Lead list builder with scheduled cold email drafts

Lead list builder with scheduled cold email drafts

An n8n workflow in two halves. A form takes the people you want to reach, by level, company size, industry and country,
and a few lines about you. An Apify Actor finds matching people with verified emails, and each one becomes a row in a
Google Sheet. Once a day, the workflow drafts a short cold email for every lead you have marked REVIEW and writes it
into that row. Nothing is sent.

What is inside

• workflow.json: the workflow as n8n exports it, sixteen nodes. Two of them are sticky notes that name the halves. It
  holds no credential.
• README.md: these notes.

What we changed from the original copy

• The Apify header held a placeholder, Bearer <token>. It now says Bearer <YOUR_APIFY_TOKEN>.
• The author's sheet and its tab id are gone. The three Google Sheets nodes ask for YOUR_SHEET_ID and find the tab by
  the name Leads.

Before you start

• An n8n instance, on n8n Cloud or your own server.
• An Apify account and its API token, with credit for the lead search.
• An OpenAI API key. The file names two models, chatgpt-4o-latest and gpt-4.1-mini, and the first one is gone (step 5).
• A Google account, and a Google Sheet for the leads.

Set it up

1. In n8n, create a workflow, open its menu and choose Import from File. Pick workflow.json.
2. Make a Google Sheet with a tab named Leads and these headers in row 1: FirstName, LastName, Title, Email,
   LinkedIn, Organization, OrganizationLinkedIn, OrganizationSite, Email Prepared and
   About You & Your Service. Copy the sheet's ID: the long part of its address between /d/ and /edit.
3. Add a Google Sheets OAuth2 credential. In Google Sheets, Get Pending Leads and Update Google Sheets, select it
   and replace YOUR_SHEET_ID with the ID.
4. Add an OpenAI credential and select it in Message a model and OpenAI Chat Model.
5. In Message a model, pick another model. OpenAI's deprecations page, read on 2 November 2026, lists
   chatgpt-4o-latest as shut down on 17 February 2026. gpt-4.1-mini, which the other half already uses, can answer
   in JSON as this node asks.
6. In HTTP Request, replace <YOUR_APIFY_TOKEN> in the Authorization header with your token. The brackets go too, so
   the header reads Bearer, a space and the token.
7. In Edit Fields, lower totalResults from 1000 to 20 or so for your first runs.
8. Save. Send the form once from its test address, then open the sheet and type REVIEW in Email Prepared on a row or
   two.
9. Run Schedule Trigger by hand to draft those rows. When the drafts read well, press Publish, and the schedule takes
   over.

How a run goes

The form half:

1. On form submission shows a form titled Lead Scraper. Job Management Level lets you pick any of Owner, Founder,
   Head, Director and Manager, and Location has a checkbox for each of ten countries. Employees Size and Industry
   are free text. About You & Your Service is the one required field.
2. Message a model tidies the answers into JSON. It matches your industry to the nearest of 148 names held in its
   prompt and fits the size to nine bands, from a company of one up to more than 10,000 people.
3. Edit Fields turns that JSON into the Actor's input: verified emails only, emails included, the four filters and a
   totalResults of 1000.
4. HTTP Request posts it to Apify's run-sync-get-dataset-items address for the Actor cdTI90GLKIsTHSjgE and waits for
   the people. Apify's listing for that ID names it Leads Finder, from the account boneswill.
5. Edit Fields1 keeps eight fields for each person. Then Filter drops anyone without an email, and
   Remove Duplicates drops a second copy of the same email within this run.
6. Google Sheets appends each person as a row, with your About You & Your Service text beside them. Email Prepared
   stays empty.

The scheduled half:

1. Schedule Trigger fires once a day at midnight, in the workflow's time zone. Its rule is empty in the file, and that
   is what n8n does with an empty rule.
2. Get Pending Leads reads every row whose Email Prepared says REVIEW.
3. Basic LLM Chain has gpt-4.1-mini write the body of a cold email of 130 words at most, with no subject, greeting or
   sign-off. It also picks your first name out of About You & Your Service. A parser holds the answer to text and
   name.
4. Update Google Sheets finds the row by its number and replaces REVIEW with the draft. The draft opens with Hey
   and the lead's first name, and closes with Best, and yours.

Change the look

• The email: the prompt and the message in Basic LLM Chain. The prompt sets the 130-word cap and the order: a line
  about the lead's company, your intro, how the two fit and a short ask.
• The frame around it: the Email Prepared value in Update Google Sheets.
• The search: the lists in the prompt of Message a model, and totalResults in Edit Fields.
• The time of the drafts: the rule in Schedule Trigger.

Good to know

• Nothing is sent. Drafts land in Email Prepared, and only for rows you mark REVIEW. That mark is your approval step,
  since the workflow never sets it.
• On 2 November 2026, the Actor's input page wrote seniority as values like director_level and company size as
  bands like 10-19. The prompt here produces Director, writes its size bands differently, and sends
  contactEmailStatus and includeEmails, which the page does not list. Run one small search and check that the people
  match. If they don't, put the Actor's own values into the lists in Message a model.
• Apify bills the run. Its listing, read the same day, charges $0.03 for each start and $0.0017 for each lead on the free
  plan, less on paid plans. OpenAI bills one call for each form and one for each draft.
• Apify ends a synchronous call after 300 seconds, according to its API reference. A large search can time out before
  the people come back.
• Remove Duplicates only compares the people of one run, so a second search can add someone already in the sheet.
• In the file, OrganizationLinkedIn receives the company's website, not its LinkedIn page. To fix it, map
  {{ $json.organizationLinkedIn }} into that column in Google Sheets.
• Title is filled from the Actor's functional field, which may hold a department more than a job title.
• Each lead's name, title, email and LinkedIn address are kept in your sheet. The form's answers and each lead's title,
  company and links go to OpenAI. Check the rules for cold email where you and your leads are.
• Once the workflow is published, anyone with the form's address can start a search on your Apify account.
• The Apify token sits in a header, so take it out before you share the workflow.