
03 · 2024 · Data pipeline
A conference prospecting pipeline: three disconnected data sources in, one CRM-ready lead import out.
Live - in production
Solo build
Per-conference run
Overview
Conference prospecting used to be hours of bridging three tools by hand. Now it's a single command.
Watershed merges LinkedIn, ZoomInfo, and Apollo against a named-account territory file, classifies each lead by verified email domain, and scores it against the conference portfolio - producing thousands of import-ready rows per run.
The build, in four parts
01 - 04
01 - The problem
Conference prospecting runs on three sources that don't connect - LinkedIn, ZoomInfo, and Apollo - plus a named-account territory file in its own spreadsheet. Every cycle meant hours of manual work: matching names, guessing domains, and assigning territory by hand. It was tedious, error-prone, and different every time.
02 - The approach
First, merge LinkedIn and ZoomInfo on a cleaned Name + Org key. Second, classify territory by verified email domain instead of guesswork. Third, score every lead against the conference portfolio so the best-fit accounts rise to the top. Each stage is auditable and can be re-run on its own.
03 - The build
A single Python CLI runs the whole pipeline, one subcommand at a time. The domain-to-territory map is stored and grows with every run, so classification gets sharper over time. An outbound skill feeds the scored output straight into sequenced outreach via Dripify - no reformatting or copy-paste.
04 - The outcome
Every run produces thousands of import-ready leads - conference-assigned, hierarchy-tagged, and Sales-Comments filled in. What used to be an afternoon of spreadsheet work is now a command I start and walk away from.
3 → 1
sources reconciled into one import
1000s
import-ready leads per run
Want more detail? Get in touch →