n8n Workflow: Score Account ICP Fit from Domain Lists

Turn domain lists into ICP-fit scoring (CSV/XLSX output) — automatically

This n8n workflow takes a list of company domains, defines your account-level ICP with OpenAI, fetches each company’s website content, and scores every account for “ICP fit”—then returns a downloadable CSV or Excel file with your results.

What this workflow does

• Ingests domains via an n8n form submission using CSV/XLSX upload or a pasted domain list, along with a run name, scoring objective, and desired output format.
• Normalizes and deduplicates domain rows, stopping early with a completion message if no valid companies are found.
• Builds an ICP profile: you provide an initial ICP description, and the workflow uses OpenAI (GPT-4.1) to generate up to three rounds of structured follow-up questions, then finalizes a structured ICP.
• Validates and fetches evidence for each domain in batches, including the company homepage and an /about page via HTTP requests.
• Extracts structured company signals: it creates guarded evidence from site content plus deterministic domain-based signals, then uses OpenAI (GPT-4.1-mini) to extract structured company signals.
• Scores deterministically against the ICP (optionally using a saved ICP library), with an option to generate a short match explanation.
• Exports results by aggregating scored and rejected rows, converting them to CSV or XLSX, and returning the file via a download form with a run summary.

Use cases

• Score inbound prospect lists or scraped domain lists for ICP-fit before routing to sales.
• Enrich and qualify target accounts for a SaaS go-to-market motion.
• Run repeatable account research to compare against different scoring objectives.

Technical details

• n8n form for receiving domain inputs and output preferences.
• Workflow logic using if, code, no op, wait, and merge.
• OpenAI for ICP question rounds (GPT-4.1), company signal extraction (GPT-4.1-mini), and optional match reasoning.
• HTTP fetching of homepage and /about pages to build evidence for scoring.