Skip to content

PLAN.md

Latest commit

 

History

History
74 lines (56 loc) · 3.61 KB

PLAN.md

File metadata and controls

74 lines (56 loc) · 3.61 KB
Raw

Research Plan — and the report spine

The goal:

  1. Real GTM research for Spoken Kitchen.
  2. A comprehensive HTML report that turns the findings into product & marketing decisions (docs/index.html).

So every phase below maps a business question → a Nimble capability. Each phase produces a finding file + the raw Nimble output, and the report synthesizes all of it into "here's what the data says, here's the decision."

The thesis (what the research must prove)

Spoken Kitchen lives in the white space between emotional legacy gifting, cookbook creation, and AI voice→text→translation. No competitor serves the immigrant-family, bilingual, elderly-voice use case end to end. Nimble's web data is how I found and sized that gap.

Phases

# Business question Nimble capability Output
1 Who are all the players in each of my 3 categories? market-finder / company-deep-dive + web search research/01-landscape/
2 What do competitors charge, and on what model (subscription / one-time / per-book)? competitor-positioning + nimble-web-expert (scrape pricing pages) research/02-pricing/
3 How do they position & market themselves? Messaging, value props, channels, SEO. competitor-positioning + seo-intel research/03-positioning/
4 Who is the ideal customer, and where do they congregate online? synthesis of 1–3 + targeted search (forums, reviews, communities) research/04-icp/
5 Where is the white space / unmet need? cross-analysis of all findings research/05-gaps/

Competitor shortlist (to confirm before running)

1. Emotional / legacy products

  • StoryWorth, Remento, Storii, Kindred (family stories), Artkive (kids' art legacy)

2. Recipe / cookbook creators (online + physical)

  • CreateMyCookbook, Heritage Cookbook, Family Cookbook Project, Paprika, Recipe Keeper, Cooked.wiki

3. AI transcription + translation

  • Otter.ai, Rev, Whisper-based tools, DeepL, Google Translate, Notta

Report structure

The synthesized deliverable is docs/index.html (self-contained, charts embedded, GitHub-Pages-ready). Built by scripts/build_report.py from the processed data. Sections:

  1. Executive summary — the thesis, proven; headline stats + decisions.
  2. The opportunity — capability heatmap + the empty-quadrant gap map.
  3. Phase findings — landscape, pricing, positioning, customer.
  4. The decision set — product + marketing decisions, each with its evidence.
  5. Risks & watch list — what could close the gap.
  6. Methodology — the reproducible Nimble commands behind every finding.

Reproducibility convention

Shell-first, with a thin Python analysis layer only where a visual lands the insight:

  • scripts/NN-*.sh — the exact nimble CLI commands per phase. Copy-pasteable; each writes raw output to data/raw/.
  • scripts/analysis/*.py — pandas, used for exactly two artifacts:
    • pricing_matrix.py → normalized competitor pricing table + chart
    • gap_map.py → the white-space visual that proves the thesis
  • scripts/build_report.py → assembles docs/index.html from the processed data.
  • Everything else stays as curated markdown findings in research/.

Status

  • Workspace scaffolded
  • Nimble CLI installed + authenticated (standard tier: no fast/--include-answer)
  • Confirm competitor shortlist (use as-is + Nimble expansion)
  • Phase 1 — landscape
  • Phase 2 — pricing
  • Phase 3 — positioning
  • Phase 4 — ICP
  • Phase 5 — gaps
  • Build HTML report (docs/index.html)