Most recruitment agencies start their day the same way: someone opens LinkedIn, types in a few search queries, and manually copies job postings into a spreadsheet. For a 22-person UK agency we worked with, this was costing 4 hours a day across three researchers.

We replaced that entirely with an automated pipeline. Here's how it works.

The Problem With Manual Job Discovery

Before we built this, their researchers were:

  • Running the same 12 saved LinkedIn searches every morning
  • Copying job titles, company names, and hiring manager info into Airtable
  • Missing jobs posted after 9am until the next morning
  • Losing deals because a competitor called the hiring manager first

The fix wasn't hiring more researchers. It was removing them from the process entirely.

The Stack

Clay — data enrichment and waterfall workflows Make (formerly Integromat) — automation orchestration LinkedIn Jobs API (via RapidAPI) — job posting source Apollo.io — contact finding for hiring managers Smartlead — email sequencing Airtable — central CRM / deal tracking

How the Pipeline Works

Step 1: Job ingestion (every 2 hours)

A Make scenario runs every 2 hours and hits the LinkedIn Jobs API with 14 pre-configured search queries — combinations of job title, location, and company size filters.

Each new job posting gets pushed into a Clay table with:

  • Job title
  • Company name
  • Location
  • Date posted
  • Original job URL
Set the Make schedule to run on the half-hour (e.g. 8:30am, 10:30am) rather than on the hour. LinkedIn's index refreshes asynchronously — you catch more fresh postings by offsetting your polling time.

Step 2: Company enrichment in Clay

Once a job lands in Clay, an enrichment waterfall runs automatically:

  1. Company size — pulled from Clearbit (via Clay's native integration)
  2. Tech stack — BuiltWith API identifies what tools the company uses
  3. Recent funding — Crunchbase enrichment flags companies that raised in the last 90 days
  4. Hiring velocity — LinkedIn company page scrape counts how many open roles they currently have

Jobs at companies with 50–500 employees, recent funding, and 5+ open roles get flagged as High Priority.

Step 3: Finding the hiring manager

For High Priority jobs, Clay runs a contact-finding waterfall:

1. Apollo.io — search by company + "Head of Talent" / "Talent Acquisition" 2. Hunter.io — email pattern guess from domain 3. LinkedIn profile scrape — confirm person is still at company 4. Fallback: flag for manual review in Airtable

Hit rate: roughly 68% of High Priority jobs get a verified hiring manager contact.

Step 4: Sequence trigger

Once a contact is verified, Make pushes them into a Smartlead sequence automatically. The first email goes out within 6 minutes of the job being posted.

That's the edge. Most competitors are still manually searching LinkedIn 8 hours later.

The Numbers

After 6 weeks running the pipeline:

Metric Before After
Jobs discovered/day 40–60 200–250
Time spent on discovery 4 hrs/day 15 mins/day
First-mover contacts ~30% ~71%
Placements in 60 days 4 11
These numbers are specific to a UK-based engineering recruitment agency targeting Series A–C fintechs. Your results will vary based on your niche, message quality, and how competitive your target market is.

What We'd Do Differently

Two things we'd change if we rebuilt this today:

1. Add a deduplication layer earlier. We initially let duplicates flow through Clay and deduplicated at the Airtable stage. This wasted Clay enrichment credits on the same company multiple times. Now we deduplicate on company domain before enrichment runs.

2. Use n8n instead of Make for the orchestration layer. Make's operations-based pricing gets expensive at volume. n8n (self-hosted) runs the same workflows at a fraction of the cost. We've since migrated three clients off Make entirely.

If you want this pipeline built for your agency, we can have it running in two weeks.