People Analytics Is Back, and This Time With a Passion

HR analytics, training analytics, and team analytics were formerly considered minor aspects of Human Resources. Perhaps an I/O Psychologist assessed engagement levels, a data analyst assessed the impact of various training programs, or a recruiter assessed job postings for better possibilities.

One or two data science tools (usually spreadsheets) were used to aid in the collecting of data in the appropriate HR area, and a limited budget was set aside for analysis tools (spreadsheets were always the most popular).

For example, let's look at the ROI of different training programs, employee engagement, where managers should concentrate on workers, pay inequity, the distribution of performance ratings, and other subjects that helped the organization enhance its various HR programs.

Nowadays, CEOs and CFOs acknowledge the need of assessing diversity, gender pay equity, skills gaps, labor utilization, retention rates, real-time feedback, and even organizational network analysis.

It is estimated that 69% of companies will establish a People Analytics database this year. A big rise in expenditures indicates that this field has grown.

Companies at Level 4 of the High-Impact People Analytics maturity model think they have accurate people data, 95 percent feel they have good privacy and security standards, and 75 percent believe they have consistent data definitions. While this is still a tiny sample, it shows how easily "world-class" may be defined.

This occurred for two reasons. First, the board expects CEOs and CHROs to report on pay fairness, diversity, and skill shortages. Second, modern integrated cloud Human Capital Management systems (about 40% of leading organizations currently use cloud-based HCM systems) need a more consistent system of record.

The recent Sierra-Cedar HR study found that the typical organization now had over 7 "systems of record" for personnel data. Employer-sponsored health insurance (HSI) But integrating this data has never been simpler, and most large firms now have Hadoop clusters and data lakes they can put up to bring it all together.

Employee data is now all over the place, and it is growing quickly. Almost all businesses have HR metrics on things like organizational compliance and how their employees move around.

Pulse surveys and continuous performance management tools aren't the only sources of data that companies now have about their employees. They also have near-real-time data about how employees communicate and work together, where they go, and how well they're doing (from wellbeing apps and voluntary data).

These include social media for both inside and outside the company, ERP systems, surveys, and evaluating information in corporate communication platforms. If you use a modern email system, you can do "organizational network analysis" (ONA) of the metadata in your emails, which lets HR analytics be done.

If you don't like it, some companies now make software that scans emails and finds "mood shifts" in team or company communication. "Stress" in a company can be shown by data, and its algorithms can look for possible fraud or failed customer projects with ease. There are now many tools that can tell if we're stressed by how we sound.

One of Josh Bersin's customers told him about a study they did to find out how well their engineering teams did. Engineers were asked to wear smart tags to find out what makes them happy and productive at work, so they could figure that out.

"Happy" engineers were the ones who traveled the most, had a lot of people they knew and spent the most time socializing. This data was used to reorganize facilities, improve meeting management, and get engineers to spend more time with their coworkers. Almost any business can now do this.

The new study shows how JetBlue uses a lot of different data sources to look at things like attrition trends, engagement factors, flight delays, and productivity, and how it does this. Doing this gives the company a complete picture of how happy their employees are and how well they serve their customers. They do this by looking at employee feedback, team and customer complaints, HR information systems (HRIS), training, flight activity data, and other information.

That dream is beginning to come true. The new study demonstrates that the capabilities of HR business partners, analysts, and team members are important predictors of Team Analytics success.
Level 4 organizations have a much more diverse collection of analytics capabilities than companies at lower levels of maturity — I'd say they've established a new "bar" for their staff. (In Level 4 organizations, 63% of HR professionals are highly analytics literate, compared to 20% in Level 1 firms.)

It's time for HR professionals to become data scientists!

It's simple. The power of analysis nowadays is not usually the line manager looking at a dashboard to find out why someone is leaving their group (they don't have the time or motivation).

To demonstrate a senior business leader that their staff has a prejudice, bad work habits, lack of skills, bad culture, or other issues that can be shown with some amounts of data.

They won't listen to an HR expert flapping their hands about how horrible the "culture" is, or how "biased" or "unfair" the corporation has become. They want plenty of data to show what's wrong and data-driven improvement suggestions. The line leader won't listen if the HR expert can't make that case, present the facts effectively, and defend their analysis.

Having evidence-based data isn't the issue; it's knowing how to explain, illustrate, and apply it to business decisions. With an MBA or a background in statistics, the business executive may question you where the data came from and how you arrived at your conclusion.

Currently, HR management teams are not analytic enough, according to many. In 2018, you should focus on "energizing your HR organization" with seminars on human resources management, programs, and exercises in statistics, big data analysis, and data-driven suggestions.

Finally, sophisticated statistics, neural networks, and other kinds of machine learning have come. A recent case study by a team of AI professionals demonstrates that courses in AI are increasing in popularity. Now in the workforce, these experts are eager to tackle challenging data challenges.

(In fact, machine learning is mostly math. So any major company may design or utilize these algorithms through public domain APIs.)

The author spoke with talent management departments looking at attrition trends, performance and retention prediction models, employee absence and grievance models, and many more types of employee productivity based on workforce analytics data from communication platforms, like Microsoft Teams. These firms are now able to discover things about their company that they never believed imaginable.

For example, one vendor now provides a solution that analyzes biannual engagement surveys and advises direct behavioral adjustments to managers to boost team engagement and productivity. Another organization created a machine-learning system to identify their top salespeople's conduct and how to coach others to perform better. Many professional services businesses study high-performing consultants' communication and travel habits to learn from them.

We used to assume "skills" were the key to productivity. Now we know that the key is in the "behaviors," "habits," and "patterns" of successful individuals. Intuitively, the software can examine and understand many of these.

The study demonstrates that the most effective people analytics teams now engage directly with the company, functioning as internal consultants and focusing on efficiency, performance, safety, enabling HR, and solving work-related issues.

"I am not in the curiosity business," said a huge technology firm's analytics leader. We need to know the commercial value before we invest time and attention in a problem."

This is the new mantra.

To sum up, it's been a long road. People analytics has reached the C-Suite, says Josh Bersin, who has studied it for over two decades. For those still undecided, he proposes investing in these technologies in the next years. It's been a long road.