Tally B2B lead scoring with OpenAI

Three OpenAI agents read the homepage and score the lead. This n8n template starts when a marketing employee submits a Tally form with a company name, LinkedIn URL, website URL and the scoring criteria that matter. A first AI agent normalizes the submission. The workflow fetches the company's homepage HTML, and a Code node strips it to readable text. A second agent extracts what the company does, its tech stack, size signals and industry. A third agent scores the lead against the chosen criteria, assigns a grade and writes a recommended action. The scored lead is appended to a Google Sheet. The page names OpenAI GPT-5.4 for the three agents.

n8nRead on 2026-11-167 steps

Steps

  1. Trigger: Marketing employee fills out a Tally formTally
    Company name, LinkedIn URL, website URL, scoring criteria.
  2. Agent 1 normalizes the payloadOpenAI
  3. HTTP Request node fetches the company's homepageHTTP Request
  4. Code node strips the HTML to text
  5. Agent 2 extracts structured intelligenceOpenAI
    What the company does, tech stack, size signals, industry.
  6. Agent 3 scores the lead 1-10OpenAI
    Assigns a grade and a recommended action.
  7. Output: Enriched lead appended to a Google SheetGoogle Sheets

Limits

  • Reads only the company's homepage.
  • The Tally form is filled by a team member, not by the prospect.

What it does well

  • Scores against the criteria chosen in each form submission.
  • Logs tech stack, grade and reasoning for every lead.

Before you import

After the import, add Tally, OpenAI and Google Sheets credentials, then create a Tally form and a leads tab with the listed columns.

Similar templates