Last updated: 13 March 2023

Published on: 14 February 2018


“Why do people leave companies?” is a complicated question. But according to OCBC Bank’s head of group human resources, the answer lies in the data.

Data analytics in the office
Who's leaving the company soon? Data analytics might have the answer for the Human Resources department.

By Nelissa Hernandez

Jason Ho, OCBC Bank's head of group human
Jason Ho, OCBC Bank's head of group human resources.


“If you torture data long enough, it will confess” goes a popular saying from economist Ronald Coase. One way to look at it: Data can yield tons of information if you dig deep enough.

Data can even predict who will leave a company in six months, if OCBC Bank’s foray into human resources (HR) analytics is anything to go by.

“The challenge for HR is that some people view it as fluffy because HR decisions or decisions relating to people matters used to be based on experiences and relationships. Many do not realise that HR is actually the most independent source in the company, and we present the most objective view of what's happening in our business,” said Jason Ho, OCBC Bank’s head of group human resources.

Tapping on the expertise of the bank's customer analytics department, OCBC began its HR analytics in 2015. Today, there are three employees who are dedicated to people analytics within their HR team.

IMPact spoke to Jason to get his insights on what can be learned from capturing HR data.


Tell us more about OCBC’s use of HR analytics. How did it come about?  

Data analytics and technology play a huge role not only in our bank’s operations but also in our HR practices. A digital strategy is not just about technology, but also people – how they adapt and use technology, and adopt a mindset to embrace changes in the organisation.

We believe that technology is one of the key enablers of HR, and data analytics will drive digital transformation. At the end of the day, it’s about how we pull the data together and use the right set of data to provide insights, make better decisions, and improve productivity.

With data analytics, we can enhance and validate on-the-ground information so that it’s no longer purely qualitative. When information is backed by data, it becomes more objective and reason-driven, complementing the relationship-based information we have gathered to enable better decision-making.

How does your HR data analytics work?

It comprises a data infrastructure, which consolidates employee data into one single information system, and self-service management dashboards that provide an easy and quick overview of business data and the ability to deep dive into issues.

There are many potential applications of people analytics, but for now, we are using analytics for talent management and retention, recruitment, and learning and development.

In the area of talent management, we are using analytics to predict attrition risk. Within recruitment, data analytics are used to provide insights on the hiring process. In terms of learning and development, we use it to drive and recommend learning initiatives to employees based on their job function, rank and recently taken courses.

Working overtime
Watching the clock or working overtime? A predictive model lets the bank predict which employees are at risk of leaving.

Speaking of attrition risk, what are some of the factors used to help OCBC predict which staff might leave?

We started by developing a model using the profile of employees who were with us as at end 2014 and left in 2015 as data points. We distilled factors that were common among these employees, and used these factors to predict employees who were at risk of leaving in 2016.

Some of the factors used in our model are the employee’s tenure of service, participation in the employee share scheme, reporting manager, and number of training days, among others.

Come 2016, we tracked employees’ attrition and validated them against the prediction done by the model.

With these findings, we continued to refine the model, added new factors and recalibrated their weightage so that attrition risk can be predicted with a high accuracy consistently across the bank, regardless of which division the employee is from.

Resignation letter
Insights from data analytics can help companies to pre-empt talented staff before they write that resignation letter.

Today, our model is able to predict which employees are at risk of leaving within the next six months, with an accuracy of 75%.

How has data analytics benefitted OCBC in terms of recruitment, streamlining your HR processes and/or boosting productivity?

We use data analytics to provide insights on executive resourcing. We have a dynamic dashboard that shows the readiness of a candidate as he or she moves along the pipeline. Our resourcing team who supports business divisions in hiring can log in to the system to know the progress, instead of relying on the traditional spreadsheets and conversations.

The system can help us identify if hiring managers are taking too long for the status of candidates' applications, and help speed up hiring processes and improve candidate experience.

It also gives us an oversight of the job openings and positions that each member of the resourcing team is working on. By doing so, it enables us to know if a member requires some support with a particular position or candidate, and eliminate any duplication of work. This helps improve productivity and create better collaboration within the HR team.

How has data analytics helped you retain top talent?

Our attrition model enables us to better pre-empt departures among our talents, so that early preventive actions can be taken to try to retain them. Once such a risk is identified, our HR business partners will work with the division heads, and a conversation takes place on concerns and how the business line manager or HR manager can engage the employee at risk.