Research Methodology

Every research task begins with a precise question, decision audience, company or sector scope, and a time boundary. A broad prompt such as ‘analyze AI’ is converted into testable subquestions: demand, supply, unit economics, competitive position, policy exposure, and evidence that could disprove the thesis.

The scope is recorded before synthesis so later evidence does not silently change the question. If the available material cannot support the requested conclusion, the report narrows its claim or states that the evidence is insufficient.

Primary materials receive priority: regulatory filings, company investor-relations releases, official statistics, court or regulator documents, standards, and direct transcripts. Reputable secondary reporting can provide context, but it does not override a conflicting primary disclosure without explanation.

For each material claim we compare date, reporting period, unit, accounting basis, and whether the number is historical, management guidance, analyst consensus, or our calculation. Conflicts remain visible instead of being averaged away.

Published work distinguishes reported facts from calculated observations and forward-looking inference. A citation proves what a source said; it does not prove that management’s forecast will occur. We therefore state assumptions, alternative explanations, and the next evidence that would strengthen or weaken the view.

Before publication, the responsible reviewer checks source accessibility, arithmetic, quote context, date consistency, heading structure, and whether the conclusion exceeds the evidence. Material updates change the review date and correction note when required.