> ## Documentation Index
> Fetch the complete documentation index at: https://docs.anysite.io/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Recruitment Agency — 6-Step Candidate Sourcing Pipeline

> How a 3-person AI agency built a recruitment pipeline processing 300+ candidates per day using Anysite's LinkedIn endpoints.

## The Challenge: SMBs Are Losing the War for Talent

In Germany, 71.7% of the skilled-worker gap falls on small and mid-sized businesses. These companies need qualified candidates — software engineers, sales professionals, operations managers — but they're competing against enterprises with dedicated recruiting teams and six-figure talent acquisition budgets.

The numbers tell the story. The average cost per hire sits at $4,700–$4,800. Time to fill an open role: 43 days. And in a market where top candidates are off the table within 10 days, a 6-week hiring process means systematically losing the best people to faster-moving competitors.

SMBs rarely have the in-house capacity to build automated recruiting systems. Most still rely on job postings and inbound applications — a passive approach that misses the 70% of professionals who are open to new opportunities but aren't actively job-hunting.

This is the gap a small European AI agency set out to close.

## The Solution: A 6-Step Candidate Sourcing Pipeline

The agency — a 3-person team, founded in 2024 — built a fully automated candidate sourcing pipeline using Anysite's LinkedIn endpoints as the data layer. Their SMB clients get a system that finds, qualifies, and surfaces candidates ready for outreach — in hours, not weeks.

Here's how the pipeline works:

| Step        | Action                                    | Endpoint            | Volume              |
| ----------- | ----------------------------------------- | ------------------- | ------------------- |
| 1. Discover | Find candidates by role, skills, location | `search_users`      | 6,006 calls (20.6%) |
| 2. Verify   | Check work history and qualifications     | `user/experience`   | 4,684 calls (16.0%) |
| 3. Research | Investigate target employers              | `company`           | 7,268 calls (24.9%) |
| 4. Expand   | Map talent at target companies            | `company/employees` | 4,194 calls (14.4%) |
| 5. Enrich   | Pull full profiles for shortlist          | `user`              | 3,612 calls (12.4%) |
| 6. Contact  | Get email for direct outreach             | `user/email`        | 1,108 calls (3.8%)  |

Each step feeds the next. The pipeline starts broad — thousands of potential candidates matched by role, skills, and location — and narrows to a shortlist of verified, enriched profiles with contact information ready for personalized outreach.

## The "Company Loop": An Account-Based Twist

The most interesting pattern in the data isn't the candidate-first flow — it's what happens at steps 3 and 4.

Company research accounts for 31.3% of all API calls (over 9,100 calls per quarter across `company`, `company/employees`, and `search_companies` endpoints). The agency doesn't just find individual candidates. They identify high-quality employers, then map those companies' entire workforce to discover additional talent.

This "company loop" is an account-based recruiting strategy: find one strong candidate, research their employer, then surface every qualified colleague at the same company. One good company yields dozens of potential candidates — a multiplier effect that a manual recruiter could never replicate at this speed.

## The Conversion Funnel: Quality Over Quantity

The pipeline is deliberately selective. Not every searched candidate makes it to outreach — and that's by design.

| Stage                | Volume | Conversion                |
| -------------------- | ------ | ------------------------- |
| Candidates searched  | 6,006  | —                         |
| Experience verified  | 4,684  | 78% pass initial filter   |
| Full profiles pulled | 3,612  | 77% pass experience check |
| Emails retrieved     | 1,108  | 31% reach outreach stage  |

**Overall: \~18% of searched candidates reach the outreach stage.** This isn't spray-and-pray mass outreach. It's a targeted pipeline that ensures every candidate who receives an email has been verified against the job requirements — leading to higher response rates and better-quality conversations.

## Results and Business Impact

**For the agency's SMB clients:**

* **Speed:** Candidates identified and contacted in hours, not the industry-average 43 days
* **Cost savings:** Each faster fill saves an estimated $6,000–$10,000 in vacancy costs (based on \~$300–$500/day mid-level role vacancy cost and \~20 days saved)
* **Reach:** Access to LinkedIn's 1B+ member network — far beyond the reach of job postings and inbound applications

**For the agency's business model:**

* **324 API calls per day** powering production client pipelines — not a prototype, a real product
* \*\*$0.057 per API call** on Anysite's Pro plan — compared to $0.49+/call on Proxycurl or \$15,000+/year minimums on enterprise platforms like ZoomInfo
* **One Pro subscription** serves multiple client pipelines simultaneously, making the unit economics work even for a 3-person operation

The cost advantage is significant. Running this same pipeline on a competitor's API would cost roughly $14,300 per quarter — over 8x more than the $1,647 spent on Anysite during the same period.

## Key Anysite Endpoints Used

| Endpoint                     | Pipeline Role                             | Why It Matters                                          |
| ---------------------------- | ----------------------------------------- | ------------------------------------------------------- |
| `linkedin/search/users`      | Find candidates by role, skills, location | The pipeline's entry point — defines the candidate pool |
| `linkedin/user/experience`   | Verify work history and seniority         | Turns raw search results into qualified candidates      |
| `linkedin/company`           | Research target employers                 | Powers the account-based "company loop" strategy        |
| `linkedin/company/employees` | Map talent at target companies            | Multiplier: one company yields many candidates          |
| `linkedin/user`              | Full profile for shortlisted candidates   | Enables personalized, informed outreach                 |
| `linkedin/user/email`        | Contact information for outreach          | The pipeline's output — actionable contact data         |
| `linkedin/search/companies`  | Discover companies by industry and size   | Feeds the company-first discovery channel               |

## Key Takeaways

* **A small team can build big products.** Three people, one Anysite Pro plan, and a well-designed pipeline process 300+ candidates per day across multiple client engagements.
* **Company intelligence is a force multiplier.** The account-based approach — researching companies to find candidates, not just searching for individuals — accounted for nearly a third of all API usage and dramatically expanded the candidate pool.
* **Selectivity drives quality.** An 18% search-to-outreach conversion rate means every contacted candidate has been vetted — leading to better response rates and faster placements.
* **Cost matters at the data layer.** At \$0.057 per call, Anysite makes it economically viable for small agencies to build recruitment products that compete with enterprise-grade solutions.