Most normal HRIS people who sit around dealing with their normal HRIS and talent management systems haven’t given much thought to business intelligence much beyond the standard reporting capabilities provided by their applications of choice. A few might have gone out and purchased new reporting tools from Cognos or Business Objects for their ad hoc reports, but these are very operational in nature. There is much more to business intelligence than ad hoc operational reports.
As we have all heard, there is a wonderful world of complex analytics that can be implemented, but there are very often the domain of the more centralized IT departments to implement. However, there isn’t much of a reason for this. In fact, I’d argue that HR want’s their own instance of the data warehouse for privacy and security purposes. After all, you don’t want just everyone looking over payroll records, ethnicity data and whatever else. Not only that, but we also have our own network of technologies that need to be brought together, and often the corporate data warehouse guys are much more concerned with ERP elements only.
When HR talks about data warehouse, we’re looking at multiple vendors for possibly benefits and payroll or other outsourced and SaaS environments, multiple applications such as all of the talent management components, and of course not to leave out the core HRMS system. It’s not always that easy though. Different divisions will operate their own programs, so you could potentially have multiples of all of the above, and if you’re global, it’s that much more complex.
However, may HRIS departments simply are not equipped to implement and manage a data warehouse on their own simply because they don’t have the specific expertise. As I noted above, we have sufficient need to have DW’s of our own. Over the next few days, I’ll go through some of the core functions in any data warehouse and how they work. Hope you enjoy.
Business Intelligence: ETL
The true beginning of any data warehouse is the ETL. So what exactly is ETL? In general, it’s known as “extract, translate, load” functionality. While it seems like a simple function, the sophistication of the ETL tools has grown exponentially over the last few years.
In terms of ETL, my opinion is that the extract and load functions are really the least of the functionality. The extract can really handle any sort of file interface that you can dream up. What it is useful for is the management and handling of many feeds.
Things really get interesting when you start talking about the translate features. When we talk about disparate applications, possibly multiple organizations, and potentially different cultures, programs and languages, the translate function really starts to shine. On the easiest level, if you look at performance scores, your corporate location might use a 5 point scale. The field sales offices might use a seven point scale, and the multiple divisions using other plans. You could argue that this functionality can easily be programmed into the data feed with some simple logic to map values from one system to another. This is completely true, but when you add the other components of HR that need translation (job code structures, organizational identifiers, salary grades…) it becomes evident that the management of the translations would be simpler if everything was in one place.
2 responses to “Business Intelligence: Introduction and ETL”
[…] Business Intelligence: Introduction and ETL | Home […]
A disparate technology environment is a reality for most HR organizations. However, companies can have full Talent Measurement capability without a data warehouse. New workforce analytic tools, like DoubleStar’s Workforce Insight OnDemand, are now available on a subscription, web access basis. While the early adopters of workforce analytics had to build and maintain BI environments, the industry is moving to SaaS models, so more organizations can achieve talent optimization with a robust talent measurment solution.