivers personalized support quickly, with an average 98% success rate of providing answers within 44 seconds.

Source: Bank Of America

2. ING’s generative AI chatbot

Facing the dual challenge of enhancing customer service and handling high inquiry volumes, ING recognized the need to upgrade its engagement through groundbreaking AI technology.

The objective was to reduce customer wait times and improve the resolution rate by launching a sophisticated AI chatbot that delivers immediate personalized assistance, thereby setting a new customer service standard within the banking industry.

ING, in collaboration with QuantumBlack, McKinsey's AI arm, accelerated the development of a bespoke generative AI chatbot within a remarkable seven-week timeframe. 

The solution incorporated a meticulously designed multi-step pipeline that provided tailored responses to customer queries, featured the ability to offer multiple solutions for consideration, and implemented strict risk mitigation guardrails.

The chatbot harnessed a combination of knowledge retrieval from expansive data stores and an intelligent ranking system to propose the most helpful responses. ING used QuantumBlack Labs' innovative tools to create custom guardrails that adhered to ING's stringent standards and industry-specific regulations.

Since its launch in September 2023, the gen AI chatbot represented a substantial leap forward in user experience. In its initial seven weeks, the AI assistant assisted 20% more customers compared to its predecessor, expediting query resolution. This success has helped ING foresee a future reduction in call center load.

ING strengthened over 50 support functions, including their risk, contact center, analytics, and technology departments. This strategic move enabled the company to scale up its generative AI initiatives significantly.

The ING and McKinsey partnership set the groundwork for a scalable model that other ING countries can adopt and implement. This technical foundation is capable of addressing a wide array of generative AI use cases across the ING Group. These endeavors laid down a blueprint for scaling the solution across 10 markets, which could influence the customer service experiences of more than 37 million customers in 40 countries.

Source: McKinsey & Company

3. G2’s Monty

The software buying experience is complex and riddled with a lot of layers, technical jargon, and a labyrinth of listings. Though customers entering the vast landscape often know what problem they want to solve, they don’t necessarily know how to identify the right tool for it.

To address this gap, we innovated Monty, an AI-driven assistant, to transform vague customer inquiries into targeted product recommendations. The goal was to guide without overwhelming.

Tim Silber, Product Manager at G2 Labs, shed light on Monty's initial role in reshaping the buyer's journey, "When a buyer doesn't know where to start, we take their unstructured ideas and requirements and let Monty sort through the hundreds of thousands of products."

Built on ChatGPT’s LLM, a strong prompting strategy, and G2’s existing site content, Monty soon evolved from a simple support chatbot that recommended products into a multifaceted AI sales representative that engages high-intent buyers and guides solution discovery.

To unlock this capability, the team aligned AI's functions closely with customer needs and plugged it into the G2 knowledge base. Tim highlighted that customizing the LLM, prompts, and content fueled Monty's effectiveness.

These differing elements further allowed Monty to cater to specific objectives across product recommendations, support, and sales. 

Monty could maintain a coherent conversation across different scenarios and also integrate product listings and visuals in the chat. These overlaps gave customers a 360-degree view of G2’s offerings, especially while assisting marketing leaders who need a nuanced understanding of the products they evaluate.

More than your average newsletter.

Every Thursday, we spill hot takes, insider knowledge, and news recaps straight to your inbox. Subscribe here

Tim commented, “Monty became capable of handling customer and vendor requests, which led to us discontinuing live support chat in favor of Monty's capabilities. With an 80% case deflection rate, we've seen great success. We've also created an AI bot creation platform to adjust Monty's strategies and improve conversation handling.”

The team overcame AI-specific challenges like hallucinations by: 

• Setting strict conversation parameters
• Significantly changing prompt strategies
• Making continuous refinements

Data-driven personalization made Monty likable, familiar, and engaging. The team used zero and first-party data to give it contextual cues. Tim also shared that his team constantly works on using G2’s existing technology stack to improve Monty's integration and accelerate product development.

A flexible and iterative approach allows for rapid evolution, adaptation to user feedback, and fine-tuning to align with customer behavior patterns. 

The implementation of Monty highlighted a surprising insight. G2 buyers actively use Monty to guide their journey.

This readiness is a testament to Monty's user-friendly design, practical functionality, and the growing trust in AI's capacity to deliver.

Keys to building your blueprint

Monty, Erica, and ING’s success stories are testaments to the integral role of AI chatbots in customer experience strategy.

Taking a leaf from each use case, trust your learnings:

• AI is a long-term investment that needs continuous improvement.
• Keep it simple and leave the doors of communication open with key stakeholders.
• Harness agile development methodologies to enhance capabilities according to customer needs.
• AI personalization simplifies complex customer journeys as long as you know your data sources and audience.
• Organized data is essential for scaling personalization.

Implementing AI can get too complex too soon. These actionable steps can transform AI from a mere tool into an essential ally for exceptional customer experiences. Happy unicorn hunting!

If you enjoyed these souvenirs from your trip to G2’s AI City with Monty, subscribe to the G2 Tea newsletter for the latest tech and marketing thought leadership.