Retailers are all-in on AI. In Amperity’s new 2025 State of AI in Retail report, 97% of retail employees expect their organizations to maintain or increase AI spending over the next 12 months. There’s optimism about AI’s potential too: 63% believe it will improve customer loyalty, and 65% expect it to increase customer lifetime value (CLV).
Despite that enthusiasm, there’s a gap between intent and execution. Less than half (43%) of retailers surveyed are currently using AI in customer-facing applications. Significantly less (23%) are using AI to resolve customer identities or prepare data for marketing.
It’s clear that AI adoption isn’t happening where it will have the greatest business impact: directly in front of customers. In this post, we’ll explore what barriers and hesitations retailers are facing and how they can build the infrastructure to best leverage customer-facing AI.
The high stakes of customer-facing AI
Retailers know that customer trust is hard to win and easy to lose. One survey respondent summed it up plainly: “Customers have said if they see AI descriptions they will not buy the product. Stopped AI descriptions immediately.”
That caution is legitimate. Retailers know that a misstep in customer-facing AI - whether it’s an awkwardly written product description or an irrelevant personalization attempt - can damage brand trust and loyalty. Getting AI use right with customers feels (and often is) more critical to business success than internal implementations.
In the long run, an arm’s-length approach to AI won’t hold. The most meaningful gains in loyalty, revenue, and CLV will come from using AI to power customer-facing experiences: hyper-personalized marketing, dynamic merchandising, predictive service. A recent Deloitte study found that, “7 in 10 retail executives expect to have AI capabilities in place within the year to help personalize experiences.” Yet our survey results showed that actions aren’t keeping pace with those expectations.
For retailers to maximize value from AI, they need to shore up the foundation AI relies on - their customer data.
Better inputs create better outputs
Many AI applications fall short in front of customers not because of the AI itself, but because of what the AI has (or doesn’t have) to work with. Without clean, unified, identity-resolved customer data, even the most sophisticated AI tools will fail to deliver accurate, relevant, and trustworthy results.
As explored in our recent post on building personalization infrastructure at scale, AI is only as strong as the data foundation it's built on. Yet in our new report, 58% of retailers admitted their customer data is fragmented or incomplete, and only 21% feel “very confident” in their ability to understand and act on customer behavior across channels.
This is why so few are applying AI to customer-facing use cases. Without reliable, orchestrated customer data, retailers can’t have confidence that outputs would actually move business forward. The risks outweigh the rewards, particularly when inaccurate or irrelevant experiences can be so damaging to customer loyalty.
Data-first AI strategies unlock real value
So what should retailers do? The solution isn’t to abandon AI in customer-facing scenarios and focus on internal use - it’s to build a customer data infrastructure that makes AI trustworthy and effective end-to-end.
AI-powered identity resolution connects customer behaviors across devices, channels, and systems into unified profiles that power better targeting, personalization, and measurement. Adopting a reliable tool for identity resolution isn’t just about fixing fragmented data, but proactively expanding the breadth and value of your data.
When retailers can tie each customer to a single, persistent identity, the AI-powered experiences that follow are more relevant, accurate, and trustworthy. Actionable, up-to-date customer profiles can then feed into every customer experience touchpoint.
From there, AI can accelerate what marketers already do:
Narrow targeting to high-value audiences
Generate personalized content at scale
Predict churn and proactively intervene
Measure CLV in real time
According to our survey, these are the exact benefits that retailers expect to gain from AI in the future.The only prerequisite is building a strong data foundation.
AI success depends on data readiness
The retail industry is clear-eyed about AI’s potential - and the internal, operational applications are already delivering results. But until retailers address the underlying data challenges that limit customer-facing AI, they’ll continue leaving value on the table.
The takeaway? Retailers should stop thinking about AI in isolation and think about it together with data infrastructure strategy. The fastest, safest, and most effective way to deliver AI-powered customer experiences is to ensure your AI has clean, unified, identity-resolved customer data to work with.
If you’re wondering where your organization stands - and how your peers are navigating these challenges - download our new research, the 2025 State of AI in Retail report.