Discover Brand

Execution Steps

Step 1: Define Discovery Scope

• Clarify the scope of the discovery with the user:
  • Full discovery: Search everywhere accessible for all brand-related content.
  • Platform-specific: Focus on a particular channel (LinkedIn, website, email templates, proposals).
  • Time-bound: Only look at content from a specific date range.
  • Content-type focused: Only look at a specific type (social posts, emails, website pages, sales materials).
• Identify available sources:
  • Local files: Documents, presentations, templates in the workspace.
  • Connected platforms: Email (Gmail), design tools (Canva, Figma), CRM, marketing platforms (Klaviyo).
  • Web presence: Company website, LinkedIn profiles, social media accounts.
  • Third-party mentions: Press coverage, partner content, directory listings.
• Confirm which sources the user wants searched and which are accessible via connected tools.

Step 2: Load Brand Framework (Layaa AI Mode)

• Read all reference files to establish the baseline for comparison:
  • domain-references/brand-voice/tone-framework.md -- The rules content should follow.
  • domain-references/brand-voice/founder-voice.md -- Individual voice profiles for attribution check.
  • domain-references/brand-voice/content-templates.md -- Channel formatting standards.
  • shared-references/company-identity.md -- Core messaging and positioning.
  • shared-references/icp-and-market.md -- Target audience definition.
• Build a checklist of brand rules to evaluate each discovered piece of content against.
• In General mode: Skip this step. Discovery will catalog materials without evaluating against a framework (unless the user provides guidelines separately).

Step 3: Search Local Files

• Use Glob to search for brand-related files in the workspace:
  • **/*.md -- Markdown content files.
  • **/*.txt -- Text documents, prompts, guidelines.
  • **/*.docx -- Word documents (note: may need conversion).
  • **/*.pdf -- Presentations, proposals, collateral.
• Use Grep to search file contents for brand-relevant terms:
  • Company name and variations.
  • Founder names.
  • Product/service names.
  • Key messaging phrases from the tone framework.
  • Taglines and positioning statements.
• Catalog each discovered file with:
  • File path.
  • Content type (proposal, social post, email template, website copy, internal doc).
  • Date modified (from file metadata).
  • Brief content summary (first 100 words or key heading).
  • Attribution (if identifiable -- which founder or team member).

Step 4: Search Connected Platforms

• Based on available connected tools, search each platform:
  • Gmail: Search for sent emails containing brand content, templates, outreach.
  • Canva: Search designs for branded materials, social media posts, presentations.
  • Figma: Search for brand design assets, UI copy, marketing materials.
  • Klaviyo: Search email campaigns, templates, automations for brand content.
  • Calendar: Check for events with brand-related descriptions or agendas.
• For each platform, catalog found materials the same way as local files.
• Note which platforms were searched and which were unavailable or returned no results.

Step 5: Search Web Presence

• Use WebSearch to find the brand's public content:
  • Company website pages (home, about, services, blog).
  • LinkedIn company page posts and articles.
  • Founder LinkedIn profiles and posts.
  • Social media accounts (Twitter/X, Instagram, YouTube).
  • Press mentions and third-party articles.
  • Review sites, directory listings, partner pages.
  • Job postings (these reflect employer brand voice).
• Use WebFetch to retrieve and analyze key pages.
• Catalog each web source with URL, content type, date (if available), and content summary.
• In General mode: This may be the primary discovery channel if no local files or connected platforms are available.

Step 6: Categorize Discovered Materials

• Organize all discovered content into categories:

## Discovery Inventory

### By Channel
- LinkedIn: [count] pieces
- Website: [count] pages
- Email: [count] templates/campaigns
- Proposals: [count] documents
- Internal: [count] documents
- Other: [count] items

### By Content Type
- Thought leadership / educational
- Sales / promotional
- Operational / internal
- Brand assets / design
- Legal / compliance

### By Attribution
- Founder 1 (Abhimanyu Singh): [count] pieces
- Founder 2 (Shubham Sharma): [count] pieces
- Company (unattributed): [count] pieces
- Third-party: [count] pieces

### By Recency
- Last 30 days: [count]
- Last 90 days: [count]
- Last 6 months: [count]
- Older: [count]

Step 7: Evaluate Voice Consistency (Layaa AI Mode)

• For each discovered piece of content, evaluate against the brand framework:
  • Tone alignment: Does it match the core voice attributes (confident, intelligent, approachable, action-oriented)?
  • Vocabulary compliance: Does it contain banned terms? Does it use preferred vocabulary?
  • Channel formatting: Does it follow channel-specific rules from the content templates?
  • ICP alignment: Does it speak to the right audience with relevant pain points?
  • Founder voice accuracy: If attributed to a founder, does it match their voice profile?
• Assign each piece a consistency rating:
  • Strong (8-10): Fully aligned with brand framework. Minor or no issues.
  • Moderate (5-7): Generally on-brand but with notable deviations. Fixable with light editing.
  • Weak (2-4): Significant misalignment. Needs substantial revision.
  • Off-brand (0-1): Does not represent the brand voice at all. Consider removal or complete rewrite.
• In General mode: Without a framework to compare against, focus on internal consistency. Note patterns and variations across content without scoring.

Step 8: Identify Drift Areas

• Analyze the consistency ratings to identify patterns of drift:
  • Channel drift: Is one channel consistently weaker than others? (e.g., social media on-brand but proposals off-brand)
  • Temporal drift: Has voice consistency changed over time? Is recent content more or less aligned?
  • Attribution drift: Does one person's content drift more than another's?
  • Topic drift: Are certain subject areas more prone to voice inconsistency?
  • Vocabulary drift: Are banned terms creeping back in? Are preferred terms being abandoned?
• For each drift area, identify:
  • The specific pattern of deviation.
  • Likely cause (new team member writing, time pressure, channel unfamiliarity, guideline gaps).
  • Impact severity (how much this affects brand perception).

Step 9: Generate Recommendations

• Based on the drift analysis, produce actionable recommendations:
  • Immediate fixes: Content that should be revised now (high-visibility, off-brand pieces).
  • Template updates: Channel templates that need refreshing or creating.
  • Guideline gaps: Areas where the brand framework needs new rules or clearer rules.
  • Training needs: Patterns suggesting someone needs a voice refresher.
  • Content opportunities: Channels or content types where the brand is underrepresented.
  • Archival candidates: Old content that no longer represents the brand and should be retired.
• Prioritize recommendations as High / Medium / Low based on:
  • Audience reach (how many people see this content).
  • Brand impact (how much it affects perception).
  • Effort to fix (quick edit vs. full rewrite).

Step 10: Compile Discovery Report

• Produce the final discovery report:

# Brand Discovery Report -- [Brand Name]
**Date:** [current date]
**Scope:** [what was searched]
**Sources Searched:** [list of platforms and locations]

## Executive Summary
- [2-3 sentence overview of findings]
- Overall voice consistency score: [X/10]
- Materials discovered: [total count]
- Items needing attention: [count]

## Discovery Inventory
[From Step 6 -- categorized inventory]

## Voice Consistency Analysis
### Consistency Distribution
| Rating | Count | Percentage |
|--------|-------|------------|
| Strong (8-10) | ... | ...% |
| Moderate (5-7) | ... | ...% |
| Weak (2-4) | ... | ...% |
| Off-brand (0-1) | ... | ...% |

### Channel Breakdown
| Channel | Avg. Score | Strongest Area | Weakest Area |
|---------|-----------|----------------|--------------|
| LinkedIn | ... | ... | ... |
| Website | ... | ... | ... |
| Email | ... | ... | ... |
| Proposals | ... | ... | ... |

## Drift Analysis
### Key Drift Areas
1. [Drift area with evidence and severity]
2. [Drift area with evidence and severity]
3. [Drift area with evidence and severity]

## Recommendations
### High Priority
- [ ] [Recommendation with specific content reference]
### Medium Priority
- [ ] [Recommendation]
### Low Priority
- [ ] [Recommendation]

## Detailed Findings
[Per-item analysis for items scoring below 7, with specific issues noted]

## Appendix: Materials Catalog
[Complete list of all discovered materials with metadata]

Step 11: Flag for Founder Review

• Highlight any findings that require founder attention:
  • Content that misrepresents the company or founders.
  • Third-party content with inaccurate claims about the brand.
  • Significant voice drift that may indicate a strategic shift (intentional or not).
  • Guideline gaps that only founders can resolve (positioning decisions, new market messaging).
• Present these as a separate, concise section at the top of the report for quick founder review.

Output Format

Deliver the complete discovery report in Markdown format. Structure it for easy scanning:

1. Executive Summary -- The top-level findings in 3-5 sentences. Founders should be able to read just this and know the state of brand voice.
2. Full Report -- The detailed report from Step 10.
3. Action Items -- Extracted as a standalone checklist, prioritized and assignable.

If the report is extensive, offer to save it as a file using the Write tool.

Edge Cases

• No connected platforms available: Focus on local files and web search. Note the limitation in the report and recommend connecting relevant platforms for a more complete audit.
• No existing brand framework (General mode): Produce a discovery-only report without consistency scoring. Instead, identify the emergent voice patterns and recommend creating formal guidelines (suggest the generate-guidelines skill).
• Very large content volume: If more than 50 pieces of content are discovered, prioritize evaluation of high-visibility content (website home page, recent LinkedIn posts, active email campaigns). Provide summary statistics for the rest. Offer to do a deep-dive on specific channels in a follow-up.
• Mixed-brand content: If discovered content serves multiple brands or sub-brands, separate the analysis by brand. Note which brand each piece belongs to.
• Stale content: If significant amounts of content are outdated (more than 12 months old), flag this separately. Stale content can still appear in search results and shape brand perception.
• Third-party content inaccuracies: If web search reveals third-party content with incorrect information about the brand, flag these for correction outreach. Include the URL and the specific inaccuracy.
• Platform access failures: If a connected platform search fails or times out, note it in the report. Do not silently skip it. Suggest the user retry or check permissions.