So your company is investing in a new talent management system and you think you'll get all kinds of great data from it, right? Well...maybe.
Just because you buy the latest HR technology product doesn’t mean you’ll get good business intelligence. Technology is not the key here. Solid business processes that result in accurate and timely information are the key.
Human resource data is particularly vulnerable to poor maintenance because the people entering the data are managers and employees that tend to approach the task with resentment (i.e. "doing HR’s work for them"). Not only is the data poorly maintained, but also the same data is reinvented many times over by different branches of HR that use the data for different purposes. What's more, most HR teams don't run quality audits on their data to ensure that data entry is accurate or that data is updated as necessary.
Here are some steps a company can take to trust their data and utilize technology tools.
Step 1: Find the Root Cause of the Inaccuracy
We often dump raw data and manipulate it with spreadsheets to make it look pretty. If the data provided to clients is wrong, find the root cause and fix it. It's often a simple fix, but you need to dig for it.
Use the five whys to find the root cause.
A termination code is wrong on a report.
- Why? Because the manager entered an incorrect code.
- Why? Because the manager was using an outdated code listing on the intranet.
- Why? Because the code listing had not been updated.
- Why? Because the HR representative who was responsible for updating the code listing on the intranet had been out for an extended leave and no one knew to make the change.
- Why? Because all the responsibility is on one person to remember to update the codes.
If you simply changed the code in the system to the correct code, it would still be wrong the next time a manager refers to the code listing on the intranet. If however, you find the root cause, you not only fix the error, but also the systemic problem that led to the error — the process was too dependent on one person.
Step 2: Make Sure Your Hierarchy is Designed to Distribute Employee Data to Managers Correctly
Many HR systems are fed from payroll that is organized around the accounting department. In our increasingly complex organizations, however, hierarchies aren’t that simple. Employees may be paid by one department, yet may report to a leader in a different department. The unit leader may be part of a specific department for accounting purposes, but in distribution performance and talent reports and processes, they should actually show under their direct manager, not their leader.
This is a great opportunity to think collectively across HR silos, identify different organizational structure needs, and design a structure that works for each need.
Step 3: Get Your Job Data Straight
An HR system is, at its core, a matching system that contains both employee data and job data to be matched depending upon the purpose. Compensation? The employee is assigned to a job code. Recruiting? The vacancy is assigned a job code. Learning and development? The competencies and learning paths are linked to a job code.
Job codes are both the most powerful and most misused elements of HR data. If an employee is assigned to the wrong job code, their salary grade, their EEO code, their bonus structure and possibly their benefits will be wrong. Companies must think collectively across HR silos and create job codes and job data that works across all HR functions.
Step 4: Make Sure Your Systems Talk to Each Other
One system for recruiting, another for learning, and yet another for the HRIS, may be nicely customized for the HR function using it, but it doesn’t translate across functions. HR needs to talk the same language for the sake of their customers, meaning that a job code to a recruiter should mean the same to a compensation representative.
Step 5: Audit Your Data
It takes work and it takes time, but your credibility is worth it.
Before you open the door to HR technology, be sure you have the key!
Photo: Can Stock