Founders explain products badly because they are too close to them. They reach for the category word, the buzz adjective, and the feature list — none of which tell a stranger what the thing does. These 12 prompts force layered explanations: one line, one paragraph, one minute, three minutes. Same product, different depth, jargon stripped at every level.
A note on placeholders: everywhere you see [brackets], swap in your own text before sending. Keep the brackets out of the final prompt.
TL;DR
- Use prompt #1 (elevator pitch) and #6 (no category word) first — they expose whether you actually know what your product does.
- Every explanation needs a named user. “For everyone” reads as “for no one.”
- Borrow Geoffrey Moore’s discipline: if you cannot finish the sentence “unlike [competitor], we [difference],” you have hope, not differentiation.
- Run these on Claude Opus 4.7 or Sonnet 4.6 for the sharpest prose, or GPT-5.5 if you want one model that also handles research and outlining. (Details below.)
- Take the winner live with the landing page copy prompts and product positioning prompts.
Which AI model to run these on (as of June 2026)
The prompts are model-agnostic, but output quality differs. Based on current creative-writing rankings:
| Model | Best for | Price (consumer) |
|---|---|---|
| Claude Opus 4.7 / Sonnet 4.6 | Sharpest, least-clichéd prose; positioning nuance | Claude Pro $20/mo (Sonnet), Max from $100/mo (Opus headroom) |
| GPT-5.5 | One model for writing + research + outlining; best value | ChatGPT Plus $20/mo |
| Gemini 3.1 Pro | Pasting long product docs (1M-token context) | Google AI Pro $19.99/mo |
For pure copy, Claude tends to produce the most natural, least template-y sentences. GPT-5.5 is the value pick when you also want the same chat to research competitors and draft the landing page. Claude (Opus 4.7 / Sonnet 4.6) and Gemini 3.1 Pro both run 1M-token context, so you can paste an entire product doc or transcript and ask for an explanation grounded in it; ChatGPT Plus handles a long brief but reserves the full 1M window for the higher Pro tier. Pricing is current as of June 2026 and changes often.
Best for
- Pitches and demo intros
- About pages and landing hero sections
- Sales emails and outbound
- Onboarding and welcome screens
- Investor updates
- Job listings (where candidates need the 30-second version)
1. One-line elevator pitch
My product: [description in 3 sentences].
Target user: [persona].
Write 8 one-line elevator pitches, each <=18 words.
Use a mix of these formats:
- "X for Y who need Z"
- "the X that does Y without Z"
- "X but for Y"
- "X like [familiar reference], except [key difference]"
- Plain English ("we help Y do Z faster")
Constraints: no "AI-powered", no "platform", no "next-gen".
Mark your top 3 and say why each works for a different audience.2. Audience-rewrite — same product, three readers
Write 3 versions of my product's one-paragraph explanation, tailored to:
(a) a designer
(b) a developer
(c) a marketer
Each <=50 words.
Use language and reference points each audience already knows.
For each version, name 1 word you swapped that the other two audiences wouldn't get.3. Explain to a smart 14-year-old
Explain [product] to a smart 14-year-old.
Constraints:
- No jargon (replace any technical term with a plain one)
- One analogy that connects to teen daily life (school, group chat, sports, games)
- <=100 words
- End with one line they could repeat to a friend
Output: the explanation, then the analogy you used, then what you swapped out.4. Explain in one tweet
Compress [product] into one tweet (<=280 characters).
Must include: what, who for, key outcome.
No hashtags, no emoji, no link.
Mention competitor or category only if it saves words.
Output 3 variants and rank by how shareable each is to a non-customer.5. Explain in a 3-minute spoken pitch
Write a 3-minute (~400 words) spoken pitch for [product].
Structure:
- 60s problem scene (one user, one moment of pain)
- 60s my insight (the thing competitors miss)
- 60s how it works (the simplest demo arc)
- 60s proof + next step
Constraints:
- Conversational, contraction-friendly
- Read aloud — no sentence longer than 20 words
- One number, one named user, one named outcome6. Explain without the category word
Explain [product] without using its category name (e.g., don't say "CRM", "newsletter platform", "AI assistant").
Describe what it actually does for the user.
<=80 words.
Then explain why the category word was hiding the real value.
This is the version to use when the category is crowded or generic.7. Explain via before-and-after
Write a before / after / how-we-bridge explanation of [product].
3 paragraphs, 50 words each.
Concrete examples, not abstract claims.
The "before" must describe a real workflow with named tools and pain points.
The "after" must describe the same workflow, end-to-end.
The bridge paragraph must name the 1 thing that changes the outcome.8. Explain via one specific user’s day
Write a 200-word "day in the life of a user" piece showing [product] in context.
Pick a specific persona: [audience].
Be concrete:
- Time of day
- What they were doing before
- The moment [product] enters
- What changed in their next action
- The downstream effect by end of day
Tone: third-person, narrative. No marketing voice.9. Explain to a skeptical investor
Write a 150-word pitch to a skeptical investor for [product].
Lead with the market insight nobody else has — phrased as a sentence.
Body: why now (3 sentences max), how we win (3 sentences max).
End with 1 piece of traction evidence (a number with a denominator, not just a number).
Banned: "$100B market", "TAM", "10x", "category-defining".
If the insight isn't sharp enough, output the insight on its own line at the top and say "this needs work".10. Explain via “we’re NOT X”
Write 5 short paragraphs in the format "We're not X".
X is a common misconception about my category.
Each paragraph: 50 words.
Each ends with "We're actually [plain-English description]".
Pick the 5 X's that customers actually confuse us with, not strawmen.
For each, name 1 feature that proves the difference.11. Explain via the “moment of value”
Describe the single moment in a user's experience where [product] delivers its value.
Constraints:
- 100 words
- Past tense, narrative
- Include: what they clicked, what they saw, how long it took, how they felt 5 minutes later
- No screenshots, no product features named directly
This is the explanation to use in onboarding, sales demos, and the hero section. It must work without showing the product.12. Explain via the test you can fail
Write a "would this be a fit?" qualifier for [product].
Format: 5 yes/no questions a prospect can self-score on.
Above the questions: 2 sentences naming who this is for.
Below: scoring guide — 5 yeses = strong fit, 3-4 = maybe, <=2 = not for you.
Constraints:
- Each question must be a real disqualifier (we'd actually decline the sale)
- No vanity questions ("are you ambitious?")
- The "not for you" outcome must be honest — point them somewhere elseHow to feed the model so it doesn’t guess
Most weak output comes from a thin brief, not a weak model. Before you run any prompt above, paste a short fact block first:
Context for everything below:
- What it does (one plain sentence, no adjectives)
- Who buys it (job title + the moment they go looking)
- The single competitor or workaround we replace
- One real outcome, with a number and a denominator
- Two words customers use that we should mirror
Don't write yet. Confirm you have these, then wait for my next message.This forces the model to anchor on your facts instead of inventing generic SaaS copy. On Claude or GPT-5.5 you can also attach your homepage, a sales call transcript, or your docs and tell it to pull the plain-English description from there rather than from your marketing page.
Common mistakes
- Jargon explanations that need 3 more jargon words to understand.
- Category-bloat (“the AI-powered next-gen platform that empowers teams to…”).
- Skipping the “for whom” anchor — every explanation needs a named user.
- Listing features when the reader asked what the product does.
- Writing one explanation and reusing it for every audience. Investor copy in onboarding kills activation.
- Accepting the first draft. These prompts are meant to give you 3-8 options to cut down, not one to ship.
FAQ
Which prompt should I start with? Run #1 (elevator pitch) and #6 (explain without the category word) back to back. If you cannot explain the product without naming its category, you do not yet understand its value — and #6 will expose that fast.
Which AI model gives the best result? For pure copy, Claude Opus 4.7 or Sonnet 4.6 (Claude Pro is $20/mo) tend to produce the most natural sentences. GPT-5.5 (ChatGPT Plus $20/mo) is the value pick if you want one chat that also researches competitors. Gemini 3.1 Pro is handy when you want to paste a long product doc. All current as of June 2026.
How do I stop the AI from writing generic marketing fluff? Two moves: feed it the fact block from the section above, and use the constraint lines in each prompt that ban specific words (“AI-powered”, “platform”, “TAM”). Banning the clichés forces the model toward concrete language.
Can I use the output as-is? Treat every draft as raw material. Pick the strongest variant, cut it by a third, and read it aloud — if you stumble, your reader will too. Then validate the claim with a real user before it goes on the page.
How is this different from a positioning statement? Explanation is what you say to a stranger in one breath; positioning is the strategic decision underneath it (which customer, which competitor, which category). Nail positioning first with the brand positioning statement prompts, then use these to express it.