Video summary

Beginners guide to AI copywriting (Claude Tutorial)

Main summary

Key takeaways

Educational

Main ideas / lessons

  • AI copywriting is a practical business skill: The creator positions AI tools (especially Claude, also ChatGPT) as ways to quickly produce marketing copy that would normally take much longer to write manually.
  • Credibility without engineering: They emphasize they’re not a software engineer or coder, but they’ve spent lots of time using AI, consuming startup/VC content, and even building a startup to demonstrate real-world familiarity.
  • Prompting quality matters: Don’t ask vague requests. You must provide clear instructions, style references, and a structure the AI should follow.
  • A repeatable VSL framework produces better results: The video provides an explicit outline/formula for writing VSLs (Video Sales Letters), with optional applicability to emails, landing pages, and headlines.
  • Workflows speed up production dramatically:
    • Use saved “projects” and reusable prompts in Claude.
    • Feed deep research (via GPT deep research) into Claude as product knowledge/context so the output becomes more accurate and value-dense.
    • Expect to edit only slightly compared to the amount of writing time saved.
  • Examples of broader copy uses: Beyond VSLs, they show how to use Claude for LinkedIn/Twitter posts, including authority positioning and a conversational tone.
  • Information quality determines output quality: The AI can structure copy, but the research you provide drives factual strength and usefulness.

Methodology / instructions (detailed)

1) Prompting AI for VSL writing (Claude/ChatGPT)

  • Start with a specific business outcome rather than a generic request.
    • Bad example: “Write me a VSSL about AI.”
    • Better example: “Write me a VSSL about how to scale your business with AI.”
  • Add style/emulation constraints:
    • Instruct the AI to emulate specific direct-response copywriters/authors (named in the video).
  • Require direct-response structure:
    • Specify that the VSL should follow classic direct-response sales letter examples.
  • Provide a detailed VSL outline (see below) so the model follows a complete persuasion path.

2) VSL structure / copy formula to follow

Use this as the required framework for the AI to generate the script:

  • Strong hook
    • Incite curiosity.
  • Education / value bridge
    • Bridge into a helpful education portion.
  • Problem amplification
    • Create problems or amplify pain.
  • Solution bridge
    • Bridge into the solution the creator discovered (framed like a “hero’s journey”).
  • Smooth start of the pitch
    • Transition cleanly into the offer.
  • Limited-time pitch with scarcity
    • Include time limitation/scarcity elements.
  • Make the offer
    • Clearly present what the audience should buy/do.
  • Reiterate the offer three times
    • Repeat the core offer multiple times to improve conversion.
  • FAQ section
    • Answer fears/objections and likely questions.

The creator also describes the formula as a loop—implying repetition/iteration within the flow.

3) Use saved “projects” in Claude to reuse prompts

  • Create a Claude Project that stores:
    • The VSL prompt instructions (framework + style constraints).
    • Any role/system-like guidance (e.g., writing type such as “AI VSSL”).
  • When generating content:
    • Provide source material only (“spit in the source material” / paste context).
    • Claude automatically follows the saved prompt structure.

Optional workflow enhancement:

  • Use ChatGPT deep research on a niche/problem.
  • Paste the research into Claude as product knowledge and additional context.

4) Add deep research to increase accuracy and value

  • Perform deep research on the niche/business problem (via GPT deep research).
  • Example research direction in the video:
    • How companies use AI to scale business, replace slow employees, and save money
    • Focus narrowed to online businesses / info product businesses.
  • After research is generated:
    • Save the research output.
    • Paste it into the Claude project field for product knowledge / context.
    • Instruct Claude to rewrite the VSL using the new context and research findings.
  • Encourage the use of specific, sourced statements:
    • The creator suggests requesting “stats,” “studies,” and named sources (e.g., “OpenAI says…”) to improve credibility.

5) Use Claude for LinkedIn/Twitter copy (saved prompts)

  • Set Claude’s role as a ghost writer for the user’s accounts.
  • Provide instructions to define positioning and tone:
    • Position the author as an authority in categories like AI, sales, SaaS, and online business.
    • Keep posts conversational (avoid sounding like corporate/robotic copy).
  • Save these instructions into a project so future posts can be generated quickly using that template.
  • Validate with engagement claims (they reference high views/likes/leads from content generated this way).

Tools and platforms mentioned

  • Claude (primary for VSL and copy generation; “Projects” feature for saved prompts)
  • ChatGPT
  • GPT Deep Research (used to collect niche/context info)
  • LinkedIn
  • Twitter/X
  • Mentions of VSL/VSSL, plus optional applicability to headlines, landing pages, and emails (other copy areas)

Speakers / sources featured

  • Primary speaker/creator (unnamed): The YouTube host who demonstrates Claude/ChatGPT workflows and provides the VSL prompt framework.
  • Copywriting figures named for emulation/style constraints:
    • Dan Kennedy
    • Jason Flad Lean (spelled this way in the subtitles)
    • David Ogilvy
  • Other named/indirect source:
    • “OpenAI” is mentioned as a type of source for credibility (example of how to cite claims), not as a direct speaker.

Original video