Fortune telling the HR way

While big data continues to be a buzzword, in HR circles, understanding and utilising this information to predict trends and events is the ultimate game-changer, as HRM investigates.

Swati Chawla, Global Head of HR Analytics, Strategy and Planning at Syngenta, says big data by itself “is just a set of data and has no standing without being analysed”.

“It is only information and does not offer insights.”

So, how can HR make sense of the often massive data that it is now able to unearth?

According to Chawla, this is where predictive analytics comes into the fray.

“Predictive analytics is about extracting insights from a comprehensive data set in order to determine patterns and predict future outcomes and trends,” she says.

“It also entails generating insights from the information to test alternatives and enable scenario planning.”

“Insight, not hindsight, is the essence of predictive analytics.”

Unsurprisingly, Chawla stresses that each of big data and predictive analytics “cannot exist without the other”.

Making sense of predictive analytics

Luca Zuccoli, Data Lab and Analytics Director, Asia-Pacific, Experian, says predictive analytics is the process of extracting relevant information from a myriad of existing data sources, and then using those insights to determine behavioural patterns and predict future outcomes.

“In other words, it is the application of statistical methods on data sets collected from different sources,” he explains. “These sources could include past customer transactions, supply chain processes, and surveys, and could predict and quantify events related to different subjects such as the impact of training.”

“Predictive analytics is often used by businesses to pre-empt customer decisions even before customers know them themselves, to test-drive marketing strategies before going to market, and to predict the success of campaigns.”

Eustace Fernandez, Head of HR, Southeast Asia, Experian, says that with recent advances in data-driven analytics, every aspect of a business has analytics imbued into its core functions.

“From finance to sales to customer experience to HR management, strategic planning has become more complex for business units with many choosing to adopt predictive analytics to guide their decision-making processes,” says Fernandez.

“HR professionals can leverage predictive analytics to make better decisions for talent acquisition, attrition risk management, employee sentiment, employee engagement and capacity planning.”

Chawla also concurs with Fernandez, highlighting that HR, like any other function, can play a significant role by engaging in data-driven decision making and planning.

“Examples vary from looking at past performance to predicting what types of employees are most productive and why, or profiling candidates to predict ‘time to productivity’, ‘sales achievements’ or recruitment forecasting,” she says.

Fernandez reveals that Experian is committed to helping its clients make smarter and better decisions through the use of analytics, so it was only natural for it to also adopt the use of analytics in its talent strategies.

 “In 2015, we developed a HR strategic dashboard that allowed us to align key metrics to our business and HR strategy, while tracking progress against relevant targets,” he says.

“The scorecard has helped us as a HR function to think about how we are adding value to the business priorities, such as how we are going to contribute towards improved pipeline ratios and financial performance to ensure that we do not become too inwardly focused.”

Predicting HR trends

Chawla says as HR professionals, she and her peers like to engage in live interactions with employees.

“To begin with, if these interactions are influenced by insights from data, it can promote a healthy, productive and a more prepared HR function and workforce,” she says.

“For example, by using past attrition data HR can predict employee attrition triggers and can proactively approach an employee before the employee acts negatively on that trigger.”

According to Chawla, another key to leveraging predictive analytics and realising maximum benefits from HR data lies in tying the data source to strategic business outcomes.

“HR needs to ensure that analytics and its outcomes are aligned with business objectives; whether they are around reducing costs, increasing revenue, maximising operational efficiency, staying profitable, or sustaining agility or growth,” she says.

Fernandez says predictive analytics can be used to quantify and improve the impact of specific employee programmes and initiatives, and to better understand the reasons behind employees behaviours.

“Predictive analytics can be used to track and predict relevant metrics such as performance, impact of training, and sentiment analysis for HR departments to gain a better understanding how to best support, engage and develop employees in order to reduce turnover,” explains Fernandez.

In addition, he says businesses can also leverage analytics and algorithms to monitor unemployment rates, employee turnover rates, business growth strategies and other workforce trends to predict future resource needs.

More data = better analysis?

Zuccoli stresses that effective predictive analysis is not dependant on the quantity of the data collected.

“Rather than the amount of data, it is more important to have the end goal in mind and to take on a selective and integrated approach of knowing what you want to achieve first and then gathering what is necessary,” he explains.

“While having more data is always useful, not all applications require huge data sets. More often than not, organisations are faced with the challenge of managing the digital deluge that they are not able to decipher the right insights from.”

Hence, he emphasises that understanding how to select the right data is at the heart of data analytics and will be crucial in every organisation’s data-driven processes.

Chawla says the quantity of data is only important to ensure statistical validity and to improve the ability to make predications.

“However, more than the quantity of data, the quality of the data holds importance,” she explains.

“Incorrect data can be more harmful than no data as it will lead to incorrect insights. A smaller quantity of data will reduce the statistical validity of data and thus, the comfort level of the decision maker. However, incorrect data will skew the predictions without any indications.”

Chawla says another factor which is crucial is the variety of data available.

“It can be in the instrument or source of data for the same data set or in data collected at different points of lifecycles and correlated,” she elaborates.

“This will add to the accuracy of the predictions. In the future, as the world gets more connected through technology advancement, the increase in the need of swift decision making will impart importance to the speed of data collection and interpretation as well.”

Time to move over?

Whereas big data and the amassing of information was once deemed to be the next big thing, has this notion been usurped by predictive analysis and the art of forecasting trends?

Zuccoli thinks otherwise.

“As opposed to the era of ‘big data’ being usurped, the era of big data will continue to flourish in today’s information economy and along with it, data analytics will develop into an integral and requisite function in the digital enterprises and cloud native organisations of the 21st century,” he explains.

“Predictive analytics is the next step forward beyond traditional business intelligence technologies that leverage big data. Predictive analytics approaches will equip businesses with the right tools to glean actionable intelligence based on historical data and provide forecasts on business insights for better decision-making.

“Operationalised predictive models can used to enhance and optimise targeted marketing efforts and drive campaign outcomes by projecting the campaign tactics on past, present and future customer behaviours against the wider macroeconomic issues and fast changing business landscape – taking businesses one step ahead of their competition.”

Specifically for HR, Zuccoli says the predictive modelling of employee behaviour and business scenarios will enable businesses to identify “at risk” employees and spot trends in key factors.

“Predictive modelling will enable businesses to recognise the strengths of the workforce and accurately predict vacancies and leadership needs – affording better talent resource management,” he states.

Chawla says predictive analytics usually follows descriptive analytics, which looks at past data and answers the questions of what happened and why.

“The findings from ‘descriptive analytics’ are used to form part of algorithms, rules and assumptions which lead to predictions,” she says.

“Prescriptive analytics goes a step further from predictive analytics and suggests decision options on how to take advantage of a future opportunity or mitigate a future risk and shows the implication of each decision option.

“At the current stage of HR analytics, we are still dealing more with descriptive analytics and going towards predictive and prescriptive analytics.”

Real-life examples

Fernandez says that Experian’s Global Workforce Analytics (GWA) team recently developed a predictive analytics model to help deal with a key business issue that the company faces – voluntary attrition levels.

“Losing valued employees is a significant drain for a business; from lost productivity, new hire costs to the knock-on effects on engagement and morale in the team left behind,” he explains.

“We estimate that a reduction of just one percent of Experian’s current global level of voluntary attrition would deliver cost savings of around $7 million a year – proving voluntary attrition rates to be a significant hit to our bottom line.”

Fernandez says the company leveraged predictive analytics tools to capture large volumes of historical data and applied statistical modelling techniques to identify the most predictive employee characteristics, such as their demographic profile, training days, absence, and team size.

“Based on the outputs, groups of employees were then identified as high, medium and low risk groups. This drastically reduced the resources spent on misaligned and inaccurate retention tactics and allowed our organisation to take calculated proactive action on high risk employees,” he adds.

Chawla says her firm has various tools for the prediction of functional workforce requirements and the skills required for the future.

“We are also working towards mechanisms for employee profiling, segmentation and productivity enhancement,” she explains.

Top tips

Zuccoli says the one tip he always offers to people before they adopt any predictive analytics technology is to have the end goal in mind.

“They need to understand their key priority first, before setting up an appropriate system to support them analytically,” he states.

“Often, companies start with methods or data infrastructures and end up missing the target. Businesses waste copious amounts of time sifting through innumerable amounts of data and gaining little to no insights at the end of the data analytics project.”

Zuccoli stresses that organisations need to also understand that predictive analysis is “not a magic crystal ball” that will provide an answer to any question.

“It should be viewed as part of a toolkit in conjunction with other influencing data and insights to help identify issues and take action,” he adds.

Chawla says there is no one size fits all approach.

“The framework and tools utilised should be determined by a combination of factors, such as the stage of the organisation in the analytics continuum, data availability and quality of data,” she says.

“The thumb rule I follow is that if the value of data extraction is more than the value the analytics would offer, look elsewhere.”

Key zones where predictive analytics can add value

  • Employee profiling and segmentation
  • Employee attrition and loyalty analysis
  • Forecasting of HR capacity and recruitment needs
  • Appropriate recruitment profile selection
  • Employee sentiment analysis
  • Employee fraud risk management

Source: Predictive Analytics in HR: A Primer, Tata Consultancy Services

 

Predictive Talent Analytics gaining traction in India

According to a new TJinsite survey from TimesJobs.com, 90% of organisations believe predictive analysis will be part of the future of talent hunting. Still, only seven percent of the firms surveyed currently utilise analytics for finding the ideal recruits.

In addition, close to 55% of all firms claimed they only utilised this data reactively and not to predict future plans.

Sixty five percent aim to utilise predictive talent analytics in in the next 12 months.

“The ability to move from gut-based judgments to data-driven decision-making is what makes predictive talent analytics the future of HR in India,” said Vivek Madhukar, Chief Operating Officer of TimesJobs.com.

 

Add new comment

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
CAPTCHA
This question is to ascertain that you are a legitimate user and prevent automated spam submissions.
2 + 3 =
Solve this simple math problem and enter the result. E.g. for 1+3, enter 4.

Contact info

HR SUMMIT 2017

Follow us on Twitter