Customer Experience (CX): Why It Has Become the Heart of Marketing in the AI Era
Igor M.
6/25/20266 min read
UX and CX: similar, but not the same
According to the Nielsen Norman Group—founded by Don Norman, who coined the term "UX" in the 1990s—User Experience describes every dimension of a person's interaction with a product, service, or company at a specific touchpoint, typically digital (such as an app, a website, or a self-service kiosk). In contrast, the concept of Customer Experience, established by consultancies like Forrester, is broader: it encompasses how a customer perceives all their interactions with a brand throughout the entire journey—both digital and non-digital—ranging from the initial advertisement they saw to telephone support, product delivery, and after-sales service.
In other words: UX is an important piece of the larger puzzle that is CX. A beautiful, easy-to-use app (good UX) won't save a brand if customer service is terrible or if deliveries are delayed without notice (poor CX).
And why does this matter so much to those in marketing and advertising? Because, today, the experience is the advertisement. Before any piece of advertising, it is the lived—and shared—experience that determines whether or not a consumer will trust the brand.
The figures that prove this
PwC’s 2025 Customer Experience Survey, which polled over 5,500 consumers and 400 executives in the U.S., reveals an alarming figure: 52% of consumers say they have stopped buying from a brand following a bad product experience, and 29% abandoned a brand specifically due to poor customer service, whether online or in person. The same study highlights a dangerous disconnect between decision-makers and consumers: while approximately 9 out of 10 executives believe customer loyalty has increased in recent years, only 4 out of 10 consumers share that view.
Other figures underscore the scale of the problem and the opportunity:
The consultancy Forrester estimates that companies willing to align customer experience and brand experience can multiply revenue growth by up to 3.5 times.
According to the Qualtrics XM Institute, poor experiences cost the global economy approximately $3.7 trillion annually.
In its CX Trends 2026 report, Zendesk points out that 74% of consumers already expect 24/7 customer service, and 88% expect faster responses than they did just a year ago.
Regarding Artificial Intelligence, Salesforce projects that by 2027, 50% of customer service cases will be resolved by AI—up from 30% in 2025—and that an automated interaction costs an average of $0.50, compared to approximately $6.00 for an interaction handled by a human agent.
AI, therefore, is no longer a differentiator; it is basic CX infrastructure. But—and this "but" is important—PwC’s own research shows that 58% of consumers still feel uncomfortable using AI tools to interact with brands, and 86% consider human interaction essential to the experience. The lesson for those planning campaigns and customer journeys: automate, yes, but without dehumanizing.
Real-life cases: the problem, the solution, and the result
Theory and statistics help us understand the scale of the challenge, but it is through case studies that transformation becomes tangible. Here are four real-world examples—featuring the problem, the solution implemented, and the measured result.
1. Magazine Luiza — fragmentation between physical and digital stores
The problem: the Brazilian retailer was facing a classic omnichannel challenge—how to maintain information consistency and rapid response times across thousands of products and promotions, integrating physical and digital service channels without creating friction for the customer.
The solution: the creation of "Lu," a virtual assistant that combines machine learning, natural language processing, and data analysis to understand the context of each question. Over the years, Lu has become virtually the face of the brand online—and today, she is a benchmark for the use of generative AI in e-commerce in Brazil.
The result: According to Hubina, the initiative delivered consistent gains in service efficiency and helped cement Lu as one of the biggest digital influencer marketing phenomena in the national retail sector, demonstrating how an operational CX issue can be transformed into a brand and advertising asset.
2. Itaú Unibanco — scaling personalization without losing the human touch
The problem: a high volume of daily interactions and complex financial products made it unfeasible to scale traditional customer service without compromising the quality of the experience.
The solution: the bank built an AI ecosystem featuring the virtual assistant "Bia" across digital channels, voice recognition in the call center, and predictive algorithms to anticipate customer needs, with an investment exceeding R$ 300 million.
The result: according to data released by the bank itself, the technology reduced credit analysis time by approximately 30% and increased fraud detection accuracy by 25%—two indicators that directly impact customer trust in the brand.
3. Banco BMG — friction identified before becoming a crisis
The problem: customers dissatisfied with charges or products often only "explode" when contacting customer service, leading to lawsuits and damage to the brand's reputation.
The solution: the institution began using generative AI in its IVR (Interactive Voice Response) system to identify—while the call was still in progress—signs that customers were likely to litigate, routing these cases to agents trained to handle high-friction situations before the problem escalated.
The result: a 15-point increase in NPS (Net Promoter Score)—a metric measuring the likelihood of a customer recommending the brand—proving that crisis prevention is also a CX strategy.
4. Sephora — Virtual Artist and the problem of low adoption
The problem: Sephora had launched Virtual Artist, an augmented reality feature that allows users to digitally "try on" makeup using their phone's camera. The tool was innovative, yet most app users didn't even know it existed—a classic UX issue of a feature being discovered but not communicated.
The solution: instead of generic campaigns, the brand precisely segmented users who had viewed makeup products in the last 30 days without using the tool and sent them notifications featuring a video tutorial that demonstrated the feature in action—while maintaining a 20% control group to measure the campaign's actual impact.
The result: the tool was used by over 8.5 million visitors, with around 200 million product shades tested virtually. This case study highlights a point often overlooked in marketing: investing in CX technology is pointless if the communication—including advertising within the product itself—fails to lead the customer to discover it.
What these case studies have in common
Looking at the four examples side by side, a pattern emerges:
The problem was never a "lack of technology." It was friction: fragmented information, slow service, hidden tools, and friction that wasn't identified in time.
AI acted as a bridge, not a human substitute. In all the cases, automation handled volume and repetition, freeing up people for moments that truly require empathy and judgment—exactly the balance PwC recommends in its report, advocating that AI should be "invisible," foster trust, and preserve the human element.
The result has become a business metric, not merely a measure of satisfaction. NPS, reduced analysis time, increased fraud detection accuracy, and usage volume—all these indicators are directly linked to revenue and brand reputation, which is, ultimately, what every advertising campaign seeks to protect and build.
What this means for those working in marketing and advertising?
If experience is the new advertising, a few practical points apply to any professional in the field:
Map out the journey before creating the campaign. A brilliant ad cannot compensate for a buying journey riddled with friction—PwC’s data showing that 52% of consumers abandon brands after a bad experience serves as a direct warning of this risk.
Use AI for personalization, but test your communication—just as Sephora did. Having the technology isn't enough; you need a rollout strategy (both internal and external) so that the audience discovers and uses the feature.
Measure what matters to the business, not just "engagement." NPS, retention, average ticket value, and cost per interaction are indicators that are harder to artificially inflate than likes and impressions.
Don't leave the human element behind. Data shows real discomfort with AI among a segment of the consumer base. A well-designed journey offers the option of automation alongside an easy exit route to a human agent whenever the customer needs it.
Sources consulted
PwC — 2025 Customer Experience Survey (pwc.com)
Forrester — Total Experience Score Research, citado via ClearlyRated e UsabilityGeek
Nielsen Norman Group — User Experience vs. Customer Experience (nngroup.com)
IBM — What is User Experience (UX)? (ibm.com)
Zendesk — CX Trends Report 2026 (zendesk.com)
Salesforce — dados sobre custo e adoção de IA em atendimento
Qualtrics XM Institute — impacto financeiro global de más experiências
Hubina — Estudos de Caso: Empresas Brasileiras que Adotaram a IA (hubina.ai)
Blue6ix — Casos de Sucesso: Como Empresas Usam IA para Transformar o Atendimento ao Cliente (blue6ix.com.br)
Braze — Sephora SEA Case Study (braze.com)
Retail Dive — cobertura sobre o lançamento do Sephora Virtual Artist (retaildive.com)