Skip to main content
Platform blog

Since COVID, countless articles have been written about the "Great Resignation", including in-depth analysis by the World Economic Forum. One key thing this research found is that in order to prevent even more people from voluntarily quitting their jobs, companies must work harder to engage their people at work. Employees are, after all, the most valuable assets of any business, and when an employee leaves, it takes a toll on colleagues and on the wider enterprise.

But when it comes to examining employee engagement, few companies rely on data as a core resource. While the last five years have seen more executives using analytics and machine learning to deliver deeper, broader insights into their customers and supply chains, they still often lack data on their people. Instead, human resources professionals have tended to focus on hiring or coaching and development, needing to use their "gut feel" instead of data to drive decisions.

One of the best-kept secrets in business is that unlocking human capital data can lead to growth and cost savings – all while preparing companies for the future of work. No surprise then, that as they become more widely recognized for their key role in driving enterprise value, leading CHROs are now increasingly relying on data.

89% of CEOs say that the CHRO should play a central role in driving long-term profitable growth!

(Details of this report can be found on Accenture.com here: CHRO as Growth Executive)

Greater data-centricity – in human resources, as well as other core functions – delivers positive business outcomes, including top-line growth. In one example, a global franchise retailer recently worked with Databricks and Accenture to mine employee data and was able to create trackable and quantifiable value in less than 18 months.

A retailers journey to use data to keep employees engaged

Here are a few key lessons learned by this big box multi-national retailer on how using data can keep employees growing in their roles, valuing their success to the bottom line – and illustrating some really good HR practices at the same time.

  1. Create a solution to solve for a true business need

On the road to digital transformation, this retailer was building a shared data cloud infrastructure. Along the way, the CHRO of the business expressed an urgent need to track corporate goals around "human capital."

Until this transformation journey began, data usage and availability in the HR domain at this organization had both been managed in an ad hoc, rather than systematic way. Once requested, an insight or analysis would take weeks or months to obtain, contain hard-to-follow data caveats and, when compared to other one-off requests, the results could not be harmonized (did not correspond).

Taking detailed business needs as its north star for change, the retailer focused on capturing actionable employee data, with valuable time spent in workshops to prioritize both data accessibility and the value of insight.

A vital goal was to align the ways in which the different inflections in an employee's journey could directly impact the business. Specifically, the team aimed to understand how people at various stages of their tenure could drive bottom-line results.

One of the hypotheses tested by the team was that product quality drops (leading to a reduction in selling price) when employees were finishing their contract periods or permanent employees contemplating putting in their notice. Seasonal contractors, for instance, showed a marked decline in both output and product quality as their departure time neared. For HR professionals, insights were needed to show how this could be avoided.

  1. Ensure privacy and data governance from the start

To ensure appropriate handling of sensitive HR data, Databricks came to the rescue with their Data Intelligence Platform built on a lakehouse architecture. Databricks enabled this retailer to pull original source data in three layers, with mining only beginning on the "silver" data (structured, clean, and consumable data). With no identifiable data being handled, irreversible hashing techniques were used to ensure that individual employees could not be traced.

As the team selected the data needed to drive business value, it was essential to ensure that the data was as accurate and clean as possible to ensure the ability to run deep analytics. [NP1]

As an initial step, the retailer outlined eight different HR sources of data (all appropriately masked and summarized by design). These included topics like:

  • Demographics
  • Capability development
  • Experiences
  • Employment lifecycle
  • Growth and performance
  • Satisfaction and other survey materials
  • Hiring and tracking details
  1. Get your wins early

In mining this data, the data team at the retailer ensured they could easily bring together different sources information from both the HR domain as well as across other areas of the business such as customer satisfaction and sales. In addition to the ability to link privacy protected individual data across the business, management also requested timely refreshes of all information.

Staying on top of both the new influx of people and, as importantly, those who are leaving, has become a staple in "people currency". Management continues to closely track the reasons for departure, and there are several identified triggers that became early indicator for employee exit that are trackable in overall business outcomes.

Once those triggers were identified at the retailer, teams were able to create AI models to predict the most "likely to leave" profiles, which was vastly successful across the organization. Given the importance of employee retention to overall business success, it was critical to share these KPIs outside headquarters with its franchisees.

There are many ways in which the core capabilities of Databricks made this journey possible, straightforward and quick to scale. The combination of Databricks' scalability, simplicity and ability to run on any cloud helped with sharing metrics and formats across different countries and regions. As a result, the retailer has found many opportunities to leverage this solution in other scenarios, like comparing sales figures (traditionally available), as well as human capital measures in hiring, job availability and role movements.

  1. Use generative AI for more than just employee retention

The retailer found that they could also use the simplicity of Databricks to help with recruitment. The team tuned the model on top of resumes for quick-to-read and ready-to-assess skills on incoming applicants. This has made it easy to review resumes, assess capabilities and quickly move candidates through varied talent levels for hiring.

To do this, AI algorithms for natural language processing are used to read through the vast amounts of text in applications and identify skills, qualifications, role descriptions, job postings and capabilities.

When the power of generative AI is applied to HR data, the model can provide the incremental ability to better parse through data and display it in an easier-to-read fashion. This gives HR teams the ability to select any random resume and rapidly identify the candidate's top skills and capabilities – talent professionals can now scan a broad set of capabilities, move faster to engage directly and increase the speed of decisions.

Using Databricks for HR Analytics

Along the way, the team at the retailer has used several capabilities that are unique to Databricks. One team member explains "in the age of exploding data, I cannot imagine using anything but Databricks moving forward," with another adding "after 25 years as a data scientist, I cannot believe how easy it is to create and run models – Mlflow has changed my life forever."

The data team at the retailer is most excited about the results. Currently, the team is using advanced analytics to identify correlations between client satisfaction scores and ongoing training initiatives, which has led to a number of quantified improvements:

  • 500+% growth in user adoption across internal users
  • 300+% decrease in refresh of core KPIs
  • 250% increase in available and trackable KPIs
  • Double-digit % reduction in staff turnover in one of five major divisions.

We're all living and working in a world where the pace of digitization keeps on accelerating. For companies striving to retain and engage their employees, one of the best-kept secrets is that gathering and utilizing people's insights can be easier than you think – and even more beneficial than you'd imagine to both the business and longer-term employee satisfaction.

If you'd like to learn more, please visit Accenture for details on The Future of Work

Try Databricks for free

Related posts

Company blog

Databricks Named a Leader in 2023 Gartner® Magic Quadrant™ for Cloud Database Management Systems

We are excited to announce that Gartner has recognized Databricks as a Leader for a third consecutive year in the 2023 Gartner® Magic...
Company blog

Databricks Data Intelligence Platform for Retail comes to NRF 2024

December 19, 2023 by Sam Sawyer and Rob Saker in Company Blog
Request a meeting with Databricks executives/thought leaders at NRF! Each January, thousands of leaders from retailers around the globe gather at Javits Center...
Industries category icon 1

Retail in the Age of Generative AI

April 13, 2023 by Sam Steiny, Rob Saker and Bryan Smith in Industries
What is generative AI? Large Language Models (LLMs) such as ChatGPT have garnered a ton of attention lately along with the broader topic...
See all Platform Blog posts