systematicHR

Continued

Business Intelligence: ODS
So once you’ve gotten through your first round of ETL, you’ll have to load the data somewhere (the “L” in ETL). The ODS, or operational data store, is where all the raw data gets loaded. This is basically a database that will house all of the raw data in a completely un-manipulated way with all the history preserved. There are several wonderful things about this. First of all, it might be the only place where you actually have HR, benefits, payroll, talent, learning, succession, recruiting data, etc. data all in one place. Certainly I can tell you with 99% confidence that you don’t have this today.

The ODS it turns out, is where you’ll run some of your operational reports and might be where you decide to drive data out to other consumer systems. I’ve had lengthy arguments in my life about the use of a data warehouse as a “data hub” but I’m going to say that it’s ok to use the ODS as a data hub. Let me explain. Let’s say you have a intranet address book that allows employees to look each other up and see some sort of an organizational hierarchy. Your problem is that it’s easy for you to provide the personal data out of the HRMS, but the reports to data is housed in the talent management system. Therefore, for this one simple interface, you need 2 file feeds. Add to that the intranet also want to have learning management so it can publish out open courses for employee registration. All of a sudden you start adding interfaces just so you can populate the intranet with accurate data. If you decided to bring all of the data into the ODS portion of the data warehouse, you could then manage all consumer systems with just one feed from a single place.

The ODS is also a very convenient place to run some of your reporting. As I stated earlier, it’s where you have most or all of your data in one place. The problem is that once we get into the next phase of data warehouse (“enrichment”) that data may no longer be usable for certain types of reports. More on this later. Suffice to say that there may be some desire to use the ODS for light reporting.

Business Intelligence: Enriched DW
Once you’ve loaded all of your multiple HR data sources into the ODS, you’ll also want to enrich the data. This means you’ll need to go through a second ETL process. This ETL process rests heavily on the “T” or translate. What the enrichment is actually doing is converting the raw data from generally unusable form from an aggregate perspective and applying data rules and definitions to it. I promised earlier not to get technical, so I’ll leave it by saying that once you have enriched the data, you still have it stored in a similar fashion as the ODS, but the data has now been modified to enhance future analytical reporting. This is an absolutely necessary step in the process.

Let’s take a simple HR situation. Let’s say for example you are a global organization. Each of your countries uses their own job code structure. You can see how each country would have liked using the ODS to report on aggregated HR data using their own job codes. However, once you aggregate the entire organization, the data would be useless without a standard set of job codes to compare each organization against. Therefore, the country’s job codes will go through a translation that eventually allows full roll-up through the organization.

More tomorrow