Skip to content

Embracing Data for Stronger People Analytics

In the last blog post, ‘The Power of Data to Transform People Analytics’, we talked about data as a cornerstone in any decision-making system. I want to spend more time here talking about what that means and how to approach data as it relates to People Analytics.

Unlike finance, marketing, and sales departments, human resource professionals have been largely deprived of access to modern platforms that exploit advancements in cloud-based machine learning, data science, and artificial intelligence. HR infrastructure, architected over twenty years ago to address pay, cost containment, and compliance functions, was never designed to handle the nuance, volume, and variability of worker information that fundamentally shapes a company’s top-line revenue and growth.  

Eskalera is looking to use our expertise on data-driven systems to help lead HR into the machine learning fold.  A good people analytics system should strive to incorporate as much data as possible to build multiple signals and model a variety of outcomes. Available data for people analytics can come not only from HR systems, but from the systems used every day to work and operate the business.  We think about potential data sources based on the following dimensions:

Machine learning techniques allow us to process and generate insights from unstructured and real-time data. People Analytics should move beyond ingesting just HR system data and take advantage of all these other data sources for richer, more in-depth analysis.

With far more data available, how do you know which of these data sources are good? In the world of big data, data does not have to be perfect. Data techniques have advanced to quickly identify clean signals.   Machine Learning allows us to start and test with small datasets, then create stronger models as we gather more data. For any data system, the more important question to ask isn’t “how much data?” but “do I have the right data?”

When evaluating data sources for any system, you should consider the following:

Is it reliable?

Is data being collected in a way that it is not biased and truly reflects the workforce? For example, if your employee engagement surveys always follow a bonus and promotion cycle, then sentiment is often skewed based on the employees’ sense of satisfaction from the recent reward.

Is it real-time?

Real-time data has the benefit of being a continuous source of insight that also helps us see quickly changing patterns. Capturing employees’ sentiments and actions through real-time collaboration systems can be more authentic and unscripted relative to surveys. This real-time, passive data collection also minimizes the burden on the employee to provide information outside their regular workflow.

Is it robust?

When making a data-driven decision, it is better to have multiple data points to help analyze and give context to the question at hand. For example, sentiment of inclusion can be captured through culture surveys, but the inclusion signal becomes stronger when paired with evidence of workplace networking opportunities. Overlaying frequency of 1:1s with managers, project assignments, and access to mentors can provide additional context in understanding inclusion and opportunities for advancement.

One of the biggest concerns companies have is around how employees may react to having their data analyzed in such a granular manner. Today’s modern HR systems are built using Privacy-by-Design frameworks, and there are data techniques that provide aggregated insights for management without identifying individuals. In addition to those assurances, insights can also be surfaced back to the employee, which is an opportunity to articulate the value back to the individual. It comes down to transparency; communicate to employees what data is being gathered, when and why, and how can it make their lives better. 

With today’s tools and techniques, you can look beyond just traditional HRIS data to power People Analytics.  By integrating real-time systems, passive data streams, and the collaboration tools that your employees are currently using every day, you can get a more complete picture about culture, collaboration, productivity, and engagement. The great news is that you don’t have to start all at once. As you add more data sources, the signal strengthens and you’ll have more context to answer a richer set of business questions. In the next blog post we’ll discuss how to think about the different data signals relative to your business objectives, which is where HR expertise is critical in harnessing the power of data.

Obtain real-time insights on culture and inclusion that serve as an early warning system for regrettable employee attrition and productivity loss. Contact us today for a demo!