How a new AI-Recruiter is impacting business at DBS
Speaking at the Mercer 2019 Singapore HR Conference, Global Pioneer and Expert in Strategy Implementation, Robin Speculand shared the DBS digital transformation story and Susan Cheong, Group Managing Director for Talent Acquisition at DBS Bank followed by sharing a case study on innovation in talent acquisition.
Speculand is writing a book on DBS and has spent many hours observing how the bank’s digital platform operates and assists the business from an HR standpoint. On a more specific level, Susan Cheong outlined how the bank has used Jim (short for Jobs Intelligence Maestro), a custom-built robotic AI system to streamline the recruitment process and free up HR professionals within the organisation. Here are the insights of both the sessions.
DBS’ Innovative Steps Towards Digitization
Speculand began by outlining some of the key changes DBS have made over the past few years:
Predicting attrition: Firstly, models have been implemented to predict when employees are going to leave. This technology identifies the predictive behaviors such as “turning up late for meetings, using up holiday time and going home early.” This has allowed DBS to develop an algorithm and in turn use data analytics to know when staff are going to leave and preempt the ones that want to keep the job. “At DBS, for every 1% of improvement in reducing turnover, it saved up to S$5 million,” Speculand said.
- Scaling skilling: DBS have eliminated management approval and sign-off on courses that cost less that S$500. Speculand added, “Now, as an employee, I can sign up for a course. The only condition is when I come back, I have to share with my peers what I’ve learned.” Therefore, learning is spread throughout the organisation.
- Onboarding Experience: Onboarding has been adopted as a customer journey. “When you join DBS, you’re live on their system before you even join the bank,” Speculand said. Therefore, before you even begin your first day, you can see what is happening within the company.
- Reverse Mentoring: One of Speculand’s favorite initiatives was the bank’s introduction of reverse mentoring. “At DBS, senior management will bring in staff members from each department to teach them what they do,” Speculand pointed out. This has proved to be a powerful tool at every level of the organisation.
- Making meetings matter: DBS has also implemented the MOJO productivity timer app. Standing for ‘Meeting Organizers and Joyful Observers’, the app aims to reduce time-wasting and make sure each meeting has a clear agenda by imposing timelines, feedback and structure.
Jim, Southeast Asia’s First Digital Recruiter
Following on from Speculand’s presentation, Susan Cheong homed-in on how Jim - DBS’ virtual bank recruiter - has helped HR teams hire more efficiently. Jim was introduced in 2018 and uses psychometric profiling, CV screening and collation of applicant responses to streamline the thousands of applicants DBS receives for most positions advertised.
“We have thousands of CVs coming in - how do you deal with that? How do you use digital technologies to deal with that?” Cheong asked. “We needed to leverage technologies to help ourselves.”
Firstly, the team had to validate the idea and prove its use to the stakeholders as quickly as possible. “The team came together and it took 2-3 months to validate the concept.”
In order to do this, the DBS Talent Acquisition Team worked with Impress AI - a startup made up of students. The DBS Innovation team established the Startup Exchange Program to facilitate the partnership between the HR department and third-party vendors and other project teams. This filled the gaps and provided advice, mentorship and feedback. “If your organisation is on a transformation journey, a structure like this could help you,” Cheong said.
How was Jim used?
The first role Jim was used to hire was senior wealth planning manager. This allowed the team to test the scalability and the data points Jim was able to collect. “We identified the key stakeholders in this process and considered them while assembling the focus groups,” Cheong described. Over 2-3 months of experimentation, the focus groups came up with different versions and iterations of how Jim could be used. Prototypes were tested. Groups of people were tested and asked what their priorities were for Jim.
At the end of this period, it was found people wanted Jim to be an end-to-end mechanism from a recruitment standpoint, starting with the time the candidate actually applied for the role to making assessments on applicants.
Speaking about the potential pitfalls, Cheong outlined a few. “Chatbots can cause very embarrassing situations if you’re not careful,” Cheong warned. “They speak in the wrong way and give the wrong answers. You really need to test that out.”
Some people have also asked how useful something like AI can be in such a people-focused area such as HR. While Cheong recognizes this point, she also highlighted how even the best-recruiters, herself included, are susceptible to “unconscious biases in the area of selection.” Jim is different. Jim doesn’t think that much. “Jim just sees the scoring from the questions. He doesn’t oversee or over-analyze.”
What’s the outcome from Jim?
“Before Jim was launched, it took us an average of 32 minutes to screen through a candidate,” Cheong said. “With Jim, we could screen candidates within 8 minutes.” This shows natural efficiency progress from an organizational standpoint.
From a staff standpoint, Jim allows workers more time to do value-added activities. “There’s very little joy,” Cheong said, in looking through thousands of potential candidates. Jim saved 24 minutes of time per candidate which translates to 75% of time per candidate. Over a two-year period, this equates to 2.7 months of time saved.
In 2017, before Jim came into use, one in every seven selects could be hired. Post-Jim, one in every three selects could be hired.
“What’s truly rewarding though, is the Talent Acquisition Team can now spend time as talent advisors and more quality time spent with hiring managers to understand the business needs without feeling they need to fill vacancies within the same amount of time. That’s the beauty of it,” Cheong said.
“Jim is not a process that is perfect overnight,” Cheong added, but with the help of stakeholder feedback and the hard work of the DBS Talent Acquisition Team, they hope to keep iterating and rewiring Jim to the benefit of the stakeholders.
Cheong ended her presentation with a reassurance that human interaction is still a high priority for the recruitment at DBS, and that Jim will not replace the “face to face interview” step of the hiring process. “That’s something that cannot be replaced,” she said.
Jim is a great example of rethinking a process putting the consumer at the center and bringing technology to augment the current teams. Speculand concluded by stating that “it’s not about digitalizing your organisation, it’s about having a strategy in a digital world.” As DBS Bank’s many strategies and immense successes illustrate, it’s more important than ever to have a solid, clear direction for your organization’s transformation journey.