Enrichment Design

Data Points Builder

Bridge the gap between research hypotheses and table enrichment. Define WHAT to research about each company before running enrichment.

When to Use

• After market-research has produced a hypothesis set
• Before list-enrichment — this skill designs the columns, that skill runs them
• When the user says "what should we research about these companies?"

Two Modes

Mode 1: Segmentation

Goal: Design columns that score or confirm hypothesis fit per company.

Input: Hypothesis set (from market-research or context file)

Process: 1. Read the hypothesis set 2. For each hypothesis, propose 1-2 columns that would confirm or deny fit 3. Discuss with user — refine, add, remove 4. Output final column_configs

Example: If hypothesis is "Database blind spot — 80-90% of targets invisible to standard tools":

• Column: "Data Infrastructure Maturity" (select: ["No CRM", "Basic CRM", "Full stack"])
• Column: "Digital Footprint Score" (grade: 1-5)

Mode 2: Personalization

Goal: Design columns that capture company-specific hooks for email personalization.

Input: Target list + what the user wants to personalize on

Process: 1. Ask what hooks matter for this campaign (leadership quotes, recent launches, hiring signals, tech stack, etc.) 2. Propose 2-4 columns with prompts 3. Discuss with user — refine 4. Output final column_configs

Example: For personalization hooks:

• Column: "Recent Product Launch" (text: describe any product launched in last 6 months)
• Column: "Leadership Public Statement" (text: find a public quote from CEO/CTO about [topic])

Interactive Column Design

Do NOT just generate columns silently. Walk through this with the user:

Step 1: Present the framework

Show the user the two modes and ask which applies (or both).

Step 2: Propose initial columns

Based on hypotheses or user input, propose 3-5 columns. For each, show:

Column: [name]
Type: [output_format]
Agent: [research_pro | llm]
Prompt: [the actual prompt text]
Why: [what this tells us for segmentation/personalization]

Step 3: Refine together

Ask:

• "Any columns to add?"
• "Any to remove or merge?"
• "Should any prompts be more specific?"

Step 4: Confirm column budget

Guidance:

• 3-5 columns is the sweet spot
• 6-7 is acceptable if each serves a clear purpose
• 8+ adds noise — push back and suggest merging

Step 5: Output column_configs

Generate the final column configs as a JSON array ready for list-enrichment:

[
  {
    "kind": "agent",
    "name": "Column Display Name",
    "key": "column_key_snake_case",
    "value": {
      "agent_type": "research_pro",
      "prompt": "Research prompt using {input} for domain...",
      "output_format": "text"
    }
  }
]

Column Design Guidelines

Agent Type Selection

Output Format Selection

Prompt Writing Tips

• Always include {input} for the company domain
• Be specific about output format in the prompt itself
• Include fallback: "If not found, return N/A" or "If unclear, return 'Unknown'"
• For select/multiselect: list the labels in the prompt too
• For hypothesis scoring: reference the specific hypothesis in the prompt
• Keep prompts under 200 words

Reference Library

See references/data-point-library.md for ~20 pre-built column configs organized by use case.

Output Handoff

After column design is complete: 1. Present the final column_configs JSON to the user 2. Tell the user: "These configs are ready for list-enrichment. Run that skill with your table ID and these columns." 3. If the user wants to run immediately, hand off to list-enrichment workflow

Data Point Library

Pre-built column configs organized by use case. Copy and customize for your campaign.

Segmentation Columns

Company Size & Stage

{
    "kind": "agent", "name": "Employee Count Bucket", "key": "employee_bucket",
    "value": {
        "agent_type": "llm",
        "prompt": "Based on the company profile for {input}, classify employee count into one of these buckets. Return ONLY the label.",
        "output_format": "select",
        "labels": ["1-10", "11-50", "51-200", "201-500", "500+"]
    }
}

{
    "kind": "agent", "name": "Funding Stage", "key": "funding_stage",
    "value": {
        "agent_type": "research_pro",
        "prompt": "Find the latest funding round for {input}. Return the stage name only (e.g., Seed, Series A, Series B). If bootstrapped or unknown, return 'Bootstrapped'. If public, return 'Public'.",
        "output_format": "select",
        "labels": ["Bootstrapped", "Seed", "Series A", "Series B", "Series C+", "Public"]
    }
}

Technology & Infrastructure

{
    "kind": "agent", "name": "CRM Platform", "key": "crm_platform",
    "value": {
        "agent_type": "research_pro",
        "prompt": "What CRM or sales platform does {input} use? Look for evidence in job postings, case studies, or tech stack databases. Return the platform name (e.g., Salesforce, HubSpot, Pipedrive). If unknown, return 'Unknown'.",
        "output_format": "text"
    }
}

{
    "kind": "agent", "name": "Tech Stack Maturity", "key": "tech_maturity",
    "value": {
        "agent_type": "research_reasoning",
        "prompt": "Assess the technology maturity of {input}. Consider: do they have a modern website? Do they use cloud services? Do job postings mention modern tools? Rate 1 (minimal tech) to 5 (advanced).",
        "output_format": "grade"
    }
}

Market Position

{
    "kind": "agent", "name": "Geographic Coverage", "key": "geo_coverage",
    "value": {
        "agent_type": "research_pro",
        "prompt": "How many locations, offices, or markets does {input} operate in? Classify as: Local (1 city), Regional (1 state/region), National (multiple states), International. Return ONLY the label.",
        "output_format": "select",
        "labels": ["Local", "Regional", "National", "International"]
    }
}

{
    "kind": "agent", "name": "Primary Vertical", "key": "primary_vertical",
    "value": {
        "agent_type": "llm",
        "prompt": "Based on the company profile for {input}, what is their primary industry vertical? Return a specific label, not a broad category. Examples: 'SaaS', 'Logistics', 'Healthcare', 'Manufacturing'. Return one label only.",
        "output_format": "text"
    }
}

Personalization Columns

Leadership & Public Signals

{
    "kind": "agent", "name": "CEO Public Statement", "key": "ceo_statement",
    "value": {
        "agent_type": "research_pro",
        "prompt": "Find a recent public statement, interview, blog post, or LinkedIn post by the CEO or founder of {input}. Focus on their views about [TOPIC]. Return a 1-2 sentence summary with the source URL. If nothing found, return 'No public statements found'.",
        "output_format": "text"
    }
}

{
    "kind": "agent", "name": "Recent News", "key": "recent_news",
    "value": {
        "agent_type": "research_pro",
        "prompt": "Find the most significant news about {input} from the last 6 months. Focus on: product launches, partnerships, funding, leadership changes, or expansion. Return a 1-sentence summary with date. If nothing found, return 'No recent news'.",
        "output_format": "text"
    }
}

Hiring Signals

{
    "kind": "agent", "name": "Key Hiring Signals", "key": "hiring_signals",
    "value": {
        "agent_type": "research_pro",
        "prompt": "Check if {input} is currently hiring for roles related to [FUNCTION]. Look at their careers page and job boards. Return: role title(s) and what it signals about their priorities. If no relevant openings, return 'No relevant openings'.",
        "output_format": "text"
    }
}

Competitive Context

{
    "kind": "agent", "name": "Current Tools", "key": "current_tools",
    "value": {
        "agent_type": "research_pro",
        "prompt": "What tools or platforms does {input} currently use for [FUNCTION]? Look in job postings, case studies, tech directories, and G2 reviews. Return tool names. If unknown, return 'Unknown'.",
        "output_format": "text"
    }
}

Hypothesis Scoring Columns

Template: Binary Hypothesis Match

{
    "kind": "agent", "name": "Hypothesis: [NAME]", "key": "hyp_[key]",
    "value": {
        "agent_type": "research_reasoning",
        "prompt": "Evaluate whether {input} matches this hypothesis: '[HYPOTHESIS DESCRIPTION]'. Look for evidence that confirms or denies this. Return a JSON object.",
        "output_format": "json",
        "output_schema": {
            "type": "object",
            "properties": {
                "match": {"type": "boolean"},
                "confidence": {"type": "string", "enum": ["high", "medium", "low"]},
                "evidence": {"type": "string"}
            }
        }
    }
}

Template: Hypothesis Fit Score

{
    "kind": "agent", "name": "Fit: [HYPOTHESIS]", "key": "fit_[key]",
    "value": {
        "agent_type": "research_reasoning",
        "prompt": "Rate how well {input} fits this hypothesis on a 1-5 scale: '[HYPOTHESIS]'. 1 = no fit, 3 = possible fit, 5 = strong fit. Consider: [SPECIFIC CRITERIA]. Return ONLY a number.",
        "output_format": "grade"
    }
}

Combo Patterns

Segmentation + Personalization (5-column setup)

1. Primary Vertical (select) — segmentation 2. Company Size Bucket (select) — segmentation 3. Hypothesis Fit Score (grade) — segmentation 4. Recent News (text) — personalization 5. Leadership Statement (text) — personalization

Pure Segmentation (4-column setup)

1. Primary Vertical (select) 2. Tech Maturity (grade) 3. Geographic Coverage (select) 4. Hypothesis Match (json with match + evidence)

Pure Personalization (3-column setup)

1. CEO Public Statement (text) 2. Recent Product Launch (text) 3. Current Tools for [function] (text)