How OCBC Used Generative AI Role-Play to Double Advisor Appointments in 90 Days

Singapore's OCBC Bank has delivered a compelling proof point for AI-powered sales training: a six-month generative AI programme that doubled advisor client appointments and grew revenue 50% in just its first three months.

The Problem with Traditional Sales Coaching

OCBC's 900 wealth advisors used to wait up to three weeks for a one-on-one training session with a supervisor. The process was slow, inconsistent across managers, and difficult to scale. So the bank built something different.

What the Programme Does

Using large language models trained on the bank's anonymised proprietary customer data, the platform simulates realistic client conversations — mimicking prospects looking to build investment portfolios, identify their risk profile, or adjust strategies amid market volatility.

After each session, supervisors receive a GenAI gap-analysis report detailing the advisor's competency levels and areas needing improvement, enabling targeted in-person coaching rather than broad, generic training.

The Results After 3 Months

• 2x more weekly client appointments for trained advisors vs. untrained peers
• 50% uplift in revenue compared with the prior three-month period
• Advisors can practice scenarios independently, as many times as needed, at their own pace

"By marrying AI precision with the human touch, we ensure our advisors are not just well-trained, but future-ready," said Sunny Quek, OCBC's head of global consumer financial services.

Key Takeaway

This model is replicable. Any organization with structured customer data can build AI training environments that compress skill-development timelines dramatically. OCBC plans to expand the programme to Malaysia and Hong Kong.

🔗 Read the full article on Computer Weekly