Deep-dive market research and analytics case studies, distilled into one answer: how should a business decide?

A collection of rigorous market research case studies. Each one starts from a real business question, investigates it in depth against public data (government statistics, industry reports, primary-source filings) with every figure traced to its source and every assumption labeled, and ends in a specific recommendation. Not a dashboard, not a chart deck: real research, distilled into a decision.

I'm a Business Analytics student at Sunway University, aiming for a career in strategy and management consulting, with a longer-term goal of leading my family's business.

Reading on a schedule?

• 20 minutes: 006 for depth (8 notebooks, one country funneled from national headline down to individual districts) or 008 for the most complete single system (a 59-country trade network, a critical-minerals concentration index, a supply chain risk index, and a validated 2030 forecast, all in one project).
• The full portfolio: everything below, roughly in the order it was built.

Question: How has Malaysia's e-commerce competitive landscape evolved since 2020, and what should a business weigh when deciding where to invest?

Sourced from Department of Statistics Malaysia, Bank Negara Malaysia, and platform market intelligence, with every figure traced to its source and labeled public, derived, or estimated.

Python SQL; Complete

Question: How did the global chip shortage affect Malaysia's electronics manufacturing and export performance, and what does it reveal about supply chain resilience?

Python SQL; Complete

Question: Is Malaysia's electricity grid decarbonizing fast enough to meet its own climate commitments, and what does that pace mean for a business deciding whether to wait on the grid vs. invest in its own renewable capacity to hit its emissions targets?

Python; Complete

Question: How much economic value in recoverable metals sits inside Malaysia's e-waste stream, how much is the formal recycling system actually capturing, and does that gap represent a real business opportunity or something else?

Python; Complete

Question: Does the "China+1" narrative (that capital is diversifying away from China toward alternative manufacturing destinations) actually show up in the data, at the global level, the ASEAN regional level, and specifically in Malaysia?

Python; Complete

Question: Does Malaysia have a graduate unemployment crisis, or an underemployment one, and does that answer hold at the national level, across every state, and down to individual districts?

Eight notebooks, the deepest funnel in this portfolio: national headline → age cohort → job quality → every state → structural explanation → every district → wage penalty → field-of-study mismatch. Includes a real data-quality bug found and corrected in a government dataset, and two notebooks where the first-draft finding was rewritten after the chart contradicted it.

Python; Complete

Question: How does consumer spending behaviour differ across global regions, and what economic, demographic, and technological factors help explain those differences?

The first case study in this portfolio to go fully global: a 182-country, 11-year panel built from World Bank, OECD, and World Bank Governance Indicators data. Data-driven K-Means market segmentation, a transparent, sensitivity-tested 5-pillar market attractiveness index, and a real solved-in-the-open technical story (the OECD SDMX category-spending API was initially descoped, then genuinely fixed by discovering the correct query structure rather than guessing).

Python; Complete

Question: How has the global supply-chain system changed over the past decade, where are the major dependencies and vulnerabilities right now, what could happen over the next 5-10 years under different scenarios, and what does all of this mean for Malaysia specifically?

The most complete system in this portfolio: a genuinely complete 59-country bilateral trade network (UN Comtrade, backfilled with OECD data for the four countries Comtrade's free tier excludes entirely) with betweenness/eigenvector/PageRank centrality analysis, a 13-material critical-minerals concentration index (HHI/CR3/CR5) built from real USGS production data, a transparent, sensitivity-tested Supply Chain Risk Index, a Malaysia export forecast to 2030 validated by a genuine 5-year rolling backtest, a 10,000-path Monte Carlo simulation on nickel prices, and a dedicated Malaysia 2030+ Vulnerability Index and evidence-graded Opportunity Radar.

Python; Complete

Eight case studies, chosen for range across industries and depth over volume. New case studies are added one at a time, each only once it's genuinely done, not before.

Business problem → objectives → data acquisition → cleaning → analysis → visualization → insight → recommendation. Every notebook opens with the question and the answer, then shows the reasoning between them, including what the data didn't support.

Every repo includes a requirements.txt and a "Reproducing This Analysis" section: clone it, pip install -r requirements.txt, and the notebooks run against the data already checked into data/processed/.

Python · SQL · Git

LinkedIn