A deep dive into HR data analytics — and why it’s broken

Data analysis forms an integral part of the HR function at many different companies today. HR teams are using data analytics to increase employee productivity, boost employee engagement, retain staff, create a better workplace culture, enhance the recruitment process, and so much more. 

But for organizations that have never utilized data-driven HR models and practices before, implementing them can be a minefield. Often, HR teams have questions about what HR analytics platform to choose, overcoming common challenges, and which data trends to watch out for in the future.

Read on to find out the answers to those questions.

Driving change in HR

Over the past few years, many different data analytics platforms have entered the market and are being used by HR leaders globally. But which are seeing the most success, and why?

Olya Panivnyk, chief HR officer at online educational platform Preply, says: “Preply has leveraged Tableau in the past to run its HR dashboard analytics, which focuses primarily on HR trends and turnover rates. We are considering a third-party provider, such as Twine Labs, to run HR analytics for us because it provides broader context into the HR community as well as external benchmarks for success.”

Kevin Gorman, head of people analytics and insights at US telecoms giant Verizon, says the most successful platforms make it easier for their customers to use data to drive change in HR.

“Most firms remain in the early stages of their data journey. Platforms that help customers create a holistic HR data model, identify gaps in data generation/collection, define data quality rules, and make data available to moderately technical users are key to accelerating this journey,” he says.

The seamless consumption of readable and accessible data by decision-makers is critical to driving change, says Gorman. He explains: “Platforms that eschew traditional BI design in favor of modern UI/UX that draws on lessons learned in social media and ecommerce platform development will accelerate adoption among decision-makers and support data-driven decisions.”

Matt Jones, senior vice president of global operations at recruitment process outsourcing provider Cielo, believes the most successful platforms share the same characteristics.

He tells UNLEASH: “Firstly, they integrate well with multiple data sources – for example, Human Capital Management (HCM) or Application Tracking Software (ATS) that adds multiple advertising, sourcing, screening and assessment tools. These platforms can digest, conform or transform disparate data sets into a coherent set of metrics and quantitative evaluations capable of telling a story.

“Secondly, these platforms have mobile-first user interfaces (UI) for users to consume analytics or reporting on demand, regardless of location. Finally, they are ‘near real time’ (as nothing is truly real time) in their refresh rates and insights, allowing decisions to made based on up-to-date data.”

Major challenges

While HR data analytics platforms offer a range of benefits, there are often areas where vendors miss the mark. For example, Panivnyk believes that many lack clear definitions for the HR metrics being collected.

Panivnyk explains: “This inconsistency leads to comparing apples to oranges. Imagine if top-line financial metrics were calculated differently by different providers? This is very often the case with HR metrics.”

Meanwhile, Gorman explains that a common concern for firms is the quality and completeness of HR data. He says:

“Many vendors are focused heavily on the front end design, data visualization and analytics, but less so on mechanisms for ensuring data quality coming into the platform. A deeper focus on tooling that helps quantify and improve data quality as it comes into analytics platforms is needed.”

He also says developing robust machine learning and artificial intelligence systems for HR is difficult due to the complexity of problems, noise in how people arrive at decisions about their careers, as well as lack of large volume and high-quality datasets.

“However, most vendors consistently oversell the capabilities of ML/AI in their products, and often deeper data science work must be done outside the platform or should not be done in the first place given the risk of bad decisions due to data limitations,” he says.

According to Matt Jones, some vendors still rely on manual data intervention, uploading, and manipulation by users. He says: “Inevitably, these platforms continue to fall behind their competitors. Additionally, vendors with closed architecture that does not integrate well with other tools in the company’s Enterprise are not well received.”

COVID-19’s impact

Jacky Cohen, VP of people and culture at global talent mobility platform Topia, says the coronavirus outbreak has also presented HR data analytics challenges.

“The main challenge is that the pandemic has caught existing HR systems by surprise. Traditionally, these systems work on the assumption employees will always be in the same location, but there is an urgent need for HRs to generate and capture data on where their employees are to comply with local employment and tax laws,” she says.

“This need is underlined by our Adapt study, which showed that since March last year only 20% globally, and 30% in the UK, are absolutely confident they have the right information to hand. In the same study, 67% of global respondents have not reported all of the days they have worked outside of the state or country to HR even though 60% knew the rules around reporting where they are working.”

Wendy Batchelder, chief data officer at US software company VMware, says separating noise from value is a big challenge throughout the HR spectrum.

“There is so much data available to us today, that often it is hard for people to sift through it all to find actionable insights. That said, HR teams need to balance effectively between supporting confidentiality and privacy, while at the same time gleaning the insights that can truly aid our employees.”

Batchelder says things get even more complicated if organisations have employees working worldwide, with privacy laws and regulations often varying by country. “For example, certain jurisdictions may have restrictions on where data may reside, resulting, in the inability to survey all employees in the same manner across the globe.”

She says using data from a range of sources to paint a bigger picture can also be challenging. “Often, HR reports HR data through an HR lens. However, data can more often come to life when it’s married with data elsewhere in the business, such as sales,” she says.

“While this creates an opportunity, at the same time this introduces even more complexity around protecting and maintaining continentality for the various data sources. As a result, HR teams often take a conservative approach when it comes to bringing these data sources together.”

A sea change in HR data

With workplace practices, trends, and technology constantly evolving, it is likely that HR data analytics will undergo massive changes over the coming years.

Gorman points out that the rise of remote working has resulted in many new metrics that HR departments must analyze. These include collaboration platform datasets such as email, messaging, video conferencing, and collaborative document editing.

“The new influx of data coupled with managers now managing more remote teams, I expect both the supply of data and demand for insights on how people interact to grow significantly. I think we’ll see dramatically more use of Organizational Network Analysis and sentiment analysis on this large body of connections and text.”

His view is that the advancement of emerging technologies will also transform HR data analytics in the foreseeable future.

“As ML/AI techniques continue to advance, particularly those well suited to small datasets and probabilistic predictions, I anticipate a more dynamic and federated approach to developing high impact, fit for purpose ML models tailored to a firm or even a segment of employees within a firm,” he says.

“Access to powerful ML/AI algorithms and the computing / storage horsepower needed to take advantage of them continues to accelerate. Continued democratization of access to these capabilities coupled with advancements in capabilities that guard against modeling errors and bias will enable even small People Analytics teams to create powerful predictions tailored to their businesses.”

Jones expects a shift from reactive, lagging data and reporting to near real-time insights built on large datasets that organizations can deploy at the point of decision.

He concludes: “Data will elevate HR and talent acquisition into the boardroom with deep, multiple Workforce Scenario Planning, allowing organizations to understand how to build, buy, borrow and automate strategies to deliver effective, engaged, and cost-effective workforces. Data will flow seamlessly with external data points to give holistic views of the world of talent.“

For modern HR departments, the analysis of data is crucial. And while modern technology platforms can help them make sense of growing employee datasets, it is clear that many challenges exist and must be solved.

On top of pre-existing challenges, HR data is set to undergo major changes as new technologies disrupt the market and workplace trends rapidly evolve.