ROI Auto-Generation

This drill creates an always-on n8n workflow that monitors Attio for deals reaching the proposal stage with sufficient pain data, then automatically generates ROI calculators and business cases. The seller reviews and sends — no manual model building required.

Input

• Attio pipeline configured with deal stages including Proposed
• Pain data populated on deals from pain-discovery-call drill
• n8n instance with Attio and Anthropic integrations configured
• Product pricing stored in .gtm-config.json or n8n environment variables

Steps

1. Build the trigger workflow in n8n

Create an n8n workflow with an Attio webhook trigger:

• Trigger: Attio deal stage changes to "Proposed" OR pain_count attribute updated on a deal in Proposed stage
• Filter: Only proceed if pain_count >= 2 AND pain_quantification_rate >= 0.5 AND roi_model_status != "generated"
• Rate limit: Maximum 10 auto-generations per day to control API costs

2. Enrich and validate pain data

When the trigger fires:

1. Pull the full deal record from Attio including all pain data, champion info, and company enrichment
2. For each pain where confidence < 0.6, re-run pain-quantification-prompt with any new context
3. Validate that pain_to_price_ratio >= 3 after re-quantification
4. If ratio < 3, skip auto-generation and send Slack alert: "Deal {name} has weak ROI ({ratio}x). Consider additional discovery before proposal."

3. Generate the ROI model

Run roi-model-generation with the validated pain data. This produces the full structured model with inputs, assumptions, savings breakdown, costs, summary, and sensitivity analysis.

Store the model in Attio as a note on the deal record.

4. Generate the business case document

Run business-case-generation with the ROI model output plus prospect context. This produces the champion-ready business case: executive summary, current state with prospect quotes, proposed solution, ROI analysis, risk analysis, alternatives comparison, and recommendation.

5. Package and notify

Build the deliverable:

1. Generate a Google Sheet ROI calculator (via Sheets API) pre-populated with the prospect's inputs
2. Generate a PDF business case (via Puppeteer or Markdown)
3. Attach both to the Attio deal record
4. Send a Slack notification to the deal owner: "ROI calculator and business case auto-generated for {company}. ROI: {X}x, payback: {Y} months. Review and send: {link}"

6. Track auto-generation metrics

Fire PostHog events:

{
  "event": "roi_auto_generated",
  "properties": {
    "deal_id": "...",
    "trigger": "stage_change|pain_update",
    "roi_percentage": 0,
    "payback_months": 0,
    "pain_count": 0,
    "generation_time_seconds": 0,
    "skipped": false,
    "skip_reason": null
  }
}

Track downstream: when the seller sends the auto-generated materials, fire roi_calculator_presented with auto_generated: true to compare conversion rates of auto-generated vs manually built ROI materials.

7. Handle edge cases

• Missing champion data: Generate the ROI model but flag the business case as incomplete (no champion perspective). Alert seller to identify champion before sending.
• Competitor mentioned in deal: Include competitive positioning in the alternatives comparison section of the business case.
• Very high ROI (>20x): Flag for human review. Extremely high projections may indicate inflated pain data or missing cost factors. Add a warning note.
• Generation failure: If Claude API returns an error or invalid JSON, retry once. On second failure, alert the seller and log the error.

Output

• Auto-generated ROI calculator (Google Sheet)
• Auto-generated business case (PDF)
• Both attached to the Attio deal record
• Seller notified via Slack with review link
• PostHog events tracking auto-generation volume and downstream effectiveness

Triggers

Runs continuously as an n8n workflow. Fires on Attio deal stage changes and pain data updates. No manual intervention needed for the generation — seller reviews before sending.