Case StudyFramework-in-ActionBefore/After Metrics

How a Mid-Market Agency Doubled Placements Using an AI Sourcing Agent

January 2, 2025

14 min read

What you'll learn: Why the team's old process stalled, how the AI sourcing agent was implemented, and the exact before/after metrics that led to more qualified replies and doubled placements.

Case study: AI sourcing agent doubles placements—overview diagram of workflow and KPIs — From slow sourcing to compounding placements with an AI agent.

Describe the company and goals

Mid-market agency (15-20 recruiters) focused on healthcare and IT staffing. Primary KPI: placements per month. Secondary: qualified reply rate and time-to-first-touch.

Stack: ATS/CRM (e.g., Bullhorn/Greenhouse), Gmail/Outlook, LinkedIn.
Constraint: tight time-to-hire; limited research capacity.
Goal: increase placements without adding headcount.

Show the baseline: before metrics

Before launch, the team struggled with inconsistent sourcing and slow first touches. Below are the key pre-deployment numbers.

Qualified reply rate (QR): 8%
Interviews booked (IB): 12 / month
Placements: 4 / month
Time-to-first-touch (TFT): 48 hours

Funnel chart showing baseline: low reply rate, limited interviews, flat placements pre-agent — Baseline funnel before the AI sourcing agent.

Reference the framework and outline steps

This deployment followed the Agent Operations Framework and Agent Evaluation Framework.

1) Define goals

Select one role (e.g., SDRs) and target a lift in qualified replies and placements.

2) Connect data

Wire ATS/CRM, enable consent flags, and enrich profiles for skills and location.

3) Create the sourcing agent

Draft prompts and tools for research, dedupe, and shortlisting; attach policy filters.

4) Launch pilot

Run feature-flagged pilot for 2–4 weeks with control vs. treatment cohorts.

5) Measure & promote

Promote versions that sustain lift; document learnings for scale-up.

Workflow swimlane: data connect, sourcing agent research, outreach, scheduling, measurement — Framework-in-action from data to placements.

Report the results: after metrics

Qualified reply rate (QR): 18% (from 8%)
Interviews booked (IB): 28 / month (from 12)
Placements: 8 / month (from 4) — ~2×
Time-to-first-touch (TFT): 6 hours (from 48)

Bar chart showing after metrics: higher replies, more interviews, doubled placements, faster time-to-first-touch — After the pilot: outcome metrics improved across the funnel.

Explain what changed and why it worked

Personalization improved relevance and response.
Sequenced follow-ups kept momentum across channels.
Scheduling automation reduced drop-offs and no-shows.
Governance & QA prevented off-brand or risky messages.

Highlights: personalization, follow-ups, scheduling, governance—top drivers of lift — Four levers that drove the biggest impact.

Add the client quote and headshot

"We didn't expect results this fast. The agent found better fits, and our recruiters focused on real conversations. Placements doubled without extra headcount."