Longitudinal and Cross-Sectional Research

Overview

Use this skill to turn a research target into a two-part deep research report:

1. Reconstruct how the subject emerged and evolved over time.
2. Compare the subject against the current field, competitors, or substitutes.

Prioritize causal explanation, narrative readability, and source-backed judgment over list-like summaries.

Fit Check

Use this skill when the user wants:

• a deep research report instead of a short answer
• both historical development and current competitive positioning
• analysis of a product, company, organization, technology, concept, or person
• a report that explains why key turning points happened, not just what happened

Do not stop at feature comparison tables. Explain how the subject got here and what that means now.

Workflow

1. Define the target and frame

Start by identifying:

• the exact research target
• the subject type: product, company, person, technology, concept, or organization
• the likely comparison set: direct competitors, close peers, indirect substitutes, or no clear peers

Adjust the comparison dimensions to fit the subject. For example, technical route matters more for protocols than for people, while career arc matters more for people than pricing.

2. Collect source material

Gather sources for both time depth and present-state comparison.

Prefer:

• official docs, company pages, release notes, papers, blogs, filings, and interviews
• reputable reporting and industry analysis
• community evidence for user sentiment when relevant

Always anchor claims with dates when the subject evolved over time. Mark any unverified or inferential statement clearly.

3. Write the longitudinal analysis first

Tell the story from origin to present. Cover:

• origin context: what problem, belief, or market condition gave rise to the subject
• founding or first-publication moment
• major milestones in chronological order
• turning points such as launches, pivots, funding, rewrites, leadership changes, partnerships, acquisitions, controversies, or adoption milestones
• decision logic at each major turning point when it can be inferred from evidence

Do not write a dry timeline. Build a narrative with cause and effect, constraints, and tradeoffs.

4. Choose the cross-sectional mode

Use one of these modes:

• No clear direct competitors: explain why the category is new, narrow, or hard to enter; then analyze likely substitutes and emerging challengers.
• One or two direct competitors: compare each one in depth.
• Three or more competitors: choose the three to five most representative ones and compare the rest briefly if needed.

Compare along dimensions that fit the target, such as:

• technical route or underlying method
• product shape, business model, or operating model
• target users and use cases
• strengths and obvious weaknesses
• pricing, scale, or resource base
• user sentiment and community reputation
• ecosystem position and likely trajectory

5. End with synthesis

Finish with a "Longitudinal-Cross-Sectional Synthesis" section that answers:

• What position does the subject occupy today?
• Which parts of that position come from historical path dependence?
• What opportunities and risks are most visible from the current landscape?
• What future direction is most likely, and which part is inference?

This section should add a fresh judgment, not just compress earlier sections.

Writing Standard

Write like a strong long-form reported analysis, not a management slide deck.

• Make readability the top priority.
• Use human language and concrete details.
• Ground judgment in evidence before interpretation.
• Separate confirmed facts from inference.
• Keep the longitudinal section narrative-driven.
• Keep the cross-sectional section analytical, but still readable and alive.

Output Shape

Default order:

1. Longitudinal analysis
2. Cross-sectional analysis
3. Longitudinal-Cross-Sectional Synthesis

Use tables only as support. The core analysis should stay in prose.

Reference

Read references/method.md when you need the full English prompt template, detailed dimension list, or the original reporting-style constraints.