Why companies don't have a true view of the customer, even though all the data is available
Share
Author: Martin BrudekReading time: approx. 5 minutesPublished: August 27, 2026Last updated: August 27, 2026
Topics: CRM Strategy, Customer Data, Data Integration, Loyalty, Golden RecordRelevant for: Heads of CRM, Heads of Loyalty, Customer Experience Leaders, Data & IT Leaders, Marketing Directors
Summary
In many larger organizations, customer data is already available. Yet a consistent view of the customer is often still missing. The issue is usually not a lack of technology, but the absence of a shared data logic.
- Systems use different definitions of the customer.CRM, ERP, e-commerce and loyalty systems each look at the same customer from a different perspective.
- Technical integration alone does not solve the problem.Without clear rules for identities, leading data sources and relationships, data can be combined but not interpreted consistently.
- Fundamental decisions are often made too late.Data logic and responsibilities are only clarified during system implementation, even though they should be defined beforehand.
- This limits the impact of CRM and loyalty.Personalization, segmentation and loyalty mechanics then rely on a customer view that has not been clearly defined.
- A reliable Golden Record requires a shared business logic first.Companies need to define which identity is leading, which systems provide the authoritative data and according to which rules records are merged.
This article explains why a unified customer view cannot be created by adding another tool and which foundations need to be in place before CRM, loyalty or customer data initiatives can deliver their full potential.
Why companies still lack a true customer view even when all the data is there
The starting point: plenty of data, little clarity
In larger organizations, CRM and loyalty initiatives usually already have access to a wide range of systems and data. Most companies therefore no longer have a classic data availability problem. There is a CRM system, transaction data in the ERP, behavioral data from e-commerce analytics, campaign tools and often dedicated loyalty management systems. From a technical perspective, much of what is needed to understand customers and manage interactions with them is already there. This quickly creates the impression that building a consistent customer view is mainly a matter of choosing the right tool.
In practice, however, a different picture often emerges. Relatively early in a project, teams start to question which data can actually be relied on. Different teams work with different figures even though they appear to be looking at the same customers. Campaigns do not perform as expected or require extensive manual correction. In my experience, this situation is initially interpreted almost entirely as a technical problem. Teams start looking for missing interfaces or systems that can “bring everything together.” That is understandable, but it does not go far enough.
Once you look more closely, the perspective changes. The data is there, often in sufficient quality, but it does not follow a shared logic. It has evolved over time in different contexts and serves specific purposes in each system. That is where the real challenge lies. The systems work well in isolation, but they were not designed to create a consistent customer view together. This is not an integration problem in the narrower sense. It is a structural issue.
Fragmented system landscapes are the norm
Most CRM and data landscapes are not the result of one central architecture decision. They evolve gradually in response to specific requirements. A campaign tool is introduced because marketing needs to work faster. A loyalty program is launched to strengthen customer retention. The e-commerce platform is expanded, the ERP is adapted and new channels are added. Each of these decisions makes sense on its own and is often necessary to achieve short-term goals. What is missing is an overarching logic that connects these systems.
In day-to-day operations, this setup often works surprisingly well for a long time. Teams know their data, understand how to interpret it and find pragmatic ways to get results. Knowledge is often held by individual people and is not documented systematically. This implicit stability works as long as complexity remains manageable. Problems arise when several systems need to be used at the same time or when decisions need to be made across systems, potentially by people from different teams.
At that point, it can appear as if the data is “not integrable.” Technically, however, it often is. What is missing is the business logic that determines how the data should be combined. This is where the issue shifts from technology to the organization. The question is no longer how to connect the data, but what the combined data is actually supposed to represent.
Missing data logic as the cause of technological limitations
What sounds obvious at first is often not clearly defined in practice: what exactly is a customer within the organization? This question is fundamental to any form of data integration. Different systems and business functions use their own definitions. These definitions make sense within their respective contexts, but they do not automatically align. In a CRM system, a customer may be a person or contact. In the ERP, it may be a billing entity or contract. In the e-commerce system, it is an account. In the loyalty system, it may be a separate identity based on its own specific logic.
One person may have several loyalty cards, use multiple accounts or be represented differently across systems. Without a clear definition of which of these identities is authoritative and how they should be linked, the customer view will remain inconsistent.
This ambiguity carries directly into the technology. CRM platforms or Customer Data Platforms can bring data together, but they cannot decide which underlying logic should apply. If it is not clear whether a person, a contract or an email address is the leading entity, even a technically clean integration cannot create a consistent view of an identity. The result is simply an aggregation of data rather than a reliable basis for decision-making. In my experience, this is one of the main reasons why integration projects can be technically successful while still falling short of expectations in day-to-day use.
External research supports this observation. McKinsey has pointed out for years that effective personalization depends not simply on the availability of data, but on robust data and analytics foundations and the ability to make that data usable across the organization. In a more recent analysis, McKinsey also argues that successful personalization increasingly depends on rethinking workflows, roles and system relationships rather than simply introducing additional tools (https://www.mckinsey.com/featured-insights/mckinsey-explainers/what-is-personalization, https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights/unlocking-the-next-frontier-of-personalized-marketing). Gartner makes a similar point in its outlook on marketing priorities for 2026, highlighting the need for “AI-ready data, content & context governance.” The underlying message is the same: technological capability is not enough if the underlying data logic and semantics have not been clarified (https://www.gartner.com/en/articles/future-of-marketing).
The problem is often addressed too late
Teams often start discussing systems, vendors and implementation before they have clarified how the data actually fits together. Once budget has been approved and the project gets underway, the focus quickly shifts to features, vendors and timelines. The underlying data logic is then defined in parallel with implementation or postponed to a later stage.
That is where problems begin. Decisions that should have been made before selecting a system are only addressed during the project. Teams then try to map data from source and target systems even though the logic that should govern that mapping has not yet been clearly defined.
This is not only a technical issue. It also affects responsibilities, processes and different interests across the participating teams. If these questions are not resolved in advance, a solution can work technically while still failing to provide a reliable foundation for CRM, loyalty and other applications.
Conclusion: what it takes to build a reliable customer view
A missing customer view is often not a data problem. The data is available and, in many cases, usable. What is missing is a shared logic for how customers, identities and relationships should be represented across systems. As long as that foundation is missing, another system will not solve the problem.
This directly affects CRM and loyalty initiatives. Personalization, segmentation and loyalty mechanics then rely on data whose relationships have not been clearly defined. As a result, the impact remains below what would technically be possible.
The goal should therefore be a reliable Golden Record: a clear, system-wide view of the customer. To achieve this, the business first needs to define which identity is authoritative, which data from which system is considered the leading source and according to which rules different records should be merged. Only then can this logic be implemented cleanly in technology and used consistently over time.
About Martin
Martin is Director Solutions Delivery at DEFACTO and is responsible for the implementation of complex Customer Engagement solutions. His focus is on translating business requirements into robust solutions and structuring projects so that they work in day-to-day operations.
He combines business requirements with system architecture, processes and implementation. For him, success is not only about whether a solution works technically, but also whether the teams involved can use it effectively, develop it further and achieve the intended purpose.
Your contact
Martin Brudek
Director Marketing Technologies & Marketing Technology Management | Marketing Automation
+49 9131 9712 2173
