Master Data and Transaction Data – Introduction
There are two major categories of data required for Business analytics and they are Transaction Data and Master data. What do they stand for and how are they used?
Lets learn with an example
Let’s say you own a company that manufactures clothing and apparels and at this point only manufactures 3 products – Shoes, coats and Caps.
The below is a master table showing the properties of these products – For example, their product codes and which plants they are manufactured in.
Now, after two days of selling your products to customers, you get the below table with details of your transactions in a transaction table
|Product Code||Date(MM/DD/YYYY)||Revenue ($)|
From looking at the above data, what kind of analysis can you do?
It’s quite obvious that nobody’s buying your Shoes, the coat sales seem to be picking up but your Cap Sales are down. Simple analytics on a really small scale – right?
Now let me ask you a simple question – What were you analyzing here?
Yes.. you were analyzing the Product code against each date. Also, how did you know the SH meant shoes? Because of the master table above. The master table also gives you more options for analysis. Since you know Shoes description stands for SH code and they are made in the plant US1, you can do deeper analysis like how many shoes from plant US1 were sold in a particular time period.
The objects under analysis in any situation are called Attributes in SAP HANA . For example – Product, Plant, customer, etc.
The fields which provide us the numbers for tangible analysis like Revenue generated, Quantity sold, Payments Receivable are called Measures in SAP HANA.
The tables containing measures from business transactions are called transactional tables like our second table above whereas the tables containing more information about the master data are called master data tables like our first table.
The central table for analysis is usually a transactional table and if we need a deeper analysis of an attribute, we refer to its attributes from the master data table(s) it may have. Like in our case, we did an analysis of revenue by plant even though we did not have plant in the transaction data. This permits the transactional tables to reduce the number of columns for characteristics since they can always be “looked up” from the master data tables they might have if required for analysis.
I hope the difference was clear and easy to understand. Please share this document using the share buttons below and subscribe to our newsletter to get alerts on the latest additions to this tutorial series.