Golden Record Management
The term ‘Golden Record’ refers to the most accurate, complete, and comprehensive representation of any master data domain and indicates that any data labelled as such can be safely used for any application across the enterprise. ‘The best version of the truth’ and ‘The single source of truth’ are two other common terms for this data standard.
When we refer to this golden record in relation to a master data management system, we are making the bold and explicit statement that the data labelled as such is safe, usable, and a guaranteed asset for all business operations and initiatives.
What is the meaning of ‘Golden Record’?
In the broad spread of data management as a discipline, ‘The Golden Record’ is a foundational principle. As in other areas like product information management, when it comes to the golden source for Master Data Management (MDM), the term identifies and describes the single version of the truth – that is, where ‘truth’ is understood as the single and unique data item which is trusted as both accurate and correct. When we compile master database tables from various data sources, we frequently find common recurring problems like duplicate data records, incomplete (and/or missing) values within a data record, and data sets with substandard quality levels. The purpose of the golden record is to resolve these problems by correcting or deduplicating, inserting values when a value is missing from a data field, and generally bringing data quality up to a previously determined standard. Thus, the golden record is what your organisation assumes to be the highest quality possible at a given moment. That gives users throughout the organisation the security that the data can safely be deployed for business purposes.
Why is the golden record important?
The most disruptive problems for the performance of product-centric organisations are duplicated records. If many of these exist in a master data management environment, it seriously affects its usefulness.
We can see a simple example of how data duplication can complicate even the simplest of information for a vendor. We could use an individual consumer – bad enough – but let’s assume it’s a B2B relationship, where the volume of purchases is large, creating a bigger disruptive impact.
An industrial distributor may record the customer as ‘S Jones Tooling’ when sending out a receipt, while the same distributor’s in-store CRM record shows a customer labelled ‘S.Jones Tooling’ in the name field. Separately ‘Samuel Jones Tooling’ sets up an account with the distributor’s ecommerce platform. So, we have the same customer with three different versions
When merging these two databases, ‘S Jones Tooling’ and ‘Samuel Jones Tooling’ will automatically become two discrete data records. The consequent problems are clear when accessing what appear to be two separate accounts. The distributor knows that this is one company, and its systems should reflect this – otherwise, CRM, marketing, sales records, and order tracking (amongst other departments) will see confusion and disruption at best, customer abandonment at worst.
Download our free PIM Buyers Guide
Golden Source/Record in Master Data Management
Master Data Management systems usually manage data across multiple domains. Information on employees, customers, financials, locations, and vendors are typical, generic MDM domains. Consequently, when planning and executing an MDM project, the scope is usually very broad and complex, requiring large amounts of resources to implement and maintain.
MDM systems exist to enhance a company’s business operations and bottom line, so use cases often cover several business objectives and functions departments across the organisation. These business objectives include:
So, the golden record is not simply a desirable state for a company’s data – it is an absolute necessity.
How to go about creating a golden record – matching and merging
Perhaps the biggest factor in implementing a master data management solution is designing workflow protocols to establish the golden record. This is where having a solid and consensualised framework of data governance, stewardship and ownership is critical.
For all data sources, what needs to be established is clarity around the fields where the data sources appear to be more reliable. That develops the criteria for deciding which system establishes primacy in determining whether to populate other MDM domains. Using our example from earlier, if you have a CRM system which captures the customer’s name, and a delivery system which also lists that name, the decision is whether one or the other of these systems tend to list customer information more correctly (or at least, with fewer discrepancies).
When creating and maintaining the golden record, the operations of matching and merging records are common, so if you have two very similar records, what is the procedure for determining which of the two is the correct one?
Of course, life cannot be so simple where the golden record is concerned. There are multiple cases where no single source field is the clear ‘winner’ in terms of correctness and reliability. Manual cleansing may be required to determine which record takes precedence. That is where workflow management tools are very useful. Using the pre-established data governance framework, the data steward assigned as responsible is able to judge which field from which record should be used as the right version. The consequent modification to the existing golden record has implications, so there should also be approval mechanisms in place – when a data record is definitively merged, it needs to be signed off by that steward. Decisions made regarding the golden record need both clear lineage and traceability.
How can consultancy provide support to achieve your golden record?
Data analytics technology is constantly evolving, and cross-functional solutions are becoming more common. For the purposes of the entire organisation, the golden record is the foundation upon which data quality is built. It informs many parts of your business, especially master data management and metadata management.
At Start with Data, we carry out a targeted three-phase Product Data Quality Assessment to guide your organisation on a prioritised roadmap for process, data governance and technology improvements for master data management.
We use the best data cleansing tools, designed to enhance the accuracy, completeness, relevance, and consistency of your product data so that when it comes to centralising those data sets, you can be sure that you have the best possible golden record.
We pride ourselves on our flexibility and adaptability to clients’ requirements and it may be the case that your organisation will need to access certain parts of our Data Quality consultancy services. Our mission is to serve your needs with our expertise, so that the outcome aligns totally with your brief. Get in touch for a conversation about how we can help you to achieve your golden record.