Data Doesn’t Have to be a Four Letter Word

Is your workforce data disparate, incomplete, inaccurate, and constantly changing? If so, you are not alone. HR data is inherently bad and difficult to integrate.

Here’s the good news: You don’t need to let dirty data hold you up on your workforce analytics journey. Here are four top tips for dealing with your dirty data.

Tip #1: Don’t cleanse data for cleansing sake

When using workforce analytics, your main goal is to get insight and information that will drive better business decisions. The insight that you can pull from the data is the most important element.

Let’s say you want to know whether you are retaining the right people to support next year’s growth target. If you only focus on cleaning your data, a year will pass by very quickly — projects that solely exist for the purpose of getting data cleaned, aligned, and normalized tend to have low business priority. All the questions you had in the first place about required skills and aptitudes will be irrelevant by the time you have all the data in hand. It is more effective to put time and resources into projects that have a direct link to business objectives.

Tip #2: Aim for accuracy instead of perfection

Indeed, good business decisions require accurate data. But the data does not need to be perfect all the time.  Let’s put on our finance hats for a moment: How often do you think financial decisions are changed because of adjusted costs, restated data, or realignments? Countless times – and those perpetual adjustments are widely accepted.

There are times when you will need to be extremely precise, however. Which brings me to my next point…..

Tip #3: Map accuracy to impactgreenzone

Analytics is about making decisions, but not all decisions are equally important. When cleansing data, you want to consider the impact of the decision. For example, there are times when you need to have near-pristine data, such as mapping succession candidates to future positions or determining changes to compensation.

On the other hand, if your organization is running analytics for a low impact decision, such as monitoring headcount change or tenure rates, striving for 100% accuracy is inefficient. In these circumstances, missing 1 person among one thousand people gives an error smaller than +/- 0.01.

In short, you want to aim for the green zone (shown here), allocating your resources appropriately.

Tip #4: Work with dirty data while improving it

What’s the best way to clean data? Start with a business question, such as: “are we retaining the right people?” Next, you can bring all your relevant data to the light, and work with it while improving it. This approach drives two favorable results:

  1. You won’t let good data go unused. Get useful information out of it while you navigate through your remaining data. This will drive more interest in improving the overall quality of data.
  2. The people who are responsible for the data and accountable for the decisions that are based on that data will be given a more compelling reason to get the data in the system, get it right, and get it done.

Follow these tips and you will be on the workforce analytics fast-track. Stay tuned: in the next post we will cover how to get started with workforce analytics using a step-by-step approach.

Author Photo
Ian Cook |
Curious about the differences between gaussian and pareto distribution? Ask Ian. Want to know what it’s like to kite ski North of the Arctic Circle? Ask Ian. Not only is he an expert in statistical analysis and HR metrics, he’s also an avid cyclist, skier and runner. At Visier, Ian helps customers drive organizational change through linking workforce analysis to business outcomes. He is responsible for the workforce domain expertise within the Visier solutions.