This blog is co-authored with Halina Pupin who has been advising clients on effective data analytics systems and approaches since 2006 following a successful career with SAP and Rogers Communications.
An earlier blog; “HR Analytics may be the answer…but what is the question” identified the objective of HR Analytics as being to answer business questions dependent upon Human Resource data and below we consider a plan for harvesting the data needed to address these critical enquiries.
Data, data everywhere and so much related to people! The typical HR department has approximately only 40% of employee records stored electronically which translates to at least 500 data points for one employee. That’s powerful data that potentially can produce analytics to support corporate objectives. But before you press the data collection button we’d like to offer some guidelines for hitting the spot.
The most important and this is so obvious we almost hate to mention it, is to clearly identify the question(s) that you want to answer about employee activity before getting started on harvesting data. Clarity on the questions will guide the process and derive meaningful and timely analytical results.
Consider the following:
- Which data elements are you going to need to feed the analysis you want to undertake?
- Where is this data stored e.g. Payroll system, recruitment platform or employee record system(ERS) and whether you are going to have to access data outside of HR e.g. Finance and Sales?
- Can you rely on your data governance to support the quality and accuracy of data that you intend to collect?
- What tools are available to you that can perform the analysis e.g. an ERS with statistical capabilities, a spreadsheet application that will extract the data from selected HRIS platforms or data warehouse and present the results?
Let’s say for example that you were looking for answers to the following:
Q1: Does training improve the performance of our employees?
Q2: What type of training has the most beneficial impact on business performance?
For both you will likely need data elements that include employee name, department, location, performance rating and completed training. You may also want to examine the position obtained and organizational tenure or even base and bonus compensation history as an indicator of prior success. Answers to training questions may depend on data sitting in finance or sales systems. Can you see how quickly the list of data elements can grow and how important it is to clearly define the enquiry in order to generate the most valuable response? Performance rating information may be in an ERS or HRIS database which is not integrated with the software used to capture training records. Compensation data will be sourced from perhaps payroll, ERS or HRIS. The more the number of sources to be accessed the more important the need to plan the gathering and expected accuracy of data to manage the scheduling of reporting your results. Considerations are compounded if elements come from systems outside of your control in other departments or even under the control of an external vendor. Planning data extraction with your IT team is strongly recommended when timelines are tight and data is coming from multiple sources.
Data governance considerations are important at this point too. Who can vouch for its quality and accuracy? This may be several people if more than one database is being accessed. This may determine whether your analytics are directional and require further validation or definitive and can be totally relied upon.
Planning an analytical study should include the availability and types of tools that are to be used for collection and reporting. Will a spreadsheet application or a data warehouse be employed and who will do the analysis? Are they qualified in statistical scrutiny and experienced in using the software tools available? Must they be part of the HR team or can an IT colleague extract and analyze the data?
Finally, is the presentation of the outcome. By this stage the results may be self explanatory to you but the boss or the Board approaching the data for the first time will require additional clarity about the results. Don’t jeopardise all of this hard work by underestimating the importance of introducing the purpose of the project before presenting its conclusions.