Everyone Has the Same AI. Why Would Customers Pick You?

AI can write the campaign, crunch the data and answer the customer. It can’t decide why anyone should care. Five marketing leaders on what still makes a brand worth choosing.

Topics

  • AI can help a marketing team produce dozens of campaign variations, analyse customer information and automate interactions across channels. But having more ways to reach customers does not necessarily give a brand more reasons to be chosen.

    The difficulty begins when the ability to produce more is mistaken for the ability to connect better.

    A shopper may move between social media, a mobile phone and a physical store before making a purchase. A mother may look for guidance during a particular stage of her child’s development, while a healthcare customer may need help understanding benefits or navigating a service process. These journeys differ, but the expectation is similar: customers want brands to understand their needs and make the experience easier.

    Kai Xin Lee of Supermom, Andros T. of ASICS, Jason Huan of Endowus, Shalini Seneviratne of Kimberly-Clark, and Dr Jaslyin Qiyu, formerly of Cigna Healthcare International Health, approach this challenge from different perspectives. Their experiences point towards a common argument: AI can make marketing more efficient, but it cannot independently establish what makes a customer relationship worth maintaining.

    The limits of automation and where human judgement matters

    Kai Xin Lee, Chief Experience and Marketing Officer at Supermom, sees AI as a test of the thinking behind a marketing strategy. 

    When a team has not established who it is speaking to, what those people need or what it wants to change for them, the technology can still produce something that looks convincing.

    “AI reflects intent,” Lee says.

    She has seen teams become preoccupied with how quickly they can generate more options. But additional variations rarely make the work more meaningful. Her experience taking Aima, Supermom’s parenting AI companion, from product development to go-to-market illustrates the difference between a compelling demonstration and a product people have a reason to use again.

    “The question I keep coming back to is: if the technology disappeared from the pitch, would the user problem still be important?” Lee says.

    Dr Jaslyin Qiyu, who was Chief Marketing Officer and Head of Customer Experience for Singapore and Australia at Cigna Healthcare International Health, approaches automation from another high-stakes industry. 

    She believes AI can play a useful role in healthcare by removing administrative friction and giving human teams more time to focus on interactions that require care and reassurance.

    “I don’t think the future is automation versus empathy. The future is automation, enabling empathy at scale,” she says.

    The risk comes when organisations begin treating efficiency as a substitute for experience. Qiyu’s advice to marketing teams is to “use AI to accelerate thinking, not replace accountability.”

    Jason Huan, Chief Marketing Officer at Endowus, sees a related change in the value of creative work. 

    AI has reduced the time and expense required to visualise campaign ideas, making it easier for teams to explore concepts before committing resources.

    But removing the production barrier has not removed the need for conviction.

    “AI removed the expense, not the problem of conviction,” Huan says. “If producing the work is no longer the scarce input, the value is entirely in the decision; the taste to pick, kill and hold a line under commercial pressure.”

    A customer journey can break down even when every channel works

    Retailers have invested in connecting digital and physical channels, but a connected technology environment does not automatically translate into a connected customer experience.

    Andros T., Regional Director for Direct-To-Consumer, Retail and Omnichannel in Southeast Asia at ASICS Corporation, sees the problem in the way organisations continue to manage channels separately, even though customers move between them.

    “The customer sees a total shopping experience for a brand as one journey, but some organisations still classify it as a different channel engagement,” he says.

    The next change may come from how customers discover products. Andros expects AI and machine learning to make shopping less linear, with customers increasingly describing what they need and relying on AI-powered interfaces to narrow down available options. But the physical store is unlikely to lose its relevance in every category. Customers buying functional footwear, for example, may still want to touch, try and feel a product before making a decision.

    The problem of disconnected experiences is not limited to retail. Qiyu identifies the handoff between customer acquisition and service as one of the biggest sources of friction in healthcare.

    “The biggest disconnect occurs during the handoff between acquisition and service,” she says.

    She believes customer experience needs to be treated as a business-wide discipline rather than a marketing initiative. The campaign creates an expectation, but the organisation has to deliver the experience that follows.

    Lee sees a similar responsibility in Supermom’s ambition to join the support system for mothers.

    Commercial performance still matters. What changes is how brands judge it.

    “If a campaign gets a response but leaves people feeling that their trust was used carelessly, that is a poor long-term result for the brand,” Lee says. “Performance matters; so does the quality of the relationship that performance creates.”

    When everyone can produce content, being chosen matters more

    AI has lowered the barriers to content production, but it has not created unlimited opportunities for brands to be discovered. Huan looks at the growing distance between the volume of information being published and the sources that customers and discovery platforms actually choose to trust.

    “AI hasn’t made it harder to publish; it’s made churning volume worth less, and being chosen worth more,” he says.

    Search and answer engines are selection systems, not storage systems. As the supply of information increases, brands have to compete for visibility within a narrower set of credible sources. For Huan, the foundations of that credibility include:

    • Proprietary data and original information
    • Primary sourcing and independent corroboration
    • A defensible point of view
    • Genuine institutional expertise

    This also changes how marketers approach Answer Engine Optimisation (AEO). Rather than treating citation frequency as an independently controllable target, he believes brands need to establish authority through accurate information and third-party corroboration. The focus should shift from the page to the claim: Can an AI model consistently state a correct, attributable fact about a brand, wherever that information appears?

    Shalini Seneviratne, Marketing Director at Kimberly-Clark, sees the same problem from the perspective of consumer attention.

    “I don’t think today’s biggest challenge is a lack of attention. Consumers are paying attention all the time; they’re just incredibly selective about what deserves it,” she says.

    “The brands that break through aren’t necessarily the loudest; they’re the ones that understand their audience deeply enough to say something that matters,” Seneviratne says.

    That understanding also changes the role creators play in marketing. Rather than bringing creators into a campaign only when the message is ready for distribution, she believes they should be involved earlier to understand consumer needs and shape innovation and communications.

    “The best partnerships make them strategic collaborators,” she says.

    More customer data does not automatically mean a better understanding

    The growing availability of customer information has made personalisation easier to implement. But knowing more about someone does not mean a brand understands why that person is making a particular decision.

    This becomes particularly complicated in healthcare, where the buyer, influencer and end user may be different people. Qiyu points to corporate healthcare, where the employer buys the plan, HR influences the decision and employees use the service. Rather than asking who interacted with a campaign, she believes organisations need to understand what problem someone is trying to solve and where they are in the decision journey.

    “The organisations that win aren’t necessarily the ones with the most data. They’re the ones that connect behavioural signals, customer feedback and service insights into a richer understanding of customer needs,” she says.

    Lee sees another dimension of the same issue in parenting, where personal information can easily cross the line from useful personalisation to intrusive targeting. A mother who chooses to share that her toddler is starting preschool may welcome guidance on that transition. But using sensitive information about a child’s development or a mother’s well-being in an unrelated commercial context can undermine trust.

    “The line is crossed when we infer more than she chose to tell us, or use one sensitive moment in a different context without a clear reason or permission,” Lee says. “More targeting isn’t automatically more relevant.”

    Seneviratne sees trust evolving alongside changing expectations of brands in parenting categories. Historically, brands often built credibility by positioning themselves as authorities on what was best for a baby. Today’s parents have access to reviews, communities, independent information and other parents’ experiences.

    Huan approaches trust from the perspective of financial services, where credibility must be supported by information customers can verify.

    “Nobody has ever been persuaded into trusting a financial institution by an ad campaign,” he says.

    For newer financial brands, licensing, transparent fee structures, clear disclosures and consistent behaviour during difficult market conditions provide a stronger foundation for trust. The implication is that brands cannot depend entirely on their own messaging.

    Regional scale still needs local understanding

    As brands expand across markets, maintaining consistency becomes more complicated. A regional strategy may create operating efficiencies, but it cannot assume that customers in every market behave in the same way.

    Andros argues that Southeast Asia cannot be treated as one uniform retail market. Singapore, Indonesia, Thailand and Vietnam have different levels of digital maturity, payment behaviours, marketplace ecosystems and relationships with physical stores.

    “Having the scale for each market should provide more capabilities, but not take away their ability to respond locally,” he says.

    Seneviratne sees a similar distinction from a brand perspective.

    “One of the biggest misconceptions is that consistency and localisation sit at opposite ends of a spectrum,” she says. “In reality, the strongest brands are consistent in what they stand for, but flexible in how they show up.”

    “Localisation isn’t simply translation; it’s interpretation,” she adds.

    Marketing needs to measure what actually changes

    As AI makes it easier to produce content, automate interactions and track customer activity, demonstrating marketing’s contribution to the business becomes increasingly important.

    Huan believes discussions about performance often become confused because contribution, attribution and causality are treated as interchangeable.

    “Most of these conversations go wrong because three claims get conflated: contribution, attribution and causality. Stakeholders are usually asking about the third, and are answered on the second,” he says.

    His recommendation is to agree on measurement standards before a campaign begins rather than selecting metrics retrospectively.

    Lee is particularly cautious about treating reach, views and follower counts as outcomes in themselves.

    “Reach and views can help diagnose distribution, but I wouldn’t use them as true measures of success,” she says.

    Her approach is to look at what happened after exposure: whether customers participated meaningfully, returned to the brand and contributed to qualified business outcomes.

    The human advantage in an automated market

    There is a common thread across these conversations. AI is changing how quickly brands can produce, analyse and distribute marketing, but it has not removed the need to understand why people make decisions.

    AI can speed up execution, but someone still needs to decide which customer problem deserves attention. Data can reveal patterns, but understanding what those patterns mean requires context. Automation can eliminate repetitive tasks, but it cannot independently determine whether a customer feels supported.

    When AI becomes available to everyone, the difference will not be who produces the most or automates the fastest. It will be whoever understands what customers value and gives them a reason to come back.

    ALSO READ: Retail Media and the Post-Click Blind Spot for CPG Brands

    Topics

    More Like This