• Written by Azhar Aliyar, Director

A purchasing manager at a large organisation must decide this week which supplier to use for a new contract, and they currently base it on the data in the purchase order table, which is their single source of truth for supplier information. A single source of truth is a repository that stores the most accurate and up-to-date information about a particular domain, in this case, supplier data. This concept is crucial in large organisations where multiple systems and stakeholders are involved, and it helps to ensure data consistency and integrity.

Definition and Clarification

The single source of truth is often confused with a data warehouse or a data lake, but it is a distinct concept. While a data warehouse is a repository that stores data from various sources in a single location, a single source of truth is a specific repository that stores the most accurate and authoritative information about a particular domain. In the context of AsscherAi, the single source of truth is the organisation's own data, which is used to answer questions in real-time.

The distinction between a single source of truth and a data warehouse is important because it highlights the need for a clear understanding of what constitutes the most accurate and up-to-date information. In a large organisation, there may be multiple systems and stakeholders involved, and it is essential to identify the single source of truth to ensure data consistency and integrity.

Differences from Related Concepts

Another concept that is often confused with the single source of truth is the system of record. While the system of record is the system that stores the official data, the single source of truth is the repository that stores the most accurate and up-to-date information. In some cases, the system of record may not be the single source of truth, as the data may be outdated or incomplete. For example, a shift supervisor checks the maintenance log, which is the system of record, but the single source of truth for maintenance information may be a separate repository that stores real-time data from sensors and other sources.

Applications in Large Organisations

In a large organisation, the single source of truth can apply to various domains, such as customer data, supplier data, or product information. For instance, the category manager compares two suppliers based on the data in the purchase order table, which is the single source of truth for supplier information. This ensures that the category manager is making decisions based on the most accurate and up-to-date information, which can help to reduce errors and improve decision-making.

The single source of truth can also be applied to other areas, such as inventory management or financial reporting. In these cases, the single source of truth would be the repository that stores the most accurate and up-to-date information about inventory levels or financial transactions. By identifying the single source of truth, organisations can ensure data consistency and integrity, which can help to improve decision-making and reduce errors.

Limitations and Misconceptions

Most people get wrong the idea that a single source of truth can be easily implemented by integrating all data sources into a single repository. However, this approach can be complex and may not always be feasible. In reality, identifying the single source of truth requires a thorough understanding of the organisation's data landscape and the various systems and stakeholders involved. It is also important to note that a single source of truth is not a static concept, as it can change over time as new systems and data sources are introduced.

For more information on how AsscherAi can help your organisation identify its single source of truth, please visit our website here. If you have any questions or would like to discuss your specific use case, please contact us.

When it is the Wrong Choice

In some cases, a single source of truth may not be the best approach, such as when dealing with unstructured data or real-time data streams. In these cases, other approaches, such as data virtualisation or event-driven architecture, may be more suitable. Additionally, a single source of truth may not be feasible in organisations with complex and distributed data landscapes, where multiple systems and stakeholders are involved. In these cases, a more nuanced approach may be required, such as implementing a data governance framework or using data integration tools to ensure data consistency and integrity.