Pipedrive Deal Handoff Brief

The Analyst synthesizes: (1) Relationship History — interaction timeline, communication patterns, rapport indicators from activities and notes. (2) Deal Context — stage progression with durations, pipeline velocity, key milestones. (3) Commitments Made — explicit and implicit promises extracted from notes, classified as fulfilled, pending, or overdue. (4) Objections Encountered — objections from deal history, classified as resolved or open, with resolution approaches. (5) Competitive Landscape — competitor mentions from notes/activities, positioning used, evaluation criteria. (6) Recommended Actions — prioritized next steps for the first 48 hours with urgency classification.

The Research Gate checks two conditions: INCLUDE_WEB_RESEARCH must be true (default) AND the deal organization name must be populated. When both are met, the Researcher uses Sonnet 4.6 with web_search to gather company context (news, market position, developments) and person context (professional background). Set INCLUDE_WEB_RESEARCH=false in the Config Loader to disable web research and rely on CRM data only.

ITP-measured: $0.08–$0.18/handoff depending on deal complexity and whether web research is enabled. Average across 20 ITP records: $0.13/handoff. Deals with web research enabled cost more due to web_search API calls. Deals with INCLUDE_WEB_RESEARCH=false cost less. The Fetcher (code-only) adds zero LLM cost.

Different trigger and purpose. DIA explains stage moves — it fires when a deal changes stage and analyzes why. PDHB fires when a deal changes owner and generates a comprehensive handoff brief. DIA provides stage transition intelligence; PDHB provides ownership transition knowledge transfer. They complement each other in Pipedrive-centric sales operations.

The Fetcher calls 6 Pipedrive API endpoints: (1) GET /deals/{id} for deal properties (title, value, stage, dates, owner). (2) GET /deals/{id}/activities for all activities (calls, meetings, emails, tasks). (3) GET /deals/{id}/notes for all deal notes. (4) GET /deals/{id}/flow for stage history with timestamps. (5) GET /persons/{id} for primary contact details. (6) GET /organizations/{id} for company information.

The Formatter generates: (1) Notion handoff brief — a detailed page with executive summary, section-by-section content, and metadata footer. (2) Slack notification — Block Kit message with key context, priority actions, open commitment/objection counts, and a link to the Notion brief. (3) Pipedrive deal note — compact summary attached to the deal record. The Pipedrive note is optional via PIPEDRIVE_NOTE setting.

The Fetcher pre-assembles all deal data from 6 Pipedrive API calls. The Analyst organizes and synthesizes pre-assembled CRM data into a structured handoff brief — classification-tier reasoning. Sonnet 4.6 handles the 6-section synthesis with high accuracy in ITP testing. Opus would add significant cost per handoff without measurable quality improvement. SINGLE-MODEL keeps cost at $0.13/handoff average.

Yes. Send a POST to the webhook URL with _is_itp: true and deal fixture data in the payload. ITP mode bypasses Pipedrive API calls and web research, using fixture data directly. The 20 ITP test records cover rich deals, sparse deals, web research on/off, no-org deals, competitor mentions, and edge cases.

Conditionally. When INCLUDE_WEB_RESEARCH=true (default) and the deal has an organization name, the Researcher uses Anthropic web_search to gather company and person context. This is configurable — set INCLUDE_WEB_RESEARCH=false to disable all web research. Non-research data comes from Pipedrive API, Notion API, and Slack API.

All sales are final after download. Review the Blueprint Dependency Matrix and prerequisites before purchase. Questions? Contact support@forgeworkflows.com before buying. Full terms at forgeworkflows.com/legal.