The Answer Was Always in the Evidence

Holistic financial planning landed because it was built on what the segment felt.

OCBC came in with a brief about financial planning for young parents. The data they held was informative about behaviour: what the segment was buying, when, in what volume. It said nothing about why the decisions felt so difficult.

The investigation turned to forums where Singapore parents have conversations they do not have in focus groups or surveys. The finding was not a product gap or a messaging problem. It was a tension: the conflict between providing every advantage for one’s children and securing one’s own financial future, present in every authentic conversation and absent from every formal research channel.

Holistic financial planning tools were launched around that tension. Research costs fell by more than 50 percent. The product landed because it was built on what the segment felt, not what they said when someone was recording the responses.

The insight was in public data the whole time.

Mastercard was watching the same domestic payment competitors every card company was watching. Tourist behaviour data from the top five arrival countries told a different story. Singapore’s MRT was consistently rated a top attraction in the actual conversations tourists were having online. No one in financial services was treating it as a commercial opportunity. The visible problem was absorbing all available attention.

An exclusive five-year first-mover partnership with SMRT followed. Cross-border transaction volume increased materially. The deal was covered in Bloomberg. The insight was in public data throughout.

The evidence pointed somewhere the brief had not anticipated.

Nestlé arrived with a health-positioning strategy for Green Coffee. The health angle was logical, internally consistent with the brief, and aligned with the category’s prevailing direction. The evidence pointed somewhere else entirely.

What Indonesian consumers were saying in their own language, in their own online spaces, pointed to beauty benefits. The angle was not in the brief. Nestlé moved from trailing the competition to leading the category, starting with social media posts and scaling to the brand’s main campaign.

When the market resets, don’t reduce price. Re-read the market.

Hilton was facing declining bookings in the post-pandemic domestic travel market without wanting to resort to price reductions, a strategy that would erode both revenue and the premium positioning the brand had spent decades building.

The investigation used publicly available data to map what post-pandemic travellers were actually choosing, not what they said they preferred. Statistical methods identified which amenities correlated most strongly with booking decisions for specific segments.

Bookings increased by 17 percent. No advertising spend. No price reductions. Premium positioning held throughout.

Measurement strategy must precede analysis.

Frasers Northpoint faced uneven customer traffic across tenants, with no systematic way of understanding which tenants were driving the highest-value footfall, or why.

The investigation built three layers of data collection (digital voucher distributions, geolocation data, redemption rates) and used them to identify the tenant locations and timing patterns that correlated with high-value customer behaviour. The result was a personalised recommendation system with dynamic vouchers adapted by location, time, and store context.

Footfall increased by 150 percent. The operating model shifted from intuition-based tenancy management to evidence-based optimisation.

From data-generating to data-driven, by building the missing measurement layer.

Great Eastern had built two functional mobile applications, one for customers, one for agents, without the measurement infrastructure to understand how either was actually performing. The applications generated data. They did not yet generate insight.

The investigation built the analytics and identity-resolution layer the applications were missing, integrated the attribution frameworks that made user behaviour legible to the product teams, and established the measurement baseline from which improvement could be tracked.

The applications moved from transactional tools to strategic assets. UI and UX evolved based on actual behaviour. Loyalty and retention data became actionable.

Talent attraction is more precisely an intelligence problem than a brand problem.

XL Axiata needed to attract senior digital talent (ML engineers, cloud architects, cybersecurity specialists) without evidence-based understanding of what motivated that specific cohort in that specific market at that specific moment.

The investigation analysed 19,948 talent and job profiles across 773 companies, identified 797 unique skills, and mapped supply and demand gaps across the market. Twelve career-motivating factors were tested through multivariate analysis.

LinkedIn following increased by more than 230 percent. XL Axiata became the destination employer for digital-transformation talent in Indonesia.

Lead allocation is as commercially significant as the sales process itself.

ASBL’s call centres handled between 10,000 and 30,000 inbound calls daily from prospective luxury apartment buyers. There was no systematic way of distinguishing high-conversion leads from low-conversion leads at the point of intake.

The investigation analysed thousands of historical calls using speech-to-text AI, identified trigger words and conversation flow patterns most strongly correlated with conversion, and built a real-time lead scoring system from the findings.

Conversion speed improved by 20 to 30 percent with the same call volume. Sales scripts were refined on evidence. Agent training became measurable and targeted.

Customers already say what they need, in spaces where no surveyor is listening.

HSBC needed to understand the financial concerns of mass-affluent Premier customers (parents aged 25 to 40) in ways traditional research methods could not access. The question was not what this segment said they needed. It was what they actually needed, as revealed in their own conversations in spaces where no surveyor was listening.

The investigation built a data pipeline to mine a full year of online forum discussions, identifying 266 distinct customer concerns. These were validated with 726 Premier customers over seven days via Facebook, tracking 3,480 interactions.

HSBC launched education consultancy services for Premier customers, a product that emerged entirely from the evidence, and that the research methods standard in banking had not surfaced.

Public digital signals are more honest than controlled research environments.

Pfizer (now Haleon) needed to understand shifting consumer attitudes toward wellness across two culturally distinct Asian markets at once. Traditional surveys could not capture authentic behaviour-level sentiment across such different cultural contexts.

The investigation used public digital signals, search data from Google in Taiwan and Baidu in China, to map actual consumer behaviour rather than stated preference. Chinese consumers were moving toward functional foods; Taiwanese consumers were engaging with lifestyle diseases and natural detox. The two markets required entirely different strategic responses.

Innovation pipelines were re-prioritised. Messaging moved to culturally resonant value propositions in each market.

Repeat engagement requires working upstream of behaviour.

Singapore Zoo needed to shift from a one-time visit model to a repeat-engagement model with local families. The question was not how to make the zoo more appealing in the abstract. It was what specific job a visit could perform for a Singapore family, and whether the zoo was currently performing that job.

The investigation identified parenting priorities and the points of family friction a zoo visit could resolve, using data analytics rather than survey methods.

KidsWorld was launched, designed around those specific family needs. Footfall increased. Time spent per visit increased.