A single version of truth can have many faces
Almost every time an organization embarks on a new enterprise data management initiative, there will at some point be mentioned the idea of a “Golden Record.”
The Golden Record, of course, is this utopian goal where a single version of the truth exists that is leveraged by every part of the organization, top to bottom, for understanding and interacting with its clients, prospects, suppliers, partners, and so on.
When to Consider a Multifaceted Golden Record
As quickly as the Golden Record is established as a goal, though, questions begin to pile up: What if two parts of the organization need to view an entity differently? What if competing regulatory requirements impose a seemingly unresolvable paradox? What should be done with information for which different parts of the organization have different access permissions? Quickly, the ideal of the Golden Record starts to seem unreachable.
That’s why it’s important to actively shift one’s mindset to adjust the idea of the Golden Record to allow for multiple facets. By doing so, it’s possible to accommodate different views of an entity when doing so is justified by the needs of the business, while simultaneously maintaining a single overarching view of an entity.
In a traditional approach to building a Golden Record, there might be a single column dedicated to the “best” name for a business. And, associated with that column would be rules that dictate how the best business name is determined from among a variety of candidate names, potentially from a variety of sources. Once the best business name is determined, it’s saved on the Golden Record for that entity and distributed throughout the enterprise.
A Multifaceted Approach to Golden Business Names
Let’s say, for example, that the rules around choosing the best business name boil down to using the legal business name wherever it’s available. That rule likely results in spectacular results for most use cases, from invoicing to risk assessment and more. But what if it doesn’t work for marketing? Marketing may have found that prospects are more responsive when marketing collateral references the fictitious name a business has established for itself, such as “Bob’s Tasty Donuts” instead of “Brenner Bakery Enterprises LLC.” But if the “best” business name is decided for them, what should they do?
The answer is to introduce an additional “best” business name onto the Golden Record, replete with its own set of rules for population. It should be stored with a column name that’s meaningful and that identifies it as intended for marketing use cases only.
Why bother putting it onto the Golden Record at all if it’s only intended for marketing purposes? Why not just let marketing build their own algorithm to select the business name to use? There are at least two reasons:
- For marketing to make a good selection, they’d need visibility to all those contributor records, which just introduces more data movements, more opportunity for inadvertent data divergence, and more risk of error.
- Consolidating the determination of the “best marketing” business name with the determination of the rest of the Golden Record values ensures that it benefits from the same level of appropriate oversight.
Avoiding Abuse
Of course, it’s possible to abuse the idea of multiple Golden Record values. Adding new columns to the Golden Record without identifying a strong business justification simply adds to governance costs while reducing the value gained from reusing existing determinations of what is “best.” So being thoughtful about when – and for what reasons – new facets are added to the Golden Record is an important part of successfully implementing this strategy.
In the end, the Golden Record approach need not be dismissed simply because difficult questions are raised about how to accommodate varying business needs throughout an enterprise; rather, by adopting a multifaceted approach to the Golden Record, organizations can find a balance that allows teams to be nimble while maximizing the value of data governance efforts.