Video summary

딱 20초만 보세요. AI로 100억 버는 6개 질문, 모두 공개합니다.(완벽한 원시인)

Main summary

Key takeaways

Business

Business goal & positioning (high level)

  • The speaker claims an AI-assisted playbook to sell “10 billion won within 1 year” for a book/product (referenced as Perfect Genshin Impact / Perfect Caveman / other variants).
  • The core emphasis is turning strategy into a repeatable process by iteratively prompting an AI (Claude/GPT/Jaminyai) for:
    • formulas,
    • marketing options,
    • prioritization,
    • schedules.

Core framework: “6-step AI revenue machine” (playbook)

The video breaks go-to-market planning into six steps, each mapped to specific AI prompts.

1) Formalization (first principles → decomposing into formulas)

  • The method is to build “laws” / formulas for how sales happen, then use AI to generate and select candidate formulas.
  • A key mistake called out:
    • Don’t ask AI to “just come up with a full strategy for 10B” without decomposition.

Splitting method (first-principles thinking)

  • Break 10B into smaller targets (e.g., revenue → units → what must happen in the market to sell those units).

Illustrative math & market-awareness assumption

  • Example decomposition discussed:
    • 10B won target
    • → 100,000 initial “explosions”/units (speaker’s formula)
    • → requires content/marketing that creates viral “explosions”
  • The approach includes a population-awareness rationale:
    • Korea adult population cited as 43 million, implying “more than half” needs awareness to reach the unit scale (presented as part of the calculation).

2) Clean up / enumerate marketing channels (taxonomy building)

  • The AI is asked to summarize and list all marketing types, both:
    • online and offline
    • plus platform-specific formats (e.g., YouTube Shorts, Reels, Threads, Card News, etc.)
  • Speaker stresses sequencing:
    • list everything first, then later combine with the formula via further decomposition.

3) Generate at least 20 marketing ideas (derived from the taxonomy)

  • Ask the AI for 20+ ideas “based on”:
    • the structured marketing taxonomy
    • the decomposition targets from Step 1.
  • Implementation detail:
    • Better results by batching (e.g., request 20 ideas in one batch, rather than asking for 50 at once).
  • Additional tactic:
    • Request ideas that can be adapted into campaigns and partnerships.

4) Priority ranking (scoring + star rating)

  • The AI rates the best ideas among the generated set.
  • Output includes:
    • top-ranked picks
    • star ratings
    • priority ordering.

5) 2-week execution plan (day-by-day)

  • The AI produces a detailed plan for the next two weeks, scheduled per day.
  • Guidance includes:
    • specify morning/afternoon tasks
    • “print it out / follow exactly.”

6) 6-month strategy + monthly plan

  • The AI is asked for:
    • a 6-month plan
    • monthly priorities
  • The macro/micro structure:
    • micro: next 2 weeks
    • macro: month-by-month up to 6 months

Key metrics / targets explicitly mentioned

  • Revenue target: 10 billion won in 1 year
  • Precedent referenced: an AI course sold for about 3 billion won (used as justification for the method).

Unit/scale targets (as stated in the video)

  • Sell 100,000 initial “explosions” (speaker’s “to sell 10 billion” formula)
  • Mentions include (wording inconsistent due to subtitle issues):
    • “500,000 won worth” and “500,000 units” in an example decomposition path
  • The playbook repeats decomposition units across explanations.

Market size anchor (awareness math)

  • Korea adult population: 43 million

Staged operational cadence

  • Concept of reaching outcomes like “100 million / 1 billion / 10 billion” over the year (not formalized as a numeric month-by-month KPI, but presented as milestones).

Example distribution / community metrics

  • Bandi Bookstore staff distribution: advance distributed to 50 employees
  • Community example: a 10,000-member group chat
  • Community volume: 108 items mentioned from a platform search (Blind/community context)

Note: Some numerical claims appear conversational and may be inconsistent due to subtitle errors, but the playbook structure remains consistent.

Concrete marketing tactics and examples (campaign ideas)

Marketing channel types (examples listed)

  • Online
    • Reels, Threads, Shorts
    • “Card News”
    • YouTube Shorts
    • podcasts
    • community formats
    • Twitter/thread-like structures
    • open chat rooms
    • influencer marketing
    • shorts-based push
  • Offline
    • posters
    • events (emphasis on publisher-meeting style)
    • in-store activities such as author talks and bookstore staff targeting
  • Named/implied community & ad placements
    • Blind (office-worker community)
    • DC Inside (mockery/discussion mentioned as risk)
    • Ridi Books (listing/entry)
    • GDN / Carrots (ad products/platforms mentioned)
    • “Aladdin main banner” (ad placement mentioned)

Specific idea examples (ranked/selected + brainstormed)

  • Viral / viral-adjacent content
    • “exploding popularity and ranking”
    • content designed to make people “get into it,” not just direct marketing
  • 4Ps-like creativity: story + authority + distribution
    • doctor/trainer/psychological expert endorsements
    • “self-assessment test” concept
    • influencer outreach
  • Event-like hooks
    • Primitive man day performance: barefoot walking in Seoul / drum game at a bonfire (as a viral narrative hook)
    • Convenience-store vs. diet comparison framing (Genshin-themed “diet” + anti-aging/health positioning)
  • Partner distribution
    • pre-distribution to bookstore staff with handwritten letter + signed copies as content
  • Community seeding
    • create or leverage large chat rooms (e.g., 10,000 members)
  • Overseas export angle
    • “news of overseas exports” and leveraging overseas bestseller ranking as a patriotic/viral narrative
    • Japan top-10 bestseller mention as a trigger for shorts/overseas push

Risk/brand management tactic (anti-backlash)

  • The stance is: “Anti-risk is being overestimated.”
  • Online mockery is mentioned (e.g., DC Inside), with recommendations for responding strategically via choices like:
    • ignore
    • respond
    • clarify
  • The decision process is treated as a key messaging/brand area.

Execution mechanics: how AI is used operationally

  • The speaker uses an AI workspace/project with “Project Knowledge”:
    • storing internal materials such as:
      • introduction/table of contents,
      • company context,
      • prior meetings/conversations,
      • business ideas.
  • Prompt sequencing produces outputs in order:
    • formula → marketing taxonomy → 20+ ideas → priority ranking → 2-week schedule → 6-month plan.
  • Iterative refinement:
    • additional instructions can be applied (e.g., ask for “over 3,000 characters” to force deeper task breakdown).
  • Operational discipline:
    • “Follow exactly” the generated plan—print and execute as written.

Implied KPI formula for sales (explicitly stated)

  • The speaker states a relationship like:
    • Sales = (ad exposure / debate?) × admission × time
  • Subtitles are garbled, but the intended model is a multiplicative funnel:
    • awareness/traffic (“debat(e)”) × conversion/admission (“admission”) × duration (“time”).

Actionable recommendations extracted from the talk

  • Decompose goals first (first principles / splitting)
    • Don’t ask AI for a direct “10B strategy”; break it into:
      • units,
      • required behaviors,
      • channel types.
  • Build a marketing channel taxonomy before ideation
    • Online/offline + format-specific list → ideate from that list.
  • Batch ideation + rank
    • Generate 20+ ideas
    • dedup/score them
    • use star ratings and move top picks into execution.
  • Turn strategy into schedules
    • Require:
      • a day-by-day two-week plan
      • a 6-month monthly roadmap
  • Plan by phases
    • Early focus: fast initial conversion target (e.g., “100,000 done quickly”)
    • Later: expand audiences, add lower-priority channels, and amplify with overseas/news triggers.
  • Use a backlash/criticism decision policy
    • Pre-decide how to react to community mockery and treat engagement/read signals as messaging advantage.

Presenters / sources mentioned

  • Presenter / speaker: “Ja-cheong”
    • described as a life hacker who “hacks AI”
    • also claims identity wording like “I was Ja-cheong”
  • Author reference: Reverse Traveler (claimed first place overall; referenced as tied to the speaker’s credential)
  • Books/products referenced:
    • Perfect Genshin Impact
    • Perfect Caveman
    • Perfect Machine
  • AI tools referenced:
    • Claude (primary)
    • GPT (also mentioned)
    • Jaminyai (also mentioned)
  • Other named individuals/entities (examples/case-like references):
    • Elon Musk (used as a first-principles reference)
    • Ellen Muns (name appears garbled; referenced as someone “earning 1 billion a second”)
    • Bookstore: Bandi Bookstore; Ridi Books
    • Communities/boards: Blind, DC Inside
    • Ad/placement mentions: Carrots, GDN, “Aladdin main banner”
    • Brand/product example: “Mori-a-bang” (mobile phone locking device)

Rate this summary

Your feedback will help improve summaries.

Improve this summary

Reprocess with a stronger model when the summary feels incomplete or inaccurate.

Pro

Translate summary in another language

Pro

Ask questions to this video

Chat for follow-up questions, clarifications, and source-backed answers.

Coming soon

Share this summary

Share on X/Twitter Share on LinkedIn Share on Reddit

Original video