Whiteboard Clarity and Jargon Stripping

Purpose

An AI system can produce competent-sounding, well-organized prose in seconds — and its default register skews toward exactly the kind of professional-sounding jargon ("synergistic," "optimized," "comprehensive end-to-end solution") that reads as confident but says almost nothing. When content is cheap to generate, the practitioners who cut cleanest through that noise become the differentiator. This skill is a deliberate, testable pass for catching and removing that jargon — not a general writing-quality check, but three specific, repeatable tests aimed at the failure mode AI-assisted writing produces most often.

Anchored in

Three practitioners' accounts, supplied by the user from source video transcripts. Notion's Max Schoening, on the failure mode and the fix: "when people make landing pages, first of all their writing skills just deteriorate immediately because they want to sound clever and marketing speak comes out of their mouth... just pretend you're standing in front of a whiteboard — what's the manic thing that you're drawing on the whiteboard to communicate this?" Khosla Ventures' Keith Rabois, on why this specifically matters more as building gets cheaper: "it's the storytelling — how do you cut through the clutter in the snappiest, most compelling possible way... the person who can say 'this is the way to frame it' is worth like all the tools in the world." Webflow's Jessica Fain, on translating a leader's own instinct rather than synthesizing around it: "marrying that leader's instincts with your ability to... really take that approach and accelerate it — holding the whiteboard marker was a great tool."

Method

1. Run the slop-strip test on any AI-drafted or AI-assisted text before it ships. Go through the text and mark every adjective and passive construction that could be deleted without losing real information — words like "synergistic," "optimized," "comprehensive," "seamless," "robust," "cutting-edge," and passive phrasing that hides who's doing what to whom. If a sentence still says something specific after those are cut, keep the cut. If it says nothing once they're gone, the sentence was never carrying real content — replace it, not just trim it.
2. Apply the whiteboard test to the core idea, out loud, before trusting any written version. Imagine explaining the idea to a specific friend, standing at a whiteboard, with 45 seconds and only a marker: what's the one thing you'd actually draw — boxes, an arrow, 3-5 words? If the honest answer is "I'd need slides for this," the idea isn't compressed enough yet to write up cleanly, no matter how polished the eventual document looks. Do this test before writing the full version, not as a check afterward — it's a compression exercise, not a proofread.
3. Compress the value proposition to one concrete, memorable line — the Steve Jobs standard. Not a category description ("an AI-powered productivity platform") but a specific, human benefit stated so plainly it sticks (the reference standard: "1,000 songs in your pocket," not "a portable digital music device with expanded storage capacity"). Test candidates against the question: could a competitor's product honestly claim this same line? If yes, it isn't specific enough yet — go back to what's actually distinctive.
4. Distinguish this pass from logical structuring — do both, in order. This skill fixes register and clarity (does it sound like a human explaining something they understand, or like AI-generated copy); executive-narrative-and-storyline fixes argument structure (is the reasoning MECE, does every claim trace to evidence). A logically perfect storyline can still read as jargon-heavy AI slop, and a punchy, clear line can still be structurally unconvincing on its own — run the storyline pass first for structure, then this pass for register, not the other way around.
5. When drafting on the AI side of this process: hold the same standard proactively, not just on request. Default toward the plainest phrasing that's still accurate, flag a sentence that reads as jargon-heavy back to the user rather than polishing it further, and prefer a concrete example or number over an abstract adjective wherever one is available.

What this skill does NOT do

• Doesn't simplify away real technical or financial precision where precision is actually required (a legal clause, a regulatory disclosure, a financial figure) — this is a register fix for narrative and pitch content, not a mandate to oversimplify everywhere.
• Doesn't replace executive-narrative-and-storyline for argument structure, or business-case-builder for the underlying substance — it improves how already-sound content is said, not what's true or well-reasoned.
• Doesn't guarantee an audience will agree with the content once it's clear — clarity makes a weak argument more visibly weak, not stronger.

Refinement notes

The three named practitioner attributions are drawn from source video transcripts supplied by the user and have not been independently cross-verified word-for-word against a second source — treat the underlying techniques (slop-strip, whiteboard test, one-line compression) as well-grounded practitioner heuristics, the exact phrasing as attributed but not independently re-confirmed.

Continue from here

• Before this, for argument structure: ../executive-narrative-and-storyline/SKILL.md
• Predicting audience reaction to a concept before writing it up: ../../../business-design-frameworks/skills/taste-emulation-heuristic/SKILL.md
• Understanding what a specific decision-maker actually needs to hear before writing to them: ../stakeholder-pressure-and-information-gap-mapping/SKILL.md
• Turning a vision into a PR-FAQ before anything is built: ../../../prototyping-and-demonstration/skills/opportunity-visioning-with-pr-faq/SKILL.md
• This pack's shared guardrails: ../../CLAUDE.md

References

• ../../CLAUDE.md — the pack's shared guardrails

---

Reminder: frontmatter has only name and description. Everything else goes into skills_index.json (run scripts/generate_index.py).