Human Capital Analytics: 6 Steps to Better Data

By Astrid Wonderley on June 9, 2022

Human capital analytics is fast becoming one of the most valuable strategies a company can use to gain competitive advantage. A 2018 report from Deloitte found that 84 percent of executives now consider people analytics important/very important, and 70 percent said they were in the midst of launching major projects to analyze and integrate human capital data into their decision-making.

Imagine if you could click a button on a dashboard and immediately see which employees in your company have a specific in-demand skill set, what skills your business will need in the next 24 months to remain competitive, or which manager is causing your best employees to quietly look for another job. All of this is possible with the right human capital analytics setup.



At the highest level, companies like IBM are using human capital analytics to predict with almost total accuracy who’s about to quit before it happens, which the company says has saved it nearly $300 million in retention costs.

Other companies are using this data to rapidly identify where their best hires come from, what skills and traits their high performers have in common, and how data captured from candidate assessments correlate with job performance. They can use these insights to hone job descriptions, prioritize job board spending, and create hiring profiles using assessment data to make more reliable recruiting decisions.

It all sounds great but there’s a catch – human capital analytics only work if the data you gather about your employees is thorough, complete, free from errors, and consistent across all databases. For that to be the case, your company needs formal guidelines in place defining every title, job description, and performance management report, and every person who records human capital data must adhere to this terminology for every field, file, and record they touch.

Don’t feel bad if you don’t fit this description. Most companies don’t.

A 2019 study conducted by PWC found that only four percent of companies had enough clean and structured human capital data to run effective analytics, while more than half (55%) received a failing grade for their data management techniques. The study found that most companies lack any consistent language or rules for capturing data, and most companies use so many different data platforms that they can’t access or analyze information in any meaningful way. 

One of the biggest negative impacts of this ‘dirty data’ trend is that it prevents companies from improving their recruiting process. When human capital data is error-prone, inconsistent, and trapped in siloed databases, meaningful analysis is impossible, forcing recruiters to rely on best guesses, their own experience, and what limited data they can access to make the best choice.


6 steps to cleaner data

Cleaning data is a complicated and time-consuming process, but the payoff can be significant. Here’s how to get started.

  1. Link workforce analytics to business results.
    Data cleaning projects will require significant financial and human resources, so build a business case that outlines how clean data will add strategic value for the company – i.e. lower recruiting costs, greater agility, skills aligned with business goals, etc.
  2. Assess your data platforms.
    Having fewer data platforms result in less data caught in silos and greater data compatibility. Evaluate which systems you currently use and why then decide whether an upgrade or elimination of outdated tools will support your clean data goals.
  3. Work collaboratively.
    These projects are equal parts technology and human capital data, so they require joint efforts from HR, who can clean the data; and IT who can update the technology and ensure all databases will work together.
  4. Rewrite the rules.
    There is no point in cleaning old data unless you simultaneously change how new data will be written, captured, and shared. That means redefining all job titles and descriptions, roles, salaries and performance management language, and setting clear metrics and goals for data collection.
  5. Assign a data monitor.
    Train everyone who captures data on how the new rules work, and why they are important, then establish quarterly data reviews to assess and clean data as appropriate, and encourage better behavior where mistakes are being made.
  6. Start with a pilot project.
    Trying to clean every piece of data all at once will feel overwhelming and could take so long that stakeholders will start to doubt its worth. Instead, pick one platform, set of data, or group of employees then clean up all of that data and run some useful analyses. It’s a quick way to demonstrate the benefits of clean human capital data, and it provides lots of opportunities to compare results between clean and dirty data sets.

When you start with a long-term plan, cross-functional team, and executive support, creating and keeping clean data is possible, and the resulting benefits could give you the edge you need to hire and keep the best talent in the marketplace.