10 years ago in most companies the functions of HR analytics were reduced to providing operational reports. Specialists were mainly engaged in measuring the costs of hiring and training staff, contributing the growth of business value more or less.
But since 2011, the market has been shifting: about 5 % of the market's leading companies, including Google, Sysco, Best Buy and Starbucks1 have invested in predictive or behavioral People Analytics technologies.
So, by 2020 the Human Resources market has formed a steady trend of implementing Predictive Analytics for making business decisions based on objective measurement and big data.
Who should we hire? How much should we pay to employees? Which employees are prone to burnout? How can we prevent staff fraud or unwanted outflow?
How does estimating and predicting of the Human Factor in business allow you to increase customer satisfaction and retention?
Find the answers out in a new Yva.ai article.
Predictive Analytics in HR: Origins
As a statistical method, Predictive Analytics proved its workable in various business fields. The analysis of Customer Data is an inherent predictive tool in marketing. And, perhaps, there’s no serious bank loan without scoring (assessing the solvency of a potential borrower). It’s Predictive Analytics too.
The world's leading researcher in the field of Talent Management Josh Bersin spoke about the prospects of Maths approach in People Analytics more than 10 years ago.
According to Bersin, Predictive Analytics opens up huge business opportunities. The world gross wages is over $4 trillion USD per year and the idea of managing this market is extremely tempting!
At the same time, the analyst noted the successful implementation of Predictive People Analytics practice needs Big Data Experience. Leading companies delegate this to statisticians, mathematicians, engineers & other “maths brain” specialists. This is not a coincidence.
"Predictive Analytics and Talent Management is a spacious Data Lake. Here you act in the same way as a venture investor, you analyze data, you predict events with various degrees, and after all you make a decision. This allows you to prevent negative scenarios before they happen, such as the dismissal of a valuable employee. And you definitely can use the opportunities opened up because you see it ahead of time. For example, you give new powers to Hi Po employees you’ve never recognized before. Or you help your employees to deal with an emergency of burnout. This definitely matters”. Daniel Hoppe, Yva.ai Expert in Remote Team Analytics Software
Why Did Predictive Analytics Come to HR?
Economic forecasts are primarily important for business. The forecasts for productivity, sales, growth of potential customers and LTV of existing ones are on one side and potential costs, risks and losses are on another side.
Salary fund is one of the extremely significant items in expenditure of a sustainable operating business. But the impact of the Human Factor remained unpredictable. Until recently.
In other words, companies invested heavily in their employees but cannot manage performance truly based on data.
But companies are interested:
- how exactly their main asset works,
- who the best performers in their teams are,
- how to improve efficiency without hiring new employees,
- how to motivate employees to succeed,
- which investments in personnel actually affect productivity and which do not.
At this point there was a business need for Predictive People Analytics. And according to the law of the market, in response to demand supply was born.
3 Levels of Predictive Analytics
It seems the more sources the information companies have to analyze, the more accurate forecasts it gets. It may seem businesses with an extensive database have a leg-up by default.
But in fact, not only the amount of data matters. The processing quality matters even more. A specialist in Talent Management, Dr. James Cecil2 from University of Wisconsin, Madison identified 3 levels of Predictive Analytics based on the depth (or height) of dealing with data.
The first stage is Descriptive Analytics. At this level the specialist uses the staff number and structure data, lateness, absenteeism and other indicators to describe what is happening right now.
The predictive value of such analysis tends to zero. Nevertheless, C-level managers need the information to be aware of what is happening to employees in real time.
The second stage is Correlation Analytics. At this level the specialist already builds relationships between variables, for e.g., between employee's morale and staff turnover.
Of course, correlations barely mean a linear causal relationship. But we suggest the possible influence of one factor to another and give grounds for certain business hypotheses. That increases the predictive value of this level.
The third level is Predictive Analytics. At the top of the model by J. Cecil is the most complex and impacting method of forecasting based on data. That allows us to establish causal relationships between specific risk factors and their influence on company state.
At this level Predictive Analytics can answer:
- whether training of salespeople can increase sales and customer satisfaction,
- why an effective employee has become less active and whether there are any risks of voluntary dismissal,
- how these or other departments interact with each other, if they have any problems and what problems they have,
- how the company's efficiency has changed after switching to remote work and many others.
How Does Predictive Analytics Work for Global Market Leaders?
Thanks to advanced Predictive Analytics some companies can accurately determine the price of increasing employee engagement. So, at Best Buy (125,000 staff) the engagement increase by only 0.1 % led to operating income growth more than $ USD 100,000.
The multinational corporation Sysco (51,000 staff) involved in marketing and distributing food products, smallwares, kitchen equipment and tabletop items began an efficiency analysis with 3 gross indicators:
- corporate climate and Employee Satisfaction,
- employee performance,
- staff retention.
The company found financial results are higher in divisions with highly satisfied employees, and the cost of employee retention and customer loyalty is really lower.
Over the past 6 years Sysco has increased the retention rate of delivery employees. The retention in Customer Success division grew from 65 % to 85 %. Retaining key employees in an economically significant division of the company Sysco saved nearly $50,000,000 in hiring and onboarding of new staff.
One of the most well managed companies in the history of management is Google, of course. And this giant has been building a talent value model for at least the last 12 years. The task of the model is to answer the question "Why do employees prefer to stay in our company?" and scale the positive impact to the entire team.
The company's hypothesis is these performers are mismanaged or stayed out of their true vocation. Detailed People Analytics has confirmed this idea.
E.g., one of Google's analytical findings performed a curious fact. Ability to show initiative demonstrated by an applicant is a more significant predictor of high performance at work than prestigious education. But this is at odds with the HR trend of many companies to trust the diploma and give preference to graduates from titled universities.
Technological Solutions in the Field of Predictive People Analytics
An obvious problem with big data in HR is we need to process all the information in real time and keep it safe. We need to take into account a lot of human variables including intelligence, experience, level of competence, stage of burnout – a number of personal parameters.
A specially trained Yva.ai neural network comes to the aid of HR-managers and business owners. It processes metadata effectively and complies with international requirements for information security of users and company data.
Do you want to tame People Analytics for your business increase? Learn more about Yva.ai from first-hand experience. Watch the webinar with CEO Yva.ai Inc. David Yang, Ph.D. about 5 Ways to Increase Remote Wellbeing in 2021. Pick it up!
- Competing on Talent Analytics // Harvard Business Review. Thomas H. Davenport, Jeanne Harris, Jeremy Shapiro. 2010.
- Applying Advanced Analytics to HR Management Decisions // James C. Cecil. Pearson FT Press. 2013.