Skip to main content

Share of Model Agent in Pencil - User Guide

The Share of Model (SOM) Agent inside Pencil is a specialized agent designed to analyze brand perception across different Large Language Models (LLMs) and provide actionable insights for improving brand positioning and content creation.

Ishita Mishra avatar
Written by Ishita Mishra
Updated over a month ago

What the Agent Does

The Share of Model Agent provides three primary capabilities:

  1. Brand Perception Insights: Analyzes how your brand is perceived across different LLMs (Claude, Gemini, GPT, Deepseek, etc.

  2. Improvement Recommendations: Provides insights on how to improve brand perception based on data analysis

  3. Content Optimization Guidance: Offers insights on improving content (images, PDP, etc.) for specific LLMs

Once insights are obtained, you can switch to specialized content generation agents (Image Generation, Video Generation, etc.) to create the recommended assets.

Data Sources

The agent uses data from Share of Model analyses executed on the . For each analysis, the agent

has access to the following data:

Share of Model Application

Brand Perception Data

  • Brand Perception Attributes: Each attribute represents either a positive (strength) or negative (weakness) perception of the brand

  • Brand Perception Cluster Groups: Groups of similar attributes that cluster together, helping identify thematic patterns in brand perception

  • Brand Perception by Source: Detailed perception data (strengths and weaknesses) broken down by each LLM source (Claude, Gemini, GPT, etc.)

  • Brand Perception by Brand: Comparative analysis showing your brand’s perception alongside competitors

Target Audience Data

  • Audience Segments: Detailed audience profiles including:

    • Demographics (gender, age range, area type, household type)

    • Behavioral attributes (challenges, decision-making patterns, purchase influences)

    • Media consumption (media channels, preferred sources)

  • Age Range Distribution: Statistical distribution of target audience across different age ranges

Analysis Metadata

  • Brand Information: The brand being analyzed

  • Countries: Geographic scope of the analysis

  • Category: Product/service category

  • Competitors: List of competitor brands included in the analysis

  • Personas: Target personas defined for the analysis

  • Sources: LLM providers included in the analysis (e.g., Anthropic, Google, OpenAI, Deepseek)

Getting Started

Prerequisites

When starting to use the Share of Model Agent, you need to:

  1. Select a Brand: The agent will list all available brands. If only one brand is available, it will be selected automatically.

  2. Select an Analysis: Choose a specific Share of Model analysis to work with. If only one analysis exists, it will be selected automatically.

Initial Setup Workflow

  1. When you first interact with the agent:

  2. The agent will automatically list available brands

  3. If multiple brands exist, you’ll be prompted to select one

  4. The agent will then list available analyses for the selected brand using

  5. If multiple analyses exist, you’ll be prompted to select one

  6. Once selected, the analysis is stored in context using

Key Capabilities

1.Brand Perception Analysis

The agent can analyze brand perception across multiple dimensions:

  • Overall Cluster Share: Understand the distribution of perception clusters (strengths and weaknesses)

  • Cluster Share by Brand: Compare your brand’s perception clusters against competitors

  • Cluster Share by Source: See how different LLMs perceive your brand differently

  • Attributes Hierarchy: Explore the detailed hierarchy of attributes within clusters

Example Queries:

  • “What are the main strengths and weaknesses of my brand?”

  • “How does my brand compare to competitors in terms of perception?”

  • “Which LLM sources have the most positive perception of my brand?”

2.Source-Specific Insights

Analyze how your brand is perceived on specific LLM platforms:

  • Identify which sources have the most favorable perception

  • Understand source-specific strengths and weaknesses

  • Get recommendations for optimizing content for specific LLM platforms

Example Queries:

  • “What are my brand’s strengths on Claude vs Gemini?”

  • “How should I optimize my content for GPT?”

3. Audience Analysis

Understand your target audience segments:

  • Demographic breakdowns

  • Behavioral patterns and preferences

  • Media consumption habits

  • Age distribution analysis

Example Queries:

  • “Who is my target audience?”

  • “What are the main challenges of my audience segments?”

  • “What media channels does my audience prefer?”

4. ContentImprovement Recommendations

Based on the analysis, the agent provides actionable recommendations:

  • Specific attributes to emphasize in content

  • Content themes that align with brand strengths

  • Areas to address to improve weaknesses

  • LLM-specific content optimization strategies

Example Queries:

  • “How can I improve my brand perception through content?”

  • “What should my next image campaign focus on?”

  • “What messaging should I use for my PDP?”

Integration with Content Generation Agents

The Share of Model Agent is designed to work in conjunction with content generation agents:

Typical Workflow

  • Analysis Phase (Share of Model Agent):

    • Analyze brand perception

    • Identify strengths and weaknesses

    • Get content recommendations

  • Content Generation Phase (Specialized Agents):

    • Use insights to create image campaigns

    • Generate video content aligned with brand strengths

    • Create optimized PDPs

    • Develop email copy and other marketing materials

Example : Creating Assets

User: "I want to create an image campaign to improve my brand perception"

1. Share of Model Agent analyzes:

- Identifies key strengths: "innovative", "sustainable", "premium quality"

- Identifies weaknesses: "price perception", "accessibility"

- Recommends: Focus on innovation and sustainability in visuals

2. User switches to Image Generation Agent:

- Uses insights from SOM Agent

- Generates images emphasizing innovation and sustainability

- Creates content that addresses accessibility concerns

Example: Improving Existing Assets

User: "I have an existing product image. How can I improve it based on brand perception?"

1. Share of Model Agent analyzes:

- Reviews current brand perception across LLM sources

- Identifies that "premium quality" is a strength but not well represented

- Finds that "accessibility" is a weakness that needs addressing

- Recommends: Enhance visual elements to emphasize premium quality while making the product appear more

accessible

2. User switches to Image Generation Agent with existing image:

- Provides the existing product image as input

- Uses SOM insights to guide improvements

- Agent edits/regenerates the image with:

- Enhanced premium quality visual cues (better lighting, refined textures)

- More accessible presentation (clearer product visibility, approachable styling)

- Creates an improved version that better aligns with brand perception goals

Best Practices

  1. Start with Broad Analysis
    Begin with overall brand perception before diving into specific dimensions.

  2. Compare Across Sources
    Always compare how different LLMs perceive your brand to identify platform-specific opportunities.

  3. Use Competitive Context
    Leverage competitor comparisons to understand your relative positioning.

  4. Link Insights to Content
    Always connect insights to actionable content recommendations.

  5. Iterate Based on Results
    Use the agent’s recommendations to create content, then analyze the impact and iterate.

Response Format

The agent provides:

  • Executive Summary: High-level insights and key findings

  • Detailed Analysis: Breakdown of data with specific metrics

  • Actionable Recommendations: Clear next steps for content creation

  • Context: Information about data sources, time periods, and limitations

All shares and percentages are expressed as percentages for clarity.

Language Support

The agent can translate insights and recommendations into your desired language. Simply request translation when needed.

Error Handling

If data is unavailable or incomplete, the agent will:

  • Transparently report limitations

  • Suggest alternative approaches

  • Continue analysis with available data while noting gaps

  • Estimate the impact of missing data on conclusions

API Documentation

For detailed API information, refer to the Share of Model API Documentation.

Terminology

  • Strengths: Positive brand perception attributes (preferred term over “pros”)

  • Weaknesses: Negative brand perception attributes (preferred term over “cons”)

  • Cluster Groups: Thematic groupings of similar attributes

  • Source: LLM provider (e.g., Claude, Gemini, GPT)

  • Analysis: A Share of Model analysis containing brand perception data

  • Persona: Target audience persona defined in the analysis

Example Use Cases

Use Case 1: Brand Perception Audit

User: "Analyze my brand's perception across all LLM sources"

Agent Response:

- Lists strengths and weaknesses

- Compares perception across different sources

- Provides percentage breakdowns

- Recommends focus areas for improvement

Use Case 2: Competitive Analysis

User: "How does my brand compare to competitors?"

Agent Response:

- Cluster share comparison by brand

- Identifies competitive advantages

- Highlights areas where competitors outperform

- Suggests differentiation strategies

Use Case 3: Content Strategy Development

User: "What content should I create to improve perception on Gemini?"

Agent Response:

- Analyzes brand perception on Gemini specifically

- Identifies key attributes to emphasize

- Provides content themes and messaging recommendations

- Suggests switching to Image/Video Generation Agent for execution

Use Case 4: Audience-Targeted Content

User: "What content resonates with my target audience?"

Agent Response:

- Analyzes audience segments

- Identifies key challenges and preferences

- Recommends content themes aligned with audience needs

- Provides media channel recommendations

Support and Troubleshooting

If you encounter issues:

  1. No Analysis Available: Ensure a Share of Model analysis has been created and is accessible

  2. Missing Data: The agent will report data limitations and suggest alternatives

  3. API Errors: Check the Share of Model API documentation for status and requirements

Did this answer your question?