Data Glossary – A critical first step
As your companies looks at Digital Transformation, being more Data-Driven. and adding Data Governance, the best first step is simply to agree to definitions, and so is born the Business Data Glossary.
What is a Business Data Glossary?
A Business Data Glossary is a glossary of business terms and/or metrics, their definition, and, most important, where the data comes from. You will find it extremely helpful to include any transformations which are needed for the different systems (and/or differences in your business).
Why does it matter?
The first step in governing your data is having a single definition of that data. Though this may seem obvious and evident, many companies don’t realize that even simple metrics like Gross Sales or Net Sales can actually NOT be consistent across their company. As an example, with a $2B manufacturer across six business segments. we found four variations on Net Sales. Some Segments included Shipping (and others did not); and Accruals for returns and rebates to distributors (and others did not).
This is important since these variations mean that you can’t compare the results across the business, nor does combining them represent your true business’s performance.
Even the most simple metrics need to be defined since it is too easy for people to miscommunicate without a common glossary of terms. The more varied and complicated the business is, the more critical it is that everyone has the same understanding of the metric.
Okay, how do you make it happen?
Use the KISS principle here, you don’t want to overpower the data governance where it isn’t needed.
Metric, Definition, Owner
Start with your most important KPIs and measures you are reporting externally. What is the Metric, and its definition? Make sure the definition is clear and concise, this is not where you want people to interpret the meaning. The Owner is the person who owns the Metric, not the Data, so the CFO probably will own all the financial terms. Even if it is apparent, you want to have this documented so if someone doesn’t like the definition, they know who to discuss. Once you have these three elements you can work out any disagreements between the stakeholders. If you are the CFO, for example, and you have financial VP’s and/or Controllers, this is the time to make sure you are all aligned on what the metrics should be.
Source, Source Details, and Transformation, Scope and Owner
For each Segment (or however you break down your business), you want to understand the Source System (like the ERP or QuickBooks); Source Details (is it automatically calculated, or manually entered after some manipulation, etc.; The transformation is how to data elements in the Source system is converted to this metric; the Scope is the scope of the data (maybe business segment, but it could be different for different areas of the business like Domestic vs International), and the Owner is the person responsible at that Segment/Division/Business Unit.
Though the Source, Scope, and Owner are simple, the Source Details and Transformation may take some work. In the Source Details, you want to understand the process that is being used. The simpler the better! Avoid manual processes and promote automated processes. If it is a manual process, what are the steps to make it happen? The Transformation is the formula used to calculate the metric. Though this could be considered part of a Technical Data Dictionary, it is important to have a clear, controlled document on how the metric is calculated so both business tools (like OneStream), and Data Lakes or Warehouses have the same formula so you have Data Integrity.
How long should this take?
For a simple company that only has one Source of Record, this process is probably a day or so process, depending on your Metrics, but the larger the company, the larger the effort. Still, this is potentially the most critical document for important KPI/metrics so that everyone understands the metrics, and HOW the information is calculated.
This document becomes the source of truth, and changes to calculation occur in the document, so everyone has visibility.
Is this all?
There are other fields you can consider (Possible Values, Data Sensitivity, Major Stakeholders, Availability, Refresh Rate, Retention, etc.), but start simple, and grow this over time as it becomes more important to your business.
Our Blog on how your Data helps Digital Transformation: Enterprise Information Management
Our YouTube Channel: Hawksroost Consulting
Check out this article from CIO magazine on Creating a Data-Driven Organization