The agent queries firmographic databases, LinkedIn, job boards, technographic providers, and first-party data simultaneously to identify every company in the addressable market. Filters apply at the source — company size, industry, geography, technology requirements — so the raw list starts with relevant accounts, not a universe to manually prune. Duplicates are resolved across sources. Known customers, competitors, and disqualified accounts are excluded. The output is a clean, sourced list of every account worth pursuing — typically built and ready for scoring within hours.

Every account on the list is scored across four dimensions: firmographic fit (size, industry, geography, growth stage), technographic fit (tools they use that make your product valuable or compatible), intent signals (third-party intent on relevant keywords), and behavioral signals (website visits, content downloads, event attendance). The composite score produces a tier — Tier 1 for high-fit, high-intent accounts that should receive full ABM treatment; Tier 2 for high-fit, lower-intent accounts that enter nurture; Tier 3 for moderate-fit accounts on a watch list. Scoring runs weekly and tiers shift as new signals arrive.

Target lists decay. Accounts get acquired, shut down, hit hypergrowth, or switch tools. The agent monitors every account on the list for changes that affect their score — new funding, headcount changes, tool additions, leadership changes — and updates tiers automatically. Accounts that meet the ICP better over time move up. Accounts that get acquired or are known to have signed a long-term competitor contract move down or off. The list reflects the market as it is today, not as it was when the spreadsheet was last refreshed.