Case Study

How an NUS Business School Alumna Turned People Data into Organisational Change at Murata Electronics Singapore

In 2025, Elise Lim Ying Qi — NUS Business School alumna and HR executive at Murata Electronics Singapore — designed a 15-item pulse survey, built and tested multiple regression models to identify the strongest drivers of employee engagement, and translated those statistical findings into HR policies employees helped shape. The result was a working analytics pipeline that moved people management from benchmarks to evidence.

• Data-Driven Decision Making
• HR Analytics
• Mixed-Method Research
• People Analytics
• Regression Modelling
• Survey Design & Analysis
• Talent Management
• Workforce Analytics

The data challenge — and what it took to solve it

The challenge Elise inherited — and why it mattered

When the data exists but the answers don’t, the survey process itself may be the problem.

Murata Electronics Singapore had been running employee surveys since 2019 — professionally executed, externally managed, and producing results across 12 dimensions of employee experience. The results were useful for benchmarking, but could not answer the question leadership most needed to resolve: which specific factors were driving engagement gaps, and what, precisely, to do about them.

The analytical constraint was structural. Murata Electronics Singapore received aggregated outputs from its global survey provider — sufficient for trend comparison, but not for the deeper modelling the HR team wanted to attempt. Without individual-level response data, segmentation by division, tenure, or role was not possible, and regression analysis could not be applied. The team needed a different data foundation.

By 2024, employee engagement stood at 70% and enablement at 77% — both short of the 2027 targets of 75% and 78%. Two themes had consistently surfaced from earlier surveys: employees wanted clearer career pathways and better advancement opportunities. But knowing what employees wanted broadly wasn’t the same as knowing which specific workplace factors were preventing it, or how much each one mattered.

Our goal was to move from knowing that engagement gaps existed to understanding exactly which workplace factors drove them — and by how much. That required individual-level data, a well-designed survey, and a regression model we could trust.

— Elise Lim Ying Qi, NUS Business School Alumna & HR Executive, Murata Electronics Singapore

Murata Electronics Singapore is Murata’s first overseas manufacturing affiliate, established in Singapore in 1972, and plays a central role in the company’s global operations. Its aspiration to operate as a “small giant” — a lean, high-performing unit delivering outsized value — depended on retaining capable, motivated people and understanding what kept them engaged.

The question Murata Electronics Singapore needed to answer was not whether engagement could be improved. It was which levers, applied in which order, would produce the greatest and most durable improvement. That required evidence that global surveys, by design, could not provide.

The better the answer

What Elise built: a survey, a model, and a plan

A locally designed pulse survey and regression modelling revealed precise, actionable levers for engagement — for the first time.

In 2025, Murata Electronics Singapore’s HR team proposed running its own internal pulse survey — making the case that global benchmarks and local explanations serve fundamentally different analytical purposes. The proposal centred on a clear argument: only individual-level response data would enable the regression modelling needed to identify which specific factors mattered most for engagement, and by how much. Leadership supported the initiative once the analytical rationale was clear.

The HR team made its case methodically: a global survey produces benchmarks; a local survey with individual-level data produces explanations. Without employee-level data, Murata Electronics Singapore could not identify which divisions were struggling, which groups were most at risk, or which factors mattered most. Over time, leadership was persuaded — not by assertion, but by the clarity and precision of the analytical ambition behind the proposal.

The pulse survey comprised 15 questions spanning career development, workplace resources, feedback quality, cross-divisional collaboration, compensation fairness, and overall job satisfaction. Crucially, every response was linked to an anonymised employee ID — giving Murata Electronics Singapore granular, individually attributable data for the first time.

The HR team developed and tested multiple regression models, systematically examining different groupings of survey items to identify which combination best explained employee engagement. The selected model used two sequential regressions: one identifying what drove job satisfaction, and a second examining what, in turn, drove engagement itself.

The findings reframed the HR team’s priorities. Removing barriers that prevented employees from doing their work well emerged as the strongest driver of engagement — more powerful, by a significant margin, than compensation. Development opportunities were the most influential factor shaping how satisfied employees felt in their jobs. These were not intuitions; they were statistically significant relationships, drawn directly from Murata Electronics Singapore’s own workforce, in its own context.

To validate the model, the team tested its predictions against the actual survey responses of a randomly selected employee — and found that the predicted engagement level matched the recorded answer. The model worked.

For years, our efforts to improve engagement scores were largely driven by gut feeling. Now, for the first time, we had a model that could tell us not just where the problems were, but which levers would actually move the needle.

— Elise Lim Ying Qi, NUS Business School Alumna & HR Executive, Murata Electronics Singapore

The better the world works

From statistical insight to real organisational change

Focus groups, deliberative polling, and employee voice transformed data into HR policies that people actually believed in.

Having strong regression results was a beginning, not a conclusion. Coefficients don’t implement themselves. Statistical significance doesn’t explain why barriers exist, what form they take in practice, or which solutions employees would trust. That required listening — structured, purposeful, and designed to surface what a survey could never fully capture.

Murata Electronics Singapore launched a series of focus group discussions using the regression model as a guide. Rather than a broad, self-selected sample, the HR team made deliberate choices about who to include. Three “vulnerable groups” were prioritised: fresh graduates, employees in selected manufacturing functions, and corporate functions. Within each group, the team ensured representation across the favourability spectrum — both satisfied employees and those who had signalled genuine dissatisfaction.

The focus groups surfaced specific, actionable themes that aligned precisely with what the regression had flagged. Employees highlighted the importance of structured mentoring, greater clarity on the link between performance and compensation, and stronger cross-team communication — each of which mapped directly onto the model’s highest-coefficient variables.

Following the focus groups, Murata Electronics Singapore introduced deliberative polling — a four-step process in which employees were presented with potential HR policy interventions, asked to vote, invited into a facilitated discussion, and then asked to vote again. The deliberative element filtered reactive first responses and allowed more considered preferences to emerge.

The buddy system stood out as the clear preference among newer employees — particularly those with lower engagement scores who wanted informal support during their early months. Employees also expressed strong support for follow-up conversations after performance reviews: not more evaluation, but more dialogue.

These employee-chosen policies moved into pilot implementation across several divisions. Manager training modules were updated to embed best practices in timely, clear feedback. The data had not made these decisions — people had. But the data had made those decisions possible by showing, with precision, where to focus.

Elise Lim Ying Qi graduated from NUS Business School with a Bachelor of Business Administration, where she was awarded the Fung Scholarship and recognised as an Asian Elite Business Scholar. During her undergraduate years, she served as HR Director and Treasurer of the NUS Investment Society, completed a student exchange at the University of British Columbia in Vancouver, and interned across GSK and CapitaLand. A member of the NUS Wind Symphony and NUS Japanese Studies Society, she developed an early affinity for Japan — a fitting foundation for where her career would take her.

At Murata Electronics Singapore, Elise inherited a pulse survey and regression analysis project that would become the foundation of this case study — her first major analytical undertaking as a newly hired HR executive. Today, she works as a Corporate Planning Analyst at Murata’s global headquarters in Kyoto, Japan, having turned the evidence-based thinking she practised in Singapore into a springboard for a broader international career.

Connect with Elise on LinkedIn ↗

Elise Lim (second from right) with her team, Dr Tan Hong Ming and his team from NUS Business School, and her analytics instructor.