Why AI, Trusted Data and Human Judgement Will Define Customer Experience for the Year Ahead

As AI becomes increasingly embedded in marketing and customer experience strategies, brands will face growing pressure to move beyond personalised content towards smarter, more responsible customer decisions. 

In 2027, competitive advantage will depend on trusted customer data, stronger AI governance and the ability to balance automation with human judgement. Bel Lloyd, Customer Success Lead at Amperity, shares four predictions for the year ahead, exploring how businesses can bridge the gap between AI adoption and meaningful customer outcomes, while building the skills, trust and data foundations needed to deliver more relevant experiences.

1. Over the next 12 months, more brands will discover that generating personalised content is easier than deciding which customer experience is actually relevant.

AI is making tailored content easier to produce. But generating something for a customer is not the same as understanding what that customer needs, or whether the brand should engage at all. I think the market is overestimating the value of more personalised output and underestimating the customer context needed to make a good decision.

Consider a customer who has just returned a product or has an unresolved service issue. An offer based only on their purchase history might look relevant to a model while feeling tone-deaf to the person receiving it. Connected context can change the decision from sending another promotion to addressing the problem first.

Amperity's 2025 State of AI in Retail provides a useful baseline. It found that 45% of surveyed retailers were using AI daily or several times per week, but only 11% said they were fully prepared to deploy it at scale. The research also found that 58% had fragmented or incomplete customer data.

Over the next year, I expect more brands to shift investment towards resolving customer identity, connecting relevant signals and improving data quality. The competitive advantage will come from better customer decisions, not simply producing more content.

2. Customer-data advantage will increasingly depend on whether brands can turn trusted, permissioned data into better customer decisions.

Richer profiles alone will not create an advantage. Brands also need reliable customer context, clear boundaries around its use and a value exchange that customers understand. First-party data is a relationship asset. Customers share information because they see a reason to trust the brand and receive value in return.

We are already seeing customers in APAC focus projects on consent and examine how to improve customer experiences around consent preferences. As AI expands what organisations can infer and act upon, those preferences need to inform how data is used, rather than sit separately from marketing decisions.

Over the next 12 months, I expect more brands to make customer preferences part of their data and activation workflows, and explain more clearly how sharing information benefits the customer. A preference to receive one type of communication should not be treated as an invitation to every form of engagement.

The opportunity is to earn and sustain permission for relevant experiences. Brands that connect customer understanding with appropriate use will be better placed to build lasting relationships as AI increases the speed and scale of engagement.

3. As AI moves from recommendations into action, leaders will need to define which customer decisions can be delegated and which require human approval.

AI literacy and governance will become a practical leadership capability. Leaders will need to set the boundaries for how AI supports customer decisions, who approves consequential actions and who remains accountable for the outcome.

For example, using AI to analyse an audience is a different decision from approving a campaign that changes how the brand treats those customers. Leaders need to consider the customer impact, the reliability of the underlying information and whether an action can be reviewed or reversed before deciding how much to delegate.

Over the next year, I expect more organisations to define approval points and escalation paths within their customer workflows. Teams should know when they can act on a recommendation, when they need to check it and when a person must make the final call.

The leaders who emerge will be those who are fluent in AI literacy and who understand where humans matter most. Human judgement should help determine what a good customer outcome looks like, as well as whether the technology is working as intended.

4. The next AI skills gap in marketing will be the ability to challenge a recommendation using customer evidence and business context.

The first wave of AI training focused on operating the tools. The next will need to teach people how to evaluate what the tools recommend. An answer that sounds convincing still needs to be checked against the customer information, business objective and assumptions behind it.

If AI recommends targeting a supposedly high-value audience, a marketer should be able to question the reasoning behind it. How was value defined? Are these distinct customers or fragmented records? What information is missing? How will we know whether acting on this recommendation improved the outcome?

Over the next 12 months, I expect more organisations to build those questions into team training and campaign reviews. AI literacy will mean knowing when to trust a recommendation, when to challenge it and when to seek additional evidence.

Leaders set the rules for AI. Teams need the skills to apply those rules and exercise judgement in their everyday work. The future skill is knowing what makes an AI answer worth acting on, and retaining responsibility for the decision.