Breaking down barriers to data-driven HR

Data can be used to predict churn, track employee engagement, measure performance and leadership – the list goes on. We have found that embracing these opportunities is vital to ensure a business remains competitive.

Data-driven HR is already in place in many leading global businesses.

Amazon has pledged an ambition to be “the most scientific HR organisation in the world”, using data to inform hiring, develop teams and retain top performers. IBM has used a social listening algorithm to monitor staff message boards and performance reviews, filtering for recurring complaints and analysing feedback to improve employee experience. As technology in this area becomes readily available and affordable, it’s also trickling down to smaller businesses. Here especially, insights can drive improvements in recruitment, retention and skills development.

Investors and shareholders are looking for deeper, more accurate insight on their teams.

Businesses often claim that employees are their greatest asset. HR data can prove that this is the case, giving investors and shareholders a detailed view of where value is, or could be, created in the business. Using HR data to provide this quantitative evidence can enable effective management decisions that reflect the value of the workforce.

Data is necessity for legal compliance.

Even companies that are not actively looking to deploy workforce analytics must engage with compliance, pay gap and data breach reporting. Not using data at all is no longer an option for businesses of any size.

…but where to begin?

Building a “data-driven HR function” might sound intimidating, particularly to teams that see analytics as a ‘nice-to-have’, rather than fundamental to day-to-day HR operations. But building the right capabilities doesn’t need to be a transformation or overhaul. It is about targeting information that is accessible and impactful.

Step 01 – Identify the end goal

It is easy to be distracted by interesting results and lose focus on the overall goal. Data collection and analytics needs a purpose, which will define the data and skills required to support decisions and actions.

In a company with unusually high attrition rates, the overarching goal could be to drive retention of valuable employees and support resource planning. Having established these objectives, the business can look to test their hypotheses using internal and external data sources to better understand the factors influencing attrition.

Example Hypotheses Data Questions answered
Employees are not happy with levels of training and development
  • Employee surveys
  • L&D records
  • Employee benchmarking
  • Are employees at all grades satisfied with the training they receive?
  • Do employees carry out the training? If not, why not?
  • Do employees grow skills and competency at all grades?
  • How does the L&D programme compare to similar programmes run by competitors?
Employees are not incentivised to stay
  • Remuneration benchmarking
  • Leaver records
  • Historical exit interviews
  • Are employees paid in line with market expectations?
  • Are employees leaving at the end of a remuneration cycle?
  • What factors incentivise employee groups in addition to remuneration? How is the businesses delivering these?
The business is not hiring the right candidates
  • Recruitment records
  • Timesheet reporting
  • Performance review records
  • Employee profiling
  • Are there candidate skillsets that thrive in particular roles?
  • Do job descriptions match the activities that employees ultimately conduct in the role?
  • Are there biases in the recruitment system that favours candidates who are less fitted to the role?


Step 02 – Prioritise use cases

Once the overall goal has been set, HR teams can devise and prioritise the steps required to achieve it. Those further down the data maturity curve often have to be realistic about what is feasible at first.

Let’s return to that company tackling unusually high attrition rates. The first step is to ensure that basic organisation data is available, particularly clear, standardised information around leavers and joiners. This will provide a reliable baseline to inform actions that target the factors influencing attrition. It also helps the business scope their requirements for robust data governance and foundations for future data initiatives.

Step 03 – Understanding the critical audiences

Business and stakeholder requirements have to shape the format, presentation and location of any outputs, or they may never be translated into actionable insights. Interactive dashboards are invaluable for audiences requiring insight on the overall picture, as well as geographical or functional drill downs. Businesses used to sharing PowerPoint reports may prefer automated publication into existing templates.

Some of our clients look to build a custom tool or module within their existing systems architecture, which can collect, interpret and present the data alongside a range of other management information to enhance senior decision making. The company aiming to reduce its attrition rates could use a tool like this to track the results of targeted interventions on attrition and a range of other business performance indicators.

Understanding the audience is a crucial step here – it is the key to ensuring that data and analytics are user-friendly and commercially impactful.

It’s not as big a challenge as it seems

This can seem like a daunting task, but the nature of HR data makes analytics as achievable as it is effective. Most companies have hundreds or thousands of employees – not millions, and the number of observations associated with those employees is rarely large enough to be considered “big data”.

The starting point of HR analytics is the data gathered on the people in the business: roles and hierarchy, payroll, demographic and skills data. Even a small number of data points on each employee – start/leave date, qualifications, demographic, manager and compensation history – can generate insights that drive employee satisfaction, save costs and deliver value across the business.

More complex analytics involving third party data, instantaneous data or even IoT sensors can always be layered on top of this basic people data for specific analytics projects in the future. But for the most part, things need not be too complicated.

We find that the same is true for all functions looking to use data more effectively to make decisions, and it has informed the JMAN approach when we are working with clients. We look for quick wins and high impact priority areas, at the same time as supporting our clients think about how to build data maturity longer term. By selecting the insights needed most urgently and targeting rapid results, we help develop actionable insight at speed, build data literacy and break down the barriers to becoming truly data-driven.