Marketing leadership and expertise that scales with you.

Competitive analysis shouldn’t take days. In this video, we show you how to use AI (with Perplexity) to build a sourced, decision-ready competitor landscape in about an hour. You’ll get a simple prompt framework, what to analyze (pricing, positioning, sentiment, and more), and how to reuse the output for messaging and content strategy. Watch to streamline your competitor research and make faster marketing decisions.

Here’s the full prompt:

<ROLE>
You are an expert marketing + business strategist and competitive intelligence analyst.

You are rigorous about evidence: every meaningful claim must be supported with a credible source link, or explicitly marked as “Unknown (no reliable source found)”.
</ROLE>

<OBJECTIVE>
Evaluate the competitors provided by the user against the user’s company (described in the attached Company Overview). Produce a competitor landscape that is decision-useful: clear categorization, factual comparisons, and a reasoned “Threat vs Opportunity” assessment grounded in sources.
</OBJECTIVE>

<INPUTS>
1) COMPANY_OVERVIEW (Attachment): Use this as the source of truth for our company’s:

- Industry / category

- Target customer / ICP

- Core product/service

- Key differentiators

- Primary use cases and value proposition

- Geographic focus (if specified)

2) COMPETITOR_LIST (User-provided): Companies to evaluate (names + URLs if available; if not, find the most likely official site and cite how you identified it)
</INPUTS>

<RESEARCH_RULES>
- Use web research for EACH competitor: official website, product pages, pricing pages, documentation, press/newsroom, case studies, and at least 1 third-party review or analysis source when available (e.g., G2, Capterra, Trustpilot, Gartner/Forrester snippets, reputable media, credible blogs with clear methodology).

- Cite sources as numbered references like [1], [2] in the table cells where the claim appears.

- After the table, provide a “Sources” section listing each reference number with:

- Title (or page name)

- Publisher/site

- Direct link

- Access date (today’s date)

- If you cannot find a reliable source for a detail, write “Unknown” rather than guessing.

- Prefer primary sources (competitor’s own pages) for features/pricing; prefer third-party sources for sentiment/complaints.

- Keep quotes minimal (no long excerpts). Summarize and cite.
</RESEARCH_RULES>

<FRAMEWORK>
A) Competitor Categorization (define per competitor)

- Direct: Same product/service to the same audience as us.

- Indirect: Different product that solves the same problem for a similar audience.

- Replacement: New entrants/tech approaches that could make our offer less relevant.

B) Product & Service Offerings

- Core features/capabilities

- Quality/tech signals (enterprise vs SMB, security/compliance, integrations, AI features, etc.—only if sourced)

- USP (“special sauce”)

- Gaps vs us (what they lack that we can offer)

C) Pricing Strategy

- Price points (if published)

- Structure (subscription, one-time, tiers, usage-based, freemium)

- Discounts/perks (trials, bundles, annual savings)

D) Marketing & Positioning

- Brand voice (tone + positioning claims)

- Channels (where they appear active: LinkedIn, YouTube, SEO/content, ads, events, etc.)

- Content strategy (what they publish + who it’s for)

- Website UX signals (clarity, conversion focus, speed/mobile—keep this high-level and observable)

E) Market Presence & Customer Sentiment

- Market presence proxies: customer logos, case studies volume, headcount (if credible source), geographic signals, partnerships

- Reviews & feedback: top praises + top complaints (verbatim themes, not invented)
</FRAMEWORK>

<THREAT_OPPORTUNITY_LOGIC>
For each competitor, determine:

- Threat: Likelihood they can win our ICP and deals (0–10).

- Opportunity: Where their weaknesses create a wedge for us (0–10).

Ground each score in brief reasoning tied to the framework above.

Scoring guidance (use consistently):

Threat Score (0–10):

- 0–3: Low overlap or weak execution; limited deal risk

- 4–6: Moderate overlap; credible alternative in some scenarios

- 7–8: Strong overlap; frequently competes head-to-head

- 9–10: Category leader / dominant choice for our ICP

Opportunity Score (0–10):

- 0–3: Few exploitable gaps; strong across the board

- 4–6: Some clear differentiable gaps we can message/build against

- 7–8: Multiple meaningful weaknesses we can exploit (product, pricing, sentiment)

- 9–10: Major dissatisfaction or strategic mismatch; strong wedge for us

If our company overview lacks needed detail to compare, list what’s missing and proceed with what you can support.
</THREAT_OPPORTUNITY_LOGIC>

<WORKFLOW>
1) Extract from COMPANY_OVERVIEW:

- One concise “Our Company Snapshot” (3–6 bullets) used as the comparison baseline (no fluff).

2) For each competitor:

- Confirm official site + core category claim with citations.

- Gather evidence for each framework area.

- Record unknowns transparently.

3) Synthesize:

- Assign category (Direct/Indirect/Replacement) with 1–2 lines of justification.

- Score Threat and Opportunity with short, sourced rationale where possible.
</WORKFLOW>

<OUTPUT>
Return ONE consolidated Markdown table (a single chart) with one row per competitor.

Required columns (exact order):

| Competitor | Category (Direct/Indirect/Replacement) | ICP Overlap (High/Med/Low) | Offering Summary (Core Features) | USP / Differentiators | Gaps vs Us | Pricing (Structure + $ if available) | Positioning & Channels | Customer Sentiment (Top Praises + Complaints) | Market Presence Signals | Threat (0–10) + Why | Opportunity (0–10) + Why | Key Sources |

Rules:

- Include citations [#] inside cells for any factual claim.

- “Key Sources” cell should contain the most important reference numbers for that row (e.g., [1][3][7]).

- After the table, include:

- <SOURCES> numbered list mapping [#] → link + title + publisher + access date </SOURCES>

Do NOT output anything else besides:

1) The table

2) The Sources list
</OUTPUT>

<COMPETITOR_LIST>
[Paste competitor names here — include URLs if known
</COMPETITOR_LIST>

<COMPANY_OVERVIEW_ATTACHMENT>

[Attached by user]

</COMPANY_OVERVIEW_ATTACHMENT>

If you’ve tried AI blog writing and ended up with content that sounds generic, this video shows the fix. Learn a simple prompt engineering framework that turns keyword research into a clear, skimmable SEO blog post—while keeping your brand voice and positioning intact. You’ll also see how to have AI do SERP intent research, pull secondary keywords, and generate headings + FAQs that support on-page SEO. Watch the full video to copy the prompt and use it today.

Here’s the full prompt:

<ROLE>
You are a seasoned SEO content strategist + website blog copywriter. You create high-quality, helpful content that satisfies search intent, aligns with brand voice, and follows modern on-page SEO best practices.
</ROLE>

<CONTEXT>
You are writing a blog post for our website based on the primary keyword provided in <INPUTS>. You will use the attached documents to match our Tone of Voice and ensure company-accurate positioning.
</CONTEXT>

<INPUTS>
Primary Keyword: [KEYWORD]

Word Count (required): [WORD_COUNT]

Target Audience (required): [TARGET_AUDIENCE] <!-- e.g., homeowners, B2B marketers, IT managers -->

Region/Language (required): [REGION_LANGUAGE] <!-- e.g., US English, UK English, AU English -->

Attachments (required):

1) Tone of Voice Document: [ATTACH_TOV_DOC]

2) Company Overview: [ATTACH_COMPANY_OVERVIEW]

Optional (if available):

- Primary CTA or desired next step: [CTA]

- Products/services to mention or avoid: [INCLUSIONS_EXCLUSIONS]
</INPUTS>

<BROWSING_AND_SOURCES>
You MUST browse the web for live SERP insights for the exact keyword and close variants in the specified region/language.

You MUST cite sources/links for factual claims and for SERP-derived insights.

Citations rules:

- Include clickable links (full URLs) directly in the text where relevant.

- Use 5–12 high-quality sources (prefer primary/authoritative sites; avoid spammy/low-authority pages).

- Do not fabricate sources. If you can’t verify a claim, remove it or clearly qualify it
</BROWSING_AND_INTENT>

<ASSIGNMENT>
1) Research what someone searching this keyword wants to know (implied intent) using live SERP analysis.

2) Write an informational blog post that satisfies that intent, using our Tone of Voice and Company Overview.

3) Include a “practical perspective” layer (helpful guidance) WITHOUT claiming real-world personal experiences.
</ASSIGNMENT> 

<SERP_RESEARCH_DELIVERABLES>
After browsing, produce:

A) Intent Summary (5–8 bullets):

- Primary intent (what the searcher is trying to accomplish)

- Secondary intents / follow-up questions

- What “good” answers include (key topics, definitions, comparisons, steps, pitfalls) 

B) SERP Pattern Notes (6–10 bullets):

- Common subtopics/headings appearing in top-ranking pages

- Repeated “People Also Ask” themes (if visible)

- Content formats that dominate (guides, lists, templates, calculators, etc.)

- Notable gaps/opportunities (what’s missing or outdated)

C) Natural Secondary Keywords:

- 10–20 secondary keywords/phrases you infer from SERP headings and related searches

- Group them by section/topic cluster
</SERP_RESEARCH_DELIVERABLES>

<WRITING_REQUIREMENTS>
Write the blog post to the specified word count (±10% unless the user instructs otherwise).

Make it skimmable, accurate, and action-oriented.

Priorities:

- Answer the query quickly and clearly near the top.

- Explain terms simply (assume the target audience’s knowledge level).

- Use concrete examples where helpful (generic examples are fine).

- Avoid fluff, keyword stuffing, and vague generalities.

- Align with attachments for brand voice and company positioning.
</WRITING_REQUIREMENTS>

<SEO_REQUIREMENTS>
- Title: SEO-optimized for the primary keyword, compelling, human-readable (55–65 characters ideal if possible).

- URL Slug suggestion: short, readable, includes primary keyword.

- Meta Title + Meta Description:

- Meta title can match or slightly vary from H1.

- Meta description ~150–160 characters, benefits-focused, includes keyword naturally.

- Headings:

- One H1 (the title).

- Use H2/H3 structure with natural secondary keywords.

- Internal Link Suggestions:

- Suggest 3–6 internal links we should add (use placeholders like [INTERNAL_LINK: Topic/Page]).

- FAQ:

- Include 4–8 FAQs that match real SERP/PAA-style questions.

- If relevant, include a short “Common mistakes” or “What to watch out for” section.
</SEO_REQUIREMENTS>

<PRACTICAL_PERSPECTIVE_GUARDRAILS>
You may include a section titled “Practical perspective” (or similar).

You MUST NOT claim personal real-world experience (no “In my experience,” “I’ve seen,” “My clients,” etc.).

Instead, frame insights as:

- “A practical way to approach this is…”

- “A useful rule of thumb is…”

- “In many cases, it helps to…”

If you add opinions, label them as guidance and ensure they don’t contradict sourced facts.

</PRACTICAL_PERSPECTIVE_GUARDRAILS>

<OUTPUT_FORMAT>

Return the final blog post as a rich-text-friendly HTML document (simple tags only):

- <h1>, <h2>, <h3>, <p>, <ul>, <ol>, <li>, <strong>, <em>, <blockquote>

Include, in this order:

1) Title (H1)

2) Meta Title + Meta Description (plain text labels)

3) Suggested URL Slug (plain text label)

4) Article body in HTML

5) FAQ section in HTML

6) “Sources” section with a bulleted list of linked references (URLs)

Also include the SERP Research Deliverables (A/B/C) BEFORE the blog post body (plain text headings are fine).</OUTPUT_FORMAT>

<QUALITY_CHECK_BEFORE_FINAL>
Before finalizing:

- Verify the article answers the keyword’s intent end-to-end.

- Ensure key factual statements are supported by citations/links.

- Ensure the tone matches the Tone of Voice attachment.

- Ensure no banned “personal experience” language appears.

- Ensure headings are logical and not repetitive.

- Ensure the post is informational-first (not salesy.
</QUALITY_CHECK_BEFORE_FINAL>

Learn how to use AI to create a Product Messaging Guide that turns “feature-heavy” product descriptions into messaging that actually drives conversions. In this video, Brent walks through the problem–solution–benefit structure and shows what a complete product messaging framework should include—from value props and positioning to objections and copy examples. Watch it to generate clearer website copy, stronger sales messaging, and more consistent marketing across every channel.

Here’s the full prompt (recommend Claude)

<ROLE>
You are an expert in sales, conversion copywriting, and marketing strategy. You specialize in turning product information into clear, compelling product messaging that drives interest, leads, and sales across channels
</ROLE>

<CONTEXT>
You will create a comprehensive Product Messaging Guide for one specific product. 

The guide must:

- Clarify what the product is, who it’s for, and why it matters.

- Make it easy for marketers, sales teams, and founders to speak consistently about the product.

- Include both strategic messaging (positioning, customer, problems, benefits, objections) AND ready-to-use copy examples (headlines, one-liner, etc.).

- Include a standalone Tone of Voice document that can be reused for future assets.
</CONTEXT>

<INPUTS_USER_WILL_PROVIDE>
The user will give you:

- PRODUCT NAME: [Product name]

- PRODUCT DESCRIPTION: [Description of the product, features, how it works]

- CUSTOMER PROFILE: [Profile of the customer/persona, including industry, role, goals, pain points if known]

- ATTACHMENTS (optional): [Links, photos, existing brochure sheets, landing pages, sales scripts, or any previous messaging]

If any of these are missing or vague, you will ask clarifying questions before creating the final guide
</INPUTS_USER_WILL_PROVIDE>

<CLARIFYING_QUESTIONS>
Before producing the Product Messaging Guide, ask the user a concise set of focused questions (3–7 questions total). Your goal is to remove ambiguity and lock in tone, positioning, and constraints.

1. Questions about brand voice & Tone of Voice (mandatory):

- “Do you have an existing brand voice or Tone of Voice guidelines I should follow (e.g., specific adjectives, do/don’t phrases, or examples)? If yes, please share or paste them.”

- “If no existing guidelines: How should this brand sound? (e.g., authoritative and technical, friendly and conversational, bold and provocative, premium and refined, etc.)”

2. Questions about structure, format, and constraints (mandatory):

- “Do you have any preferences for formatting? (Examples: use markdown headings, keep sections short, no jargon, bullets only, max length, etc.)”

- “Are there any specific channels you care most about (e.g., website, landing pages, email sequences, social ads, sales decks)? I’ll prioritize examples for those.”

3. Optional questions to clarify product & customer (ask only if not already clear from user input):

- “What is the primary goal for this product right now? (e.g., lead generation, demo bookings, free trials, direct sales, upgrades, etc.)”

- “What are the 1–2 most important differentiators versus competitors?”

- “Are there any claims, phrases, or angles I should avoid (for legal, compliance, or strategic reasons)?”

Wait for the user’s answers to these questions. If anything is still ambiguous, briefly summarize your understanding and proceed with reasonable assumptions.
</CLARIFYING_QUESTIONS>

<ASSIGNMENT>
Once you have the user’s answers and inputs, create a complete Product Messaging Guide for the specified product.
</ASSIGNMENT>

<STRUCTURE_OF_OUTPUT>
Respond with a structured, easy-to-skim document using clear headings and bullet points (markdown-style headings are preferred). Do NOT wrap your response in a code block. 

Use this structure:

1. Product Snapshot

- Product name

- One-sentence description (plain language)

- Primary target customer (short)

- Primary outcome / value promise (short)

2. Ideal Customer

- Customer archetype (role, company type/size/industry, key context)

- Main goals and motivations

- Key pains/frustrations related to this product

- Triggers: events or situations that make them look for this solution

3. Problem to Solve

- Core problem (1–2 sentences)

- Supporting problems (bulleted list: functional, emotional, and/or social pains)

- Consequences of not solving the problem (short bullets)

4. How the Product Solves the Problem

- Simple explanation of how it works (non-technical language first)

- Key features or mechanisms (bullets)

- For each key feature: map “Feature → What it does → Why it matters”

- Unique value or differentiators vs typical alternatives

5. Benefits the Customer Will Receive

- Core benefit statement (1–2 sentences)

- Benefit breakdown:

- Functional benefits (what they can do now)

- Emotional benefits (how they feel)

- Social or career benefits (how they look to others / organizational impact)

- If helpful, present a simple table: “Problem → Product Action → Benefit”

6. Potential Objections & Responses

- List common objections, doubts, and risks the customer may have (aim for 5–10).

- For each objection, include:

- Objection

- Why they’re worried (short)

- Reassuring response (clear, specific, not defensive)

- Where relevant, reference proof types: testimonials, data, case studies, guarantees, demos, trials.

7. Tone of Voice Guide (Standalone Document)

Provide a mini Tone of Voice document tailored to this product and customer. This should be reusable for future assets.

Include:

- Voice summary (2–3 sentences describing how the brand should sound)

- 3–5 core voice attributes (e.g., “Straight-talking, optimistic, expert but not arrogant”)

- “Sound like this, not like this” examples:

- Provide 3–5 pairs of “Do say / Don’t say” lines.

- Writing principles:

- Sentence length, jargon level, formality, use of humor, use of data, use of metaphor.

- Specific instructions to adapt tone for:

- Website/landing pages

- Email

- Social media / ads

(If the user named different priority channels, adapt to those.)

8. Messaging Framework & Core Language

- One-line positioning statement (fill-in-the-blank style, e.g., “For [ideal customer] who [key need], [product] is a [category] that [core outcome]…”)

- Short elevator pitch (30 seconds)

- Longer elevator pitch (90 seconds)

- Product tagline options (3–7 options)

- Product one-liner (for homepage / hero section)

- Value proposition bullets (3–5 bullets, benefits-focused)

9. Copy Examples for Key Channels

Tailor this to the channels the user said they care about most. If they didn’t specify, default to website + email + social.

a) Website / Landing Page Examples

- Hero section:

- 3–7 headline options

- 3–5 supporting subhead options

- 3–5 CTA button copy options

- Short “How it works” section copy (3 simple steps)

- Short credibility/proof block (e.g., how you’d phrase testimonials, logos, or data points if available or inferred)

b) Email / Outreach Examples

- 5–10 subject line options

- 2–3 short email opening lines/hooks tailored to the ideal customer’s pain

- 1 short email body example (outbound or nurture), using the defined Tone of Voice

c) Social / Ad Examples

- 5 hook/headline-style lines for social or ads

- 2–3 short post/ad examples (e.g., LinkedIn, Facebook, or similar), using the defined Tone of Voice

10. Implementation Notes & Next Steps

- Suggestions on where and how to use this messaging (e.g., website sections, email flows, sales scripts).

- Suggestions for what user could test (e.g., which headlines, which offers, which angles).

- Any risks, caveats, or important compliance considerations if relevant.
</STRUCTURE_OF_OUTPUT>

<STYLE_AND_CONSTRAINTS>
- Prioritize clarity over cleverness. Use simple, concrete language.

- Avoid generic buzzwords unless they are absolutely necessary for the audience.

- Use the customer’s words and perspective wherever possible (“you” focused).

- Make everything skimmable: short paragraphs, bullet points, clear headings.

- When creating examples, keep them realistic and specific to the given product and customer profile.

- Where assumptions are made, make them reasonable, and if important, briefly note them
</STYLE_AND_CONSTRAINTS>

<FINAL_INSTRUCTION>
When the user pastes product information into this prompt:

1. Ask your clarifying questions as described in <CLARIFYING_QUESTIONS>.

2. Once the user responds, generate the full Product Messaging Guide following <STRUCTURE_OF_OUTPUT> and <STYLE_AND_CONSTRAINTS>.

3. Ensure you include the dedicated “Tone of Voice Guide (Standalone Document)” section in every final guide.
</FINAL_INSTRUCTION>

In this AI Advent Calendar episode, you’ll learn how to create a Tone of Voice guide that makes ChatGPT and Claude write in your brand voice—not generic AI copy.  We’ll walk you through the prompt, what to include (voice pillars, writing principles, and do/don’t examples), and how to save it as a reusable .txt source file for consistent on-brand marketing content.    Watch the video to build your voice once—and reuse it everywhere.

Here’s the full prompt (copy and paste into your preferred LLM):

<ROLE>
You are an expert in branding, copywriting, and on-brand content systems.
</ROLE>

<ASSIGNMENT>
Your task is to create a comprehensive, practical Tone of Voice (TOV) document for a company.
</ASSIGNMENT>

<CONTEXT>
Here is the details on the company
- Company name: [ENTER COMPANY NAME HERE]
- Company website: [ENTER COMPANY URL HERE]
- [Optional: Add any additional brand description or strategy notes]

<ATTACHMENTS>
Attached are a set of writing samples that represent the desired tone of voice.
[Attach documents]
</ATTACHMENTS>

<REASONING>
Here is the priority of the the inputs you're being provided that should be used in developing the tone of voice:
1. Writing samples (highest priority, source of truth for voice).
2. User-provided brand description / strategy notes.
3. Company website content.

 If there are tensions between these sources:
- Assume the writing samples are the definitive reference for tone.
- Use the brand description to clarify intent and positioning.
- Use the website only to fill factual/contextual gaps or support positioning.
</REASONING>

<PROCESS>
This is how I'd like you to analyze the content provided.
1) Carefully read and analyze the writing samples first. From them, infer:
- Overall personality and emotional temperature (e.g., warm, authoritative, playful, formal, etc.).
- Typical sentence length and rhythm (short/ punchy vs. long/ flowing).
- Level of formality and complexity (reading level, jargon use, plain language vs. technical).
- Point of view (1st person “we/I”, 2nd person “you”, or 3rd person).
- Common rhetorical devices (metaphors, analogies, questions, repetition, numbered lists, emphatic punctuation, etc.).
- Attitude toward the reader (mentor, peer, expert, cheerleader, challenger, etc.).
- Distinctive verbal quirks (favorite phrases, structures, recurring patterns, transitions, types of headlines).

2) Use the brand description and website to:
- Clarify who the brand is, who they serve, and what they stand for.
- Align tone guidance with brand values, audience needs, and category norms.
- Add any missing context that helps position the voice (e.g., “premium but friendly,” “mass-market and accessible,” etc.).

Do NOT describe your step-by-step thought process; only output the final Tone of Voice document
<OUTLINE_OF_DOCUMENT>
Your output should follow this structure (adapt language as needed, but keep the logic): 

--------------------------------

A. Snapshot: Voice at a Glance

--------------------------------

Provide a concise overview of the voice:
- 1–2 short sentences that summarize the tone.
- 3–6 bullet points capturing the core qualities (e.g., “Direct but kind,” “Optimistic and practical,” “Expert, not academic”). 

This section should be something a writer or LLM can glance at in 10 seconds and “get” the voice.

--------------------------------

B. Brand & Audience Context (Brief)

--------------------------------

Summarize only what’s needed to anchor the voice (do not turn this into a full brand strategy deck):
- Who the brand is (1–3 sentences).
- Who the primary audience(s) are.
- What the brand promises or stands for.
- How the tone should make the audience feel.

Tie this directly to tone, e.g., “Because the audience is time-poor executives, the voice is concise and no-nonsense.”

--------------------------------

C. Voice Pillars / Personality Traits

--------------------------------

Identify 3–6 key voice pillars (e.g., “Plainspoken Expert,” “Encouraging Coach,” “Bold Challenger”).

For each pillar:
- Name: A short, memorable label.
- Description: 2–4 sentences explaining what this means in practice.
- “Sounds like”: 3–5 bullet points describing how this pillar shows up in language.
- “Does NOT sound like”: 3–5 bullet points describing what to avoid (contrast is crucial). 

Where helpful, root these in observations from the samples (e.g., “In the samples, the brand often uses short, decisive sentences and avoids hedging language.”) but keep the document written as guidance, not an analysis report.

--------------------------------

D. Core Writing Principles

--------------------------------

Spell out actionable rules that any writer or LLM should follow. Use bullets.

Cover items such as:
- Clarity vs. cleverness (which is more important for this brand?).
- Level of directness vs. diplomacy.
- How to handle complexity: explain, simplify, or lean into nuance?
- Use of storytelling vs. straight facts.
- How persuasive or “salesy” the tone should be.
- How much emotion vs. rational argument to use.
- How to handle authority and humility (e.g., “confident but never arrogant”). 

Make these rules specific and behavior-based, not vague (e.g., “Avoid filler like ‘in today’s world’ and ‘in many ways’”).

--------------------------------

E. Language & Style Guidelines

--------------------------------

Provide detailed guidance on how the voice appears in the actual words and sentences. Use headings and bullets where helpful.

Address:
- Reading level and complexity (e.g., “aim roughly at 8th–10th grade reading level unless technical detail is required”).
- Sentence length and structure (short, varied, long/rolling).
- Point of view (we/you/they) and how consistently it should be used.
- Tense preferences (present vs. past vs. future).
- Vocabulary:
- Words and phrases to prefer (e.g., “clients” vs. “customers”).
- Words and phrases to avoid.
- Approach to jargon (embrace, translate, or avoid).
- Grammar and punctuation:
- Oxford comma or not.
- Use of contractions.
- Use of exclamation points, ellipses, and ALL CAPS (if any).
- Tone markers:
- Use of questions.
- Use of metaphors or analogies (and what kind).
- Use of humor (yes/no, and what kind).
- Formatting nuances:
- Use of headings, subheadings, lists, and bold/italics for emphasis.
- Preferences for short paragraphs vs. dense blocks.
- Inclusivity & accessibility:
- Inclusive language expectations.
- Sensitivity to particular topics (if evident from context).
- Locale:
- Default language and locale (e.g., US English spelling) unless specified otherwise by the user. 

--------------------------------

F. Do / Don’t Examples

--------------------------------

Create concrete examples that show how to apply the tone. This section is critical for both humans and LLMs.

For each area where misuse is likely (e.g., too formal, too fluffy, too salesy), provide:
- A “Neutral or Wrong” version (label it clearly, e.g., “❌ Don’t”).
- A “On-Brand” version (label it clearly, e.g., “✅ Do”). 

Cover at least:
- A short website headline + subhead.
- A short social post.
- A short paragraph of blog/long-form copy.
- A simple call-to-action line. 

Make sure the “Do” examples closely mirror the patterns you observed in the writing samples (rhythm, phrasing, level of energy, etc.).

--------------------------------

G. Channel-Specific Guidance

--------------------------------

Explain how the core voice stays consistent but flexes across key channels.

For each relevant channel, provide 3–5 bullets:
- Website / landing pages.
- Blog / articles / thought leadership.
- Social media posts.
- Email (newsletters, announcements, nurturing).
- Performance/Ad copy (if relevant).
- Other channels specific to this brand (if evident).

For each channel, specify:
- Typical length and pace.
- How direct vs. conversational.
- How much context vs. punch.
- Any specific formats (e.g., “Use strong hooks in the first sentence for LinkedIn posts,” “Lead with outcome in ad headlines”).

--------------------------------

H. Structural & Formatting Preferences

--------------------------------

Provide guidance for how content should generally be structured, especially for LLMs generating longer copy: 
- How to open (e.g., “Start with the reader’s problem,” “Lead with an outcome,” “Open with a short, vivid line”).
- Preferred use of subheadings, bullet lists, and section breaks.
- Whether to include summaries, key takeaways, or CTAs at the end—and how they should sound.
- Any patterns observed in the samples (e.g., “Often uses 1–2 sentence paragraphs to emphasize key lines”). 

--------------------------------

I. Invitation to Refine

--------------------------------

Close with a short note inviting feedback and clarifications from the user, such as:

“If any part of this Tone of Voice description feels off, too narrow, or incomplete based on how you see the brand, please highlight:
- Which section or example feels misaligned.
- How you would describe the desired tone instead.

I can then help refine and tighten this document so it more perfectly reflects your brand’s true voice.”

<OUTPUT_FORMAT>
- Use clear headings, subheadings, and bullet points. Make the document easy to scan and practical for everyday use and for LLMs.
- Provide ONE cohesive Tone of Voice document that follows the structure above (you may rename headings but must preserve the intent).
- Use clear headings, subheadings, and bullet points.
- Include multiple concrete “Do/Don’t” examples modeled closely on the provided writing samples.
- Do not refer to yourself or describe your reasoning steps.
- Do not restate the prompt or instructions; only output the finished Tone of Voice document.
- At the end of your document, explicitly invite the user to flag any major conflicts between how you’ve described the tone and how they see the brand, so they can refine it.
</OUTPUT_FORMAT>