Häufig gestellte Fragen
What does single source of truth mean for master data?
For master data, single source of truth (SSoT) means that every business-critical data record — for example on customers, suppliers, articles or accounts — is maintained in exactly one leading location and distributed from there without contradiction to all dependent systems. Instead of several parallel and often divergent copies, there is one authoritative source from which downstream applications derive their view. The distinction between a logical and a physical SSoT is important: what is meant is one functionally unambiguous source per attribute, while technically controlled replicas in several systems are entirely permissible. A master data strategy translates this principle into concrete decisions on the data model, responsibilities and interfaces.
Why do single-source-of-truth initiatives often fail in practice?
A common reason is that business departments create their own shadow data stores — such as Excel lists, local databases or notes — because the leading system does not cover their requirements or maintenance is perceived as too cumbersome. Studies and field reports cite human error, lack of standardisation, unclear responsibilities and missing processes as typical causes of poor master data quality. An SSoT is therefore primarily an organisational concept of governance and processes, and not solely a question of tooling. Only clear responsibilities, binding maintenance rules and suitable technical support make the strategy sustainable.
Is the ERP automatically the single source of truth?
In many companies, the ERP system is the obvious leading source for central domains such as articles, business partners or accounts, because order, procurement and posting processes converge there. However, that does not automatically make it the single source of truth, since most organisations run additional systems such as CRM, PIM or a web shop that also need reliable master data. A master data strategy therefore defines domain by domain which system leads for which data — for example terms and cost data in the ERP, customer contacts in the CRM, and enriched product attributes in the PIM. What matters is the clean delineation of data ownership per domain, not the existence of a single system.
How do a master data strategy and Master Data Management (MDM) differ?
The master data strategy is the organisational guideline that defines which data is maintained authoritatively where and how it is distributed, including responsibilities and quality targets. Master Data Management, by contrast, is the operational discipline and the tooling with which this principle is implemented in the system landscape — for example through consolidation, cleansing and synchronisation of data records. MDM software alone, however, does not create a single source of truth as long as governance and maintenance processes are missing. In practice, the two interlock: the strategy defines the goal, MDM provides the machinery for its lasting implementation.
What is a golden record and how does it relate to the master data strategy?
A golden record is the most complete and correct possible representation of an object — for example a customer or an article — that merges information from different sources into one consistent data record. It is, in a sense, the practical result of the single-source-of-truth principle at the level of the individual record. It is typically created through matching, duplicate cleansing and consolidation with survivorship rules as part of Master Data Management. The master data strategy creates the preconditions for this by defining unique keys, maintenance rules and responsibilities that make a stable golden record possible over time in the first place.
How do you measure the quality of master data?
The quality of master data is usually assessed across several dimensions, including completeness, consistency, timeliness, correctness and uniqueness of the records. In practice, metrics can be derived from these, such as the share of filled mandatory fields, the number of detected duplicates or the age of records since their last review. Continuous measurement is important because a single source of truth is not a one-off action but a permanent maintenance and governance process. Field and industry reports also point out that poor master data quality has a particularly strong impact on production planning, sales and material management — which is why quality metrics there should be closely tied to the operational processes.