Ik bouw een merk in 24 uur

Summary (business execution focus)

A creator demonstrates a “brand-from-scratch” build process—using AI tools (Cloud Code / ChatGPT / image generation) to go from identifying a customer pain point to producing a first, mobile-ready one-page e-commerce website and a brand book within roughly “12 hours” (later suggesting it’s closer to ~2 hours of active work in this demo).

---

Core strategy / playbook (pain-point first)

• Start with customer pain, not products

  • Instead of finding a product and then advertising it, first identify the biggest pain points people actively worry about.

• AI-assisted pain discovery workflow

  • Run broad prompts to extract pain categories.
  • Iterate by tightening the prompt to find pain that is:
    • Emotionally sticky (keeps people “awake at night”)
    • High-insecurity (e.g., appearance/odor/sweat stigma)
    • Purchase-worthy because the solution must remove the pain, not just “offer a benefit”

• Pick the pain → derive product concepts

  • For the chosen pain (sweat stains / underarm sweat visible through shirts), generate solution options such as:
    • Sweat-proof undershirt with waterproof underarm panels (primary pick)
    • Anti-perspirant products
    • Clothing fresheners / garment refresh concepts
    • Disposable underarm pads (considered mainly as an upsell)

---

Frameworks / processes used (implicit and explicit)

• Pain-point funnel

  1. Broad pain scan
  2. Filter for “real emotional urgency”
  3. Select top pain
  4. Pain → solution categories → shortlist product concepts

• Competitor targeting framework

  • Review competitors and differentiate via audience specificity
  • The goal isn’t “who does it better,” but how to narrow the niche.

• Brand book / system prompt approach (prompt “control” method)

  • Generate a brand book (name, font, colors, voice, story, etc.).
  • Use a tight prompt so the AI doesn’t produce inconsistent or random outputs.

• Creative-to-build pipeline

  • Generate/collect brand visuals first.
  • Feed them into Cloud Code to assemble the site.
  • Treat Cloud Code as an assembler requiring “perfect inputs,” not a creator from scratch.

---

Concrete example: brand concept built in the demo

Selected pain point

• Visible sweat stains / underarm sweat through clothing
  • Other pains are rejected as “not painful enough” (examples mentioned: litter box mess, ironing pile, cycling glasses, etc.).
  • The focus is on insecurity and embarrassment—and brands can “play with emotions” only if the product truly fixes the pain.

Product direction (initial pick)

• “Sweatproof undershirt”
  • Prevent sweat from reaching the outer shirt.
  • Notes on strategy:
    • Fit/appearance is less critical since it’s worn underneath.
    • Comfort and effectiveness still matter.

Target niche chosen via competitor review

• Competitors skew toward men, so he chooses women due to faster online purchase behavior for pain-point products.
• Then narrows further:
  • Women who feel insecure while working out and sweating
• Differentiation thesis:
  • Specialize for an underserved audience segment rather than selling a generic “sweat solution.”

---

Operational workflow (step-by-step execution)

1) AI planning and prompt library

• He claims to keep “mega prompts” updated (not relying on outdated prompts).
• Uses iterative prompt refinement when results are off-target.

2) Competitor research (store targeting)

• Reviews competitors for:
  • who they target (age/gender)
  • how specific their positioning is
• Output decision:
  • focus on women, then further on athletic/sweating insecurity.

3) Build the brand book (system design)

• Generate:
  • brand story / voice / tone
  • color palette
  • fonts
  • tagline(s)
• Emphasis:
  • control output by doing research on fonts/colors first, then locking them into subsequent prompts to prevent “weird things.”

4) Generate visuals before website assembly

• Image generation tasks (examples):
  • product page badges/logos
  • hero image for homepage
  • explainer photos (how it works)
  • social image (woman wearing it)
  • “how it blocks sweat” proof images
  • review-style visuals
• Key execution issue flagged:
  • early images accidentally included the logo on the shirt; he re-prompts to remove logos to support marketing flexibility/dropshipping realism.

5) Cloud Code website assembly (prompt-driven)

• Provides Cloud Code with:
  • brand book + photo pack + product reference images
  • folder structure for inputs
• Output:
  • a one-page site foundation including explainer sections, reviews, and Q&A
• Iterative improvements:
  • copywriting and layout fixes (copy was “weak” at first)
  • remove duplicate product variations (“one product page, not two colors”)
  • add more visuals so the site feels complete

6) Mobile-first optimization (observed)

• Checks the site on phone early:
  • layout and photo sizing look clearer on mobile
  • prioritizes clarity since most traffic is mobile.

---

KPIs / metrics mentioned

• No direct targets provided for revenue, CAC, LTV, or ROAS.
• Indirect “cost/ops” metric:
  • estimates cost of goods from supplier research/AliExpress:
    • supplier price ~€6 estimate
    • implies potential listing around ~€8
    • suggests bundle pricing to increase order value
• Explicitly deferred:
  • “We don’t know about the supplier quality yet”
  • “We don’t know break-even ROAS yet”

---

Pricing / offer tactic (bundle logic)

• He recommends selling in bundles to:
  • raise average order value
  • improve per-piece economics (e.g., buying 3 becomes more advantageous)
  • make the purchase feel like a “no-brainer”
• Upsells are considered (e.g., disposable underarm pads), but the core product is prioritized.

---

Actionable recommendations extracted from the demo

• Choose a pain that is emotional + urgent

  • If the product doesn’t remove the pain (fix insecurity), the strategy is weak.

• Differentiate by audience specificity

  • Target the segment with the strongest pain and purchase behavior (he chooses women working out).

• Treat AI site builders like “render engines”

  • Provide perfect brand assets, exact visuals, and explicit instructions.
  • Don’t let the AI freely “fill in”—quality drops.

• Generate enough proof visuals

  • Early builds feel incomplete without many images showing:
    • what the product does
    • why it works
    • social proof/reviews

• Iterate with tight feedback loops

  • Re-prompt when logos or visual details don’t match dropshipping reality.

• Validate mobile-first early

  • Mobile clarity is treated as a key success condition.

---

Presenters / sources

• Presenter: The YouTube creator/host (name not provided in the subtitles).
• Referenced source: Daniel Dalen (inspiration for “launching a company in two days”).
• Tools mentioned: “Cloud Code” (CloudCode), “ChatGPT” / “Chat,” image generation tools (e.g., “Midjourney” mentioned as a comparison), and AliExpress for supplier/product reference.