
Retail Personalization Is Only as Good as the Data Behind It
Retailers have spent years investing in customer data and personalization, yet the experience customers receive is often still surprisingly generic. The challenge is no longer simply collecting more information. It is making sure the business can recognize the customer, understand the context and act on accurate information at the right moment.
Amperity’s 2026 State of Personalization in Retail survey highlights that disconnect. While 83% of consumers said personalized shopping was important, 57% still described their experiences as generic, and 79% said retailers frequently got personalization wrong.
To understand why that gap persists, CTO Magazine spoke with Derek Slager, co-CEO and co-founder of Amperity. We asked Slager where personalization breaks down, why customer identity remains such a difficult problem for large enterprises, and what role AI can realistically play in building more accurate and trusted customer experiences.

The personalization gap is becoming a data problem
Consumers are asking for more relevant experiences, but retailers are still struggling to deliver them consistently. Amperity’s 2026 survey points to a disconnect between customer expectations and the quality of personalization being delivered. The issue suggests that personalization is increasingly dependent on the accuracy and accessibility of customer data, not simply the sophistication of marketing tools.
Derek, your research shows a pretty significant gap between what consumers expect from personalization and what retailers are actually delivering. Why do you think that gap is still so wide?
Slager: Most organizations still limit their use of personalization to a messaging problem. They think of it in terms of how to deliver the right offer, recommendation, or campaign, when in reality, personalization is an accuracy problem that spans the entire enterprise. Our 2026 State of Personalization in Retail survey finds that while 83% of consumers say personalized shopping is important, 57% still describe their experiences as generic, and 79% say retailers frequently get personalization wrong. The real issue is that most companies still struggle to maintain accurate, unified customer data in real time. The result is a growing gap between what businesses know about customers and what they can actually act on in the moment that decision matters.
If a company cannot accurately identify a customer across systems, channels, and moments in real time, then every downstream experience becomes unreliable. You end up with duplicate profiles, outdated preferences, disconnected purchase histories, and recommendations that feel irrelevant.
Personalization cannot live solely inside marketing. It depends on data engineering, analytics, commerce, customer service, privacy, and operations, all working from the same trusted customer foundation. When a customer contacts support, that agent should have the same view of that person that your email platform does. When they don’t, the experience breaks.
Where do you see the biggest breakdowns in data accuracy across the enterprise, and how do these gaps directly impact customer experience and business outcomes?
Slager: Enterprises collect enormous amounts of customer data, but it lives in disconnected systems. Between e-commerce, mobile apps, customer support, advertising platforms, and data warehouses, most enterprises are unknowingly operating dozens of disconnected versions of the same customer. Because of this, we usually find that the biggest breakdown happens at the identity layer, where multiple versions of the customer exist.
When customer identities go unresolved, every business function suffers. Marketing sends mistimed or irrelevant offers, customer service lacks context, analytics teams work from incomplete data, and product teams don’t fully understand customer behavior.
Our survey finds that relevance, accuracy, and timing are the three attributes shoppers value most in personalization. If any one of those breaks down, customers notice and trust erodes quickly.
From a business perspective, inaccurate data creates wasted media spend, lower conversion rates, weaker loyalty, and missed revenue opportunities. If the underlying customer profile is fragmented, the experience still feels generic—no matter how sophisticated your personalization stack.
Many companies invest heavily in data infrastructure, but still struggle with fragmented customer views. What structural or organizational barriers tend to prevent true data unification?
The challenge is rarely caused by a lack of technology investment. Most large enterprises already have the infrastructure, but the true barrier is fragmentation on both the technical and organizational sides.
Different parts of the organization interact with customer data in very different ways. Marketing focuses on campaigns and segmentation, e-commerce on digital behavior, support teams on service history, and data teams on infrastructure and analytics.
Without shared standards and coordination, each team ends up building its own version of the customer. The result is often fragmented customer views and multiple “sources of truth” that contradict each other. Breaking that pattern requires deliberate cross-functional alignment around shared definitions of the customer – not just better tooling.
Derek Slager, co-CEO and co-founder of Amperity
The companies moving fastest are the ones reorganizing around the customer instead of around channels or internal systems.
More customer data does not necessarily mean better customer understanding
Large retailers have invested heavily in data platforms, but information often remains divided across commerce, marketing, support and analytics systems. That fragmentation can make it difficult to establish a consistent customer identity, even when an organization has no shortage of data.
How should enterprises rethink ownership of customer data across teams (marketing, product, data, engineering) to ensure consistency and accountability?
Slager: Customer data should not be owned by a single department. The most successful organizations stop arguing over who owns it and start treating it like shared infrastructure—the same way they think about their data warehouse or their network.
Marketing, engineering, analytics, and customer experience teams all interact with customer data differently, but they need to operate from the same foundation. That requires shared customer definitions, consistent identity standards, and aligned governance practices.
The key is treating customer accuracy as a company-wide responsibility. Once organizations align around a common customer foundation, personalization stops being a marketing problem and starts being a competitive advantage.
From your experience, what distinguishes organizations that successfully operationalize personalization at scale from those that remain stuck in pilot phases?
Slager: The companies that scale personalization successfully tend to approach it very differently from the ones stuck in pilots. They focus less on isolated campaigns and more on building the operational systems behind customer decisioning.
In practice, that usually means they can consistently recognize customers across channels and devices, act on customer signals in real time, and measure personalization based on business outcomes instead of engagement metrics alone. The data does not sit in dashboards waiting for analysis. It directly influences recommendations, suppression logic, service experiences, and next-best actions while the interaction is still happening.
That gap matters because consumers increasingly expect personalization to feel immediate and relevant. Our survey found that 74% of consumers are more likely to purchase when they receive truly personalized recommendations, and 69% are more likely to buy when recommendations adjust instantly while they browse.
That’s where many organizations still struggle. They may have the data, but they cannot operationalize it fast enough for the experience to actually feel relevant to the customer.
How do you approach resolving identity fragmentation across channels while maintaining data integrity and privacy compliance?
Slager: Identity resolution starts with being able to recognize the same customer across different systems and channels in a reliable and privacy-conscious way.
At Amperity, we use AI to connect customer records that traditional systems often miss. The goal is not just to connect records faster. It’s to create customer context reliable enough to support real-time decisions. Equally important is maintaining clear guardrails around consent, permissions, and data usage. Privacy compliance cannot be layered on afterward. It has to be built directly into the customer data architecture.
The balance organizations need to find is recognizing customers in ways that feel helpful and relevant without crossing into experiences that feel invasive. Confidence in how data is used responsibly and transparently remains one of the top factors influencing customer spending and loyalty. That trust becomes a competitive differentiator.
With increasing regulatory scrutiny, how can companies balance deep personalization with responsible data usage and governance?
Slager: The future of personalization comes down to trust. Consumers are willing to share data when the experience feels relevant and helpful, but they also expect companies to handle that data responsibly.
The companies doing this well are transparent about how customer data is used and build privacy directly into their systems and processes from the start. That includes consent management, access controls, and clear governance on who can use data and how.
AI also has a role to play. Companies need to understand how their systems make decisions rather than relying on models they cannot explain. In practice, that means being able to answer basic questions: Why did this customer receive this offer? What data drove that recommendation? If a company cannot explain its decisions internally, customers won’t trust them externally.
What role does AI play in improving data accuracy and decisioning in personalization and where should organizations be cautious about over-reliance?
Slager: AI is becoming essential to personalization because companies are dealing with more customer data than manual processes can realistically handle. As we’ve seen, it’s especially useful for connecting customer records, identifying patterns, predicting intent, and helping companies respond in real time. Implementing it in these ways allows organizations to move beyond broad audience segments and deliver more relevant, individualized customer experiences.
At the same time, AI is only as good as the data behind it. If the underlying customer data is fragmented or inaccurate, AI will simply scale those problems faster. Additionally, consumers still value human judgment, especially in service and high-trust interactions. We found that many consumers prefer a mix of AI-driven personalization and human support rather than AI alone. The best use of AI is to improve accuracy and support better decision-making, not replace human accountability.
There’s also a trust issue here. Customers want recommendations that feel relevant, but there’s a point where personalization can start feeling intrusive. How should retailers think about that line?
Slager: Engagement metrics alone can be misleading because “clicks” don’t necessarily reflect customer trust, loyalty, or long-term business impact.
Leaders should evaluate personalization through a broader operational and financial lens – customer lifetime value, retention, loyalty participation, and repeat purchase rates. Accuracy shows up in whether you can consistently recognize the right customer, deliver relevant experiences across channels, and respond while the customer intent still exists. A practical example: if a customer browsed a product online, bought it in-store, and then received an email promoting the same item, that’s a measurement failure.
Fixing it shows up as reduced unsubscribes, higher repurchase rates, and stronger loyalty enrollment.
If you look three to five years ahead, what changes most about personalization? Are we moving toward a world where every customer interaction is continuously adjusted in real time?
Slager: Personalization is moving toward something much more real-time, contextual, and adaptive. Consumers increasingly expect brands to recognize them immediately, understand what matters in the moment, and respond while that intent still exists. Batch-and-blast personalization is already starting to feel outdated.
At the same time, the industry is shifting away from third-party signals and toward trusted first-party customer context. That makes identity and data accuracy even more important because AI systems can only make good decisions when the underlying customer understanding is reliable. AI will absolutely help companies scale personalization faster, but trust and transparency are becoming just as important as the technology itself.
Ultimately, personalization is evolving beyond campaigns and segmentation. The companies that win will be those that can continuously understand customers, make real-time decisions, and adapt experiences as customer behavior changes.