A worked example: a staffing and recruiting firm.
This is the one-page plan a free review produces, fully worked for a typical nine-person staffing shop. Read it as an owner would: if the shape of the hours below looks like your desks, your version of this page takes one hour to earn.
Where your hours are going
Nine people in Grove City: six recruiters on a mix of clerical temp, contract, and perm desks, two in the back office, plus you. Bullhorn for the ATS — core only, no Automation add-on — LinkedIn Recruiter seats, Indeed, and one client that makes you work through Fieldglass. The industry time-audits say about half a recruiter's week is admin; here's what that looks like across your six desks:
- Resume reformatting and submittal write-ups: ~35 submittals a week, ~20 minutes each to strip, re-template, and write up (published range is 10–25 minutes) = ~12 hrs
- Screening notes written twice: ~40 screens a week; notes typed after the call, then re-written for the submittal — or never captured at all, ~15 minutes each = ~10 hrs
- Monday chasing: timesheets and onboarding packets, by email and text, every week = ~5 hrs
- Interview scheduling tag between clients and candidates = ~4 hrs
That's roughly 31 hours a week — most of a full recruiting desk — spent on copy-paste, formatting, and chasing instead of talking to candidates and clients.
Do this week — free, ~1 afternoon, no help needed
Set up LinkedIn Recruiter saved-search alerts and a Project per req. Those seats cost you around $10,000 a year each, and most recruiters use a fraction of them — running the same search fresh every morning. Build one Project per open req and save the search with alerts on, and new matching candidates surface themselves daily instead of being re-hunted. No new tools, no cost, one afternoon of setup discipline.
The four that matter
| # | Recommendation | Hours/wk back | Setup | Cost |
|---|---|---|---|---|
| 1 | Screening calls summarize themselves into Bullhorn | ~8 | 1–2 weeks | ~$20–50/recruiter/mo |
| 2 | Resume reformat + submittal draft, human-reviewed | ~7 | 2 weekends | ~$30–60/mo |
| 3 | Redeployment and dead-database re-engagement | ~3 | 2–3 weekends | ~$50/mo + texting tool |
| 4 | Timesheet chasing and VMS reconciliation assist | ~4 | 1 weekend | ~$20–30/mo |
Recruiter runs the screen as normal; the call is recorded (with consent), AI writes the structured summary — availability, pay expectations, must-haves, red flags — straight into the Bullhorn candidate record, and drafts the submittal blurb from it. The recruiter edits instead of typing from memory — no more double writing, no more screening intelligence dying in notebooks. A purpose-built tool or a generic recorder plus an AI step both work. "Done" looks like: every screen from last quarter is searchable in the ATS. Trial it yourself in a week, or I set it up and template the outputs, flat fee.
Candidate resume in, agency-template resume out — contact details stripped, your branding on, formatting consistent — with a draft submittal summary assembled from the resume and the screening notes from #1. A recruiter reviews and sends; nothing goes to a client unread. At ~35 submittals a week and ~20 minutes each, this is your single biggest block of recoverable hours. Runs in your accounts on a business-tier AI with a Word template. Build it yourself over two weekends, or I build it in two weeks.
Your ATS knows every contractor's assignment end date, and your database is full of people nobody has spoken to since they weren't placed. Automated plays: an alert 30 days before each end date so a recruiter calls first, plus drip check-ins to finishing contractors and dormant candidates. Industry-average redeployment is 30–40%; best-in-class firms run over 70% — every point in between is a placement you already paid to source. Bullhorn Automation sells this for roughly $750 a month; both plays can be built on Make plus a texting tool for a fraction of that. One hard requirement: texting recruits is TCPA territory, so consent capture, opt-out handling, and quiet hours get built in from day one, not bolted on. DIY if someone owns it, or I build it with the compliance pieces included.
Monday reminders go out automatically, escalating only the stragglers to a human, and the Fieldglass export gets compared against your Bullhorn records with mismatches flagged before payroll runs instead of after invoices bounce. "Done" looks like: the back office starts Monday with a short exception list, not a phone list. A weekend of setup in Make with your existing exports — yours or mine.
What I'd not automate, and why
Any AI that decides who advances. AI drafts, formats, summarizes, reminds, and reconciles — a human makes every advance-or-reject call. The EEOC has already settled with a company whose tool auto-rejected older applicants ($365,000), and the Workday litigation is testing whether the liability reaches everyone in the chain — including staffing firms. The admin around decisions, never the decision.
AI video-interview screening. Candidate drop-off is real, pricing is enterprise, and states are regulating it one by one. Not at nine seats.
Fully autonomous outreach. Robotic sequences burn a local candidate pool fast, and Columbus is a small town when your business runs on reputation. Candidate PII — resumes, SSNs, I-9 docs, background results — stays in your systems and your purpose-built vendors; automations run in your accounts, and I never possess your data.
If you want help
The review is free and everything above works without me. If you'd rather have any of it built and running for you, that's a flat-fee engagement. Either way, I'm around.
Worked example — not a client engagement. The firm is illustrative; the benchmarks are real. Tool prices are illustrative.
Your firm's version of this page takes one hour to earn.
Claim a free reviewTen free reviews over the next few weeks, Columbus-area firms of 5–25 people. How the review works.