Over the past couple of years, artificial intelligence has dominated the business conversation. Senior executives are asking how they can use AI to reduce costs, automate tasks, and make processes more efficient. But that’s only a partial starting point. The broader benefits revolve around how AI can reinvent and improve the customer experience, for those companies willing to remake entire processes rather than tinker with small steps.

AI is transforming how customers research, purchase, and interact with companies. For example, a Bain & Company survey at the end of 2024 found that about 80% of consumers rely on “zero-click” results in at least 40% of their searches, reducing organic web traffic by an estimated 15% to 25%. During the first half of 2025, the number of ChatGPT prompts grew by nearly 70%, according to a sample by market intelligence firm Sensor Tower. Chatbot usage continues to grow rapidly.

80% of consumers rely on “zero-click” results in at least 40% of their searches, reducing organic web traffic by an estimated 15% to 25%

Generative AI has helped make experiences simpler, faster, more personalized, and more under the customer’s control. Now agentic AI, which can reason and act autonomously, is spurring another step change in the customer experience. In turn, a superior experience can produce a triple play: happier customers who remain loyal, more engaged employees spending less time on rework and resolving problems, and higher shareholder performance due to lower cost and faster revenue growth.

From front stage to backstage

To understand AI’s potential, it’s useful to think of the customer experience as operating in two arenas: the front stage—what the customer sees and feels—and the backstage systems and workflows that power the front.

AI helps companies remove friction from every part of the front stage. Customers no longer need to navigate complex websites, memorize product names, or wait on hold. Instead, they can describe what they want in their own words and receive an immediate, tailored response.

Company-customer interfaces are shifting from static menus to dynamic conversations. Instead of pushing customers down a linear flow, companies are giving them more control—responding to needs as they arise, in context, with relevance.

Delivering that seamless experience on the front stage requires a backstage with connected functions, such as marketing and distribution, that previously may have worked in silos. Data flows, business rules, fulfillment operations, and other processes must synchronize more tightly than before.

Leading companies are radically rethinking the backstage in parallel with the front, often using agentic AI. They’re integrating AI across core processes, using the technology to autonomously route service requests, generate content, summarize customer history, and even detect issues before they become complaints. This dual focus—reinventing both the customer-facing layer and the underlying machine—is essential to creating repeatable, scalable outcomes (see Figure 1 for an insurance example).

Figure 1

Innovative examples from the field

Let’s look at how companies in multiple industries have started to redesign their customer experience through AI.

Banking. Bradesco, one of Latin America’s largest banks, pioneered the use of AI in the financial sector nearly a decade ago. More recently, Bradesco built a generative AI chatbot that resolves customer problems without human intervention in 90% of cases, serving millions of customers every day. And Bradesco now uses agentic AI to provide money transfers in seconds. Its Smart PIX conversational AI assistant allows customers to move money via voice command, text message, or photo directly on WhatsApp using Pix, Brazil’s instant payment system created by the Central Bank. For the customer, the result is a faster, more intuitive payment experience.

In the US, Capital One, a major provider of auto loans, has launched Chat Concierge, an AI agent that aims to lift the “cognitive burden” of purchasing a vehicle. Chat Concierge, which customers find on participating dealers’ websites, orchestrates tasks that range from estimating the value of a trade-in to scheduling appointments with sales staff.

A short explainer on generative AI in retail banking offers additional context on these types of use cases.

Insurance. Allstate has discovered another benefit of AI models, for insurance claims: The models produce more empathetic emails than many of Allstate’s human representatives. The company’s 23,000 reps send out about 50,000 daily communications to claimants, either seeking information or negotiating settlements. Now, AI generates almost all of these emails, which the company has found to be less accusatory, less jargony, and more empathetic than communications written by human reps, who still oversee the emails for accuracy.

Telecommunications. Verizon uses generative AI to accurately predict the reason behind 80% of incoming service center calls and then connect callers with a suitable human agent. This has reduced store visits and overall churn, staving off defection of an estimated 100,000 customers in 2024.  

Airlines. Delta’s touchless ID uses biometric screening to speed up travelers’ journeys through the airport. The technology uses facial recognition software to confirm a traveler’s identity, check them in, print out bag tags, and allow them to walk through security without showing a boarding pass or pulling out identification.

Retail. For many years it was difficult for consumers to see exactly how different shades of a beauty product would look on their face. L’Oréal first offered virtual makeup try-ons in 2018, through a technology that used facial images from women of diverse ages and ethnicities for a superior color match. Recently L’Oréal expanded its AI tools to launch Beauty Genius, which provides personalized diagnostics, education, product recommendations, and try-ons across hair care, hair color, makeup, and skin care.

When it comes to agentic AI, many retailers are still in the early stages, typically in discrete domains such as inventory, pricing, or store layout. Walmart, however, recently unveiled its flagship AI agent, Sparky, designed to make decisions and act without considerable human oversight. Sparky, accessed through the Walmart mobile app, helps customers search for products tied to specific events, compare different options, and sift through reviews. Over time, Walmart plans to upgrade the tool with capabilities such as automatic reordering.

These companies didn’t set out simply to add more technology to the customer journey. Instead, they focused on removing friction, adding relevance, and increasing value from the customer’s perspective.

A new experience playbook

As the pace of AI deployment accelerates, the biggest risk isn’t moving too fast but rather moving too narrowly. Companies need a comprehensive strategy for an AI-enabled customer experience. Once that strategy is in place, here’s where to focus.

1. Start with the customer’s priorities. Understand what matters to customers and be open to reinventing how to meet their needs. Identify one or two high-gain journeys, where friction is high and experience matters. Ask: How could this journey look if we redesigned it from scratch with AI at the core? How do we want customers to feel? In banking, for instance, it’s not about “mortgage origination” but about “buying my first home with confidence.” In retail, it’s not “order fulfillment” but “getting my product when and how I need it.” AI helps companies design journeys the way customers experience them.

2. Take a clean-sheet approach to redesign the backstage. Backstage processes underpin the experience, and technology and data platforms underpin the processes. So it’s critical to identify where AI agents can reason and make decisions and how to get them working together behind the scenes.

One important step is to test where AI agents can take on responsibility and how they coordinate with each other and with humans. Agents can already prefill insurance claims or loan applications using data pulled from multiple sources. They can trigger payments or refunds directly in core systems. Looking ahead, why couldn’t an insurer start by immediately covering customers for basic risks, then add additional coverage based on what the customer’s personal AI agent is willing to provide to the insurer’s AI agent?

3. Reinvent roles and operating models. AI isn’t just a tool—it’s a team member collaborating with humans. The work shifts from a large operational team to a more automated process that requires fewer humans, who will mainly supervise and check the automated outputs. In call centers and other groups, humans will increasingly handle only complex issues, with AI agents taking care of the simple requests.

Map the dependencies, because changing how the work gets done depends on orchestration to tie all the data, technology, and business processes together. Teams should organize across functions, not in silos, in the service of fulfilling individual customer journeys. In tandem, employees will need to be reskilled and redeployed. New roles such as AI supervisors, AI trainers, and journey owners will become core to the organization.

Managing this substantial change involves giving people the right skills, confidence, and clear pathways to more valuable work. An effective, redesigned customer experience hinges on engaged employees, not just slick AI tools.

4. Embrace an iterative form of delivery. AI evolves too quickly for long-cycle, waterfall-style projects. Companies that cling to old delivery models will fall behind. While agile delivery is well established, the rise of AI exposes any cracks in the execution. Some processes require a common technology platform to scale up, while other technology elements should be tailored to the specific user experience. Getting that balance right will entail a number of iterations and careful orchestration.

Executives should insist on test-and-learn programs with tight feedback loops, transparent governance, and the ability to shift direction as technology matures. What this means in practice is shorter planning cycles; cross-functional squads blending business, data, and technology; and governance that focuses on outcomes, not activity.

Rewired for the customer, not just for efficiency

In the race to capture AI’s potential, too many companies still only look inward, using AI to boost productivity, automate processes, and trim costs. That’s valuable but limited. The bigger rewards will flow from addressing customers’ most pressing priorities. Ambitious companies are demonstrating how AI can make the experience not just more efficient but also more responsive and empowering for customers.