Case study
How a staffing firm cut sourcing from 20+ hours a week to under 1 hour a week
A staffing firm’s recruiters spent 20+ hours a week sourcing candidates by hand. We built an AI workflow that pulls and ranks matching candidates the moment a job order lands and drafts the outreach. Sourcing dropped to under 1 hour a week, and placements per recruiter went from 3 to 5 a month with the same headcount.
- Trigger
- AI step
- Person decides
- Your systems
- TriggerJob order lands
- AI stepRead the requirements
- AI stepPull and rank matching candidates
- AI stepDraft personalized outreach
- Person decidesRecruiter reviews, adjusts and sends
- hours a week spent sourcing candidates
- 20+tounder 1
- placements per recruiter per month, same headcount
- 3to5
At a glance
- Client
- A staffing firm (anonymized)
- Problem
- Recruiters spent 20+ hours a week sourcing and updating the ATS instead of talking to candidates
- What we built
- An AI sourcing workflow triggered by each new job order: pull, rank, draft outreach
- Results
- Sourcing 20+ hours a week → under 1 hour a week; placements per recruiter 3 → 5 a month, same headcount
- Tools
- The firm’s existing ATS and candidate sources
The client is anonymized. The figures are the real before-and-after numbers from the project, reported as-is and not rounded up.
What was the problem?
Every new job order kicked off the same manual search: recruiters combing the ATS and other sources, building a shortlist, then writing outreach one candidate at a time. It added up to 20+ hours a week, time recruiters weren’t spending with candidates or clients.
In staffing, speed matters: whoever gets a qualified candidate in front of the client first usually wins the placement.
What did we build?
An AI workflow that starts the moment a job order lands. It reads the requirements, pulls matching candidates, ranks them against the role and drafts personalized outreach. Recruiters review the shortlist and the drafts, adjust, and send.
Recruiters stay in control of who gets contacted and what is said. The workflow removes the searching and the first-draft writing.
What changed?
| Measure | Before | After |
|---|---|---|
| Time spent sourcing | 20+ hours a week | under 1 hour a week |
| Placements per recruiter per month | 3 | 5, same headcount |
- hours a week spent sourcing candidates20+tounder 1staffing firm case study: hours a week spent sourcing candidates
- placements per recruiter per month, same headcount3to5staffing firm case study: placements per recruiter per month, same headcount
A related result. A first-pass agent that matches, ranks and drafts the candidate write-up cut time-to-submit by more than 60%. Recruiters review and send. That figure comes from a first-pass agent built for staffing work; it is reported on its own and not combined with the numbers above.
- drop in time-to-submit with a first-pass matching agent60%+staffing case study: drop in time-to-submit with a first-pass matching agent
What should you know before building something similar?
- Start with ranking and drafting, not auto-sending. Recruiters trust the shortlist faster when they approve every message at first.
- Write down what a “good match” means for your desk (must-haves versus nice-to-haves). The ranking is only as good as that definition.
- Log every candidate the workflow touches back to the ATS, so the database gets better instead of drifting.
Could this work for you?
If your team does similar work by hand, probably. Request a free workflow audit and you get a one-page plan: the 3 tasks worth automating first, hours saved, and what a build would cost. See also AI automation for recruiting and staffing and AI workflow automation.