Here's the setup: you paste company websites into column A of a Google Sheet, and a few minutes later the founder's name, title, email and phone number appear next to them. No copying CSVs in and out.
I'll show it in Make first, then the same thing in n8n with a single HTTP request. The scraping step uses an Apify actor that reads company websites. Any actor that takes a list of URLs and returns rows will slot in the same way.
The sheet
Make one tab called Leads with these headers in row 1:
| Col | Header | Filled by |
|---|---|---|
| A | website | you |
| B | status | automation |
| C | founder_name | automation |
| D | founder_title | automation |
| E | founder_email | automation |
| F | email_confidence | automation |
| G | shared_inbox | automation |
| H | socials | automation |
| I | phones | automation |
Column B is the important one. The automation only picks up rows where status is empty, and writes done or no contacts when it's finished. That stops it re-processing the same row and paying twice.
Paste domains or full URLs into A. acme.com and https://www.acme.com/ both work with most scrapers, but be consistent, because you'll match on this value later.
The Make scenario
Five modules in a row:
1. Google Sheets: Search Rows. Pick your spreadsheet and the Leads tab. Filter: status (column B) "does not exist" / is empty. Set a limit, say 50. Small batches are easier to debug and stay well under time limits.
2. Tools: Array aggregator. Aggregate the website values into one array. Without this, Make runs the actor once per row, which is slower and starts a separate run every time. In the aggregator's source module pick Search Rows, and aggregate only column A.
3. Apify: Run an Actor. Connect your Apify account with an API token (Apify console → Settings → Integrations). Choose the actor. Set "Run synchronously" to yes, so Make waits for the run to finish. In the input JSON, map the array:
{ "websites": {{map(2.array; "website")}} } (The exact mapping expression depends on what your aggregator outputs. Make's mapping panel will show it after one test run.)
4. Apify: Get Dataset Items. Map the dataset ID from the previous module. Each item is one company.
5. Google Sheets: Update a Row. This is where the results go back. The catch is that you need the row number for each website. Two ways:
- Simple: put a Google Sheets: Search Rows module before it, searching column A for the item's website, and use its row number.
- Cleaner: keep the row numbers from step 1 in the aggregator alongside the websites, then look them up.
Map the fields: founder name to C, title to D, email to E, the confidence label to F, and so on. Write done into B. If the item has no contacts, write no contacts into B so the row isn't retried forever.
Run it once with three rows and check every column before you schedule it. Then set the scenario to run every hour, or trigger it manually when you've pasted a new batch.
A tip on joining arrays. Phones and socials come back as lists. Use Make's join() function to turn them into one cell, e.g. join(phones; ", "), or you'll get only the first value.
The n8n version
n8n has an Apify node, but the plain HTTP Request node is simpler because Apify has an endpoint that runs an actor and returns its results in one call.
1. Google Sheets node: operation "Get Rows", with a filter on status being empty. Add a Limit node (50) after it.
2. Code node to build the request body from all incoming items:
const websites = $input.all().map(i => i.json.website).filter(Boolean); return [{ json: { websites } }]; 3. HTTP Request node:
- Method:
POST - URL:
https://api.apify.com/v2/acts/intropi~company-contacts-tech-lookup/run-sync-get-dataset-items - Authentication: header
Authorization: Bearer <your Apify token>(store it as a credential, not in the URL) - Body: JSON,
{{ $json }}, which sends{"websites": ["acme.com", "..."]} - Timeout: raise it to 300000 ms
The sync endpoint waits for the run and returns the dataset items as an array, so the next node receives one item per company. Apify caps sync runs at five minutes. In my tests 63 small-business sites took 81 seconds, so batches of 50 are comfortable. If a batch times out, make it smaller.
4. Google Sheets node: operation "Update Row", matching on column website. Map the fields to C–I and set status to done.
One snag: the actor may normalise the URL (adding https:// or dropping www.). If the match column doesn't line up, add a Set node that normalises both sides the same way, or match on the domain only.
Before you run it on 5,000 rows
- Start with 10. Read the results. Does the "founder" look like the founder? Are the emails on the right domain?
- Keep the confidence column. A published address and a guessed one are not the same thing. Sort by it and treat them differently in your outreach.
- Know the cost. Check the actor's pricing before scheduling anything hourly.
- Send like a person. If this feeds a cold email tool, keep volume low, write to people who'd plausibly care, and honour opt-outs. CAN-SPAM applies in the US, GDPR and PECR in the EU and UK.
The actor in the example
The actor ID in the n8n URL is mine: Founder Email Finder (https://founders.intropi.in). It reads each site's home, about, team, contact and imprint pages and returns the founder or owner with their title, the email marked published-on-site or a firstname@ guess on a domain with MX records, the best shared inbox, phones, socials and tech stack. It costs $1.50 per 1,000 websites with contacts, failed sites are free, and Apify's free monthly credit covers a few thousand.
Disclosure: I built it. The sheet layout and modules work the same with any other website scraper on Apify.
Written with AI assistance; the numbers come from my own test runs.