Partner GTM Engine

Partners send account lists in inconsistent formats. Domain is often missing even though it is the matching key — so ownership, whitespace, and conflicts stay invisible.

There is no reliable view of who owns what across partners and the internal book, or which unclaimed accounts are worth pursuing.

Without this: Co-sell stays goodwill. Conflicts have no partner context. Whitespace is not scored.

Primary: AE/SDR reviewing their own territory and whitespace (v1 is a personal view, not an admin console)

Secondary: Aligned partners receiving filtered exports of their own accounts only

Job: Ingest sheets, classify ownership by domain, score whitespace, and keep partner relationship context on every conflict.

Metadot channel program: distribution sales +25% to $3.4M (2017) and +33% to $4.5M (2018). The v1 product is Channel Territory Mapping (secretsafari): column mapping per partner, domain as sole source of truth, Exa + human review when domain is missing — never an LLM guessing a domain from memory.

Goals

• Ownership classes: internal_only, single_partner, multi_partner_conflict, unclaimed.
• Score unclaimed/internal-only accounts (100 pts: tech 30, employees 25, ICP 25, funding 20).
• Sync assignments and scores to Account Intelligence via API/webhook.

Non-goals

• Salesforce IDs, map viz, live Nooks/Outreach, multi-rep admin, Clay (Sumble + Exa only).
• Fuzzy name matching when a domain already exists.

Upload 2+ partner sheets with different columns; classify ownership; resolve or flag missing domains; rank whitespace; retrieve partner contacts from a conflict view.

Question the MVP tests: Does domain-based matching plus scored whitespace beat spreadsheet co-sell?

1. Upload partner sheet + map columns
2. →Normalize domain (deduce only via search, never from memory)
3. →Classify ownership
4. →Enrich and score whitespace
5. →Export partner-safe lists
6. →Sync assignments to Account Intelligence

Plain language

1. Partner sheets
2. →Domain match
3. →Ownership
4. →Enrichment
5. →Score
6. →Export / sync

Technical

1. Next.js App Router
2. →Supabase Postgres
3. →SheetJS
4. →Sumble + Exa
5. →Claude API for name cleaning and disambiguation only

• accounts
• partners
• partner_sheets
• ownership_mappings
• enrichment_data
• account_scores
• rep_assignments
• export_logs

Operational

• Sheets with different schemas correctly classified
• Unresolved domains clearly flagged

Behavioral

• Whitespace ranking that survives manual review
• Conflict view used for partner contact

Business

• Measured: Metadot channel growth. Hypothesis: matching and scored whitespace increase partner-sourced pipeline.

If overlap, conflicts, and whitespace are structured around domain, partner activation and sourced pipeline should rise. Metadot is historical proof of the motion; Channel Territory Mapping is the system.

• Domain as sole match key vs. fuzzy names.
• Personal v1 vs. team admin.
• Search-backed domain deduction vs. model recall.

PRD for Channel Territory Mapping. Companion: Account Intelligence (account-signal-engine), independently deployable — no shared database.

Intended value

• Increased partner activation
• More account overlap identified
• Faster co-selling
• Increased partner-sourced / influenced pipeline

• Partner-auth’d enablement packs (battlecards + per-account messaging) in v2.
• Keep Clay out of this build.