Let me first define “Data Dictionary” — It’s a document that lists data fields/metrics and their standardized definition to be used across the org.
The key here is: Standardized.
Imagine that a management team meeting is going on and you have CEO, VP of Sales, VP of Marketing, CFO, COO among others in the room.
Meeting Agenda: why they didn’t hit the $100M profit goal in the first quarter. So each of them start with the reports they had access to.
VP of Sales says they missed it by $5M
CFO says that they missed it by $9M
COO says that they missed it by $7M
VP of Marketing has three different versions on her report and she is confused!
No ONE talks about the “Why” they missed the goal but instead spends next hour reconciling the numbers!
It was a hypothetical scenario but these things happen all the time! Of course it could be any team meeting and the metric could be something else or it could just that someone is working on something on their own and end up spending a lot of time digging through all the metric definitions and trying to makes sense of it all. This is where data dictionary could help! Let’s take this a step further:
What’s one of the most important characteristic of a good data analysis/science?
It needs to be Actionable.
It needs to help business decision makers take action based on the insights that they found or were shared with them. And before they take that decision, business decision makers need the data they can TRUST!
For data to be trusted, it needs to be understood. It needs to have a definition that everyone agrees upon.
This is what data dictionary is for. It lists data fields/metrics and their standardized definition so that everyone in the org understands what the field/metric means and don’t have to worry about aligning their meaning. They could focus on Analyzing and extracting insights that would change the business and the world!