The operations review at a four-location outpatient group runs long again. Fourteen clinical reqs are open: six medical assistants, four LPNs, two RNs and two radiology techs. Two LPN shifts have been covered by a per diem agency since August, and the founder is looking at a vendor proposal for a talent sourcing platform that promises "more qualified candidates." The question on the table is simple: how would the team know, inside one month, whether that promise holds on its own reqs?
A talent sourcing platform pilot is a 30-day, side-by-side test in which half of a clinic's open clinical reqs run through the new platform and half stay on the current process, with both arms measured on the same four numbers: qualified screens per req, time to first contact, screen-to-interview rate and agency fill share. If the pilot arm does not beat the control arm on those numbers, the platform is not worth the contract, however strong the demo looked.
That framing matters because small healthcare employers rarely get a second chance at a buying decision. A year-long contract that fails to move the agency bill becomes a sunk cost the clinic carries through the next headcount plan.
Why clinics need a talent sourcing platform pilot, not a demo
A demo shows what a platform can do on a vendor's sample data; a pilot shows what it does on your applicant flow, your shifts and your pay bands. For a clinic group, those are the variables that decide whether an MA req fills in two weeks or six.
Demand is the first reason the distinction matters. The U.S. Bureau of Labor Statistics projects that employment of medical assistants will grow 13 percent from 2025 to 2035, with about 109,700 openings projected each year. Every outpatient practice in your metro is fishing in that same pool, which means a platform's reach claims only count if they turn into conversations with people in your commute radius who hold the right credential.
Cost pressure is the second reason. In a SHRM analysis of Appcast's 2026 benchmark data, Appcast's chief economist noted that 95% of net new job gains in 2025 were in health care, while both cost-per-application and cost-per-hire climbed sharply. When every application costs more, buying more applications is not a strategy. A pilot forces the question of what each extra applicant turns into.
The third reason is mobility. A 2026 Harris Poll conducted for Strategic Education and Workforce Edge found that 59% of direct-care healthcare workers say they are likely to look for another role within the next year, while employers estimated only 39%. The candidates are out there and moving. The constraint for a small clinic is usually reaching them and responding before someone else does.
What a talent sourcing platform has to prove in 30 days
A talent sourcing platform earns its contract by producing more qualified, reachable candidates per req at a lower cost per filled role than the clinic's current mix of job boards and agency. Applicant volume alone proves nothing.
Most clinic groups already know their applicant count. What they rarely measure is how many of those applicants hold an active license, can work the posted shift and answered a recruiter within 48 hours. That is the population that turns into hires, and it is where small teams lose ground: the office manager or the single recruiter reviews applications between other duties, first contact slips to day four or five, and the strongest LPN in the stack has already taken a call from the hospital across town.
Nurses describe the cost of that delay from their side. Incredible Health's 2026 State of Nursing report found that 65% of job-searching nurses accepted or seriously considered a role they had reservations about simply to end the search. The same report notes that among healthcare employers using AI, only 4% use it to screen or hire. Read together, those two findings describe a market where candidates take the first credible offer and most employers still respond at human-inbox speed.
So the pilot should test the full path from posting to scheduled interview, not just sourcing reach. Four capabilities decide that path for clinical roles:
- Distribution to the boards clinicians actually use, including healthcare-specific boards alongside Indeed and ZipRecruiter.
- Rediscovery of past applicants already sitting in your inbox or ATS, ranked against the current req.
- Structured first contact that captures license type and state, certifications, shift availability and commute within hours, at any time of day.
- Clean handoff of shortlisted candidates into your ATS, or into a lightweight one if the clinic runs hiring from email and spreadsheets.
How to design the 30-day pilot
The cleanest pilot design splits comparable reqs into two arms and runs them in parallel, so seasonality, pay changes and local market shifts hit both arms equally.
Pair reqs by role and location: one MA req at Clinic A goes to the pilot, the matching MA req at Clinic B stays on the current process. Do the same for LPNs and techs. Keep pay, posting copy and hiring-manager availability identical across arms. If a clinic manager in one arm takes a week to schedule interviews, the comparison breaks no matter which platform is in use.
Agree on definitions before day one. "Qualified" should mean something written down: active license in your state, the required certification (BLS, ARRT, phlebotomy), availability for the posted shift and an acceptable commute. Agree also on what counts as first contact (a two-way exchange, not an automated acknowledgment) and which date starts the clock (application timestamp).
The pilot scorecard
Copy this table into a shared sheet and fill both columns weekly. The pass thresholds are a starting point; set your own before the pilot begins, not after you see the results.
| Metric | How to measure it | Suggested pass threshold for the pilot arm |
|---|---|---|
| Qualified screens per req | Completed screens that meet the written qualified definition, divided by reqs in the arm | At least 1.5x the control arm |
| Median time to first contact | Hours from application timestamp to first two-way exchange | Under 24 hours, including nights and weekends |
| Screen-to-interview rate | Interviews scheduled divided by qualified screens | Equal to or better than control (quality check on the shortlist) |
| Recruiter hours per qualified screen | Hours logged on review, calls and scheduling, divided by qualified screens | At least 50% lower than control |
| Agency fill share | Reqs or shifts escalated to agency after day 21, by arm | Fewer escalations than control |
| Offer-accept rate | Accepted offers divided by offers extended | No worse than control |
A worked example: 14 clinical reqs across four clinics
The numbers below are illustrative assumptions for the outpatient group in the opening scene, not benchmarks or vendor results. They show how a 30-day pilot readout should be built.
Setup. Fourteen reqs, split into two arms of seven (three MAs, two LPNs, one RN, one rad tech in each). The current process is job-board postings reviewed by the office manager and one recruiter, with phone screens booked by email. Assumptions: manual review takes 4 minutes per application; a phone screen, including scheduling back-and-forth, takes 25 minutes of recruiter time.
Control arm (current process). 210 applications. Review time: 210 × 4 minutes = 14 hours. The team attempts 30 phone screens (30 × 25 minutes = 12.5 hours) and completes 18 that meet the qualified definition; the rest are no-shows or candidates who have already accepted elsewhere. Median first contact: 4 days. Total recruiter time: 26.5 hours, or about 1.5 hours per qualified screen. Ten interviews, three offers, two accepts. One LPN req is escalated to the agency on day 22.
Pilot arm (talent sourcing platform plus structured screening). 230 applications, of which 40 are past applicants rediscovered from the clinic's own records. Every applicant receives a structured screen within hours that captures license, certification, shift and commute. 31 complete it and meet the qualified definition. The recruiter reads the 31 screen summaries at 10 minutes each (about 5.2 hours) and spends 3 hours spot-checking the ranked list, for 8.2 hours total, or about 0.26 hours per qualified screen. Median first contact: under 12 hours. Seventeen interviews, five offers, four accepts. No agency escalations.
Reading the result. Qualified screens per req: 4.4 in the pilot versus 2.6 in control, which clears the 1.5x threshold (2.6 × 1.5 = 3.9). Screen-to-interview rate: 55% versus 56%, effectively equal, which says the shortlist quality held as volume rose. Recruiter hours per qualified screen fell by more than 80%.
Agency share is where the business case lands. Assume the agency bills an LPN at $75 an hour against a $45 fully loaded direct cost, a $30 premium. At 36 hours a week over 4.3 weeks, one escalated LPN req costs about $4,644 a month in premium (30 × 36 × 4.3). If the clinic group escalates two LPN reqs a quarter on its current process, the avoided premium alone is the number to set against the platform's annual price. Run that arithmetic with your own bill rates before the renewal conversation, not after.
A pilot can also fail, and that's useful. If the pilot arm produces more applicants but the qualified-screen count barely moves, the problem is likely posting reach in your local market or pay competitiveness, and no sourcing tool fixes either. For more on where sourcing tends to break once applicant volume rises, see why healthcare talent sourcing breaks at volume.
What to check before you sign
The pilot tells you whether the platform works; the contract review tells you whether it will keep working once the pilot team moves on. Small clinic groups should confirm a few things in writing.
- Where candidate data lands. Shortlisted candidates should flow into your ATS automatically. If the clinic has no ATS, confirm the platform includes one that can post jobs, keep a talent pool and schedule interviews.
- Which boards are included. Ask for the list of job boards connected, and whether healthcare-specific boards are part of it.
- Who owns the hiring decision. Screening should collect structured answers and surface qualified candidates; the clinic manager and the recruiter decide who advances and who gets the offer.
- How candidates experience the screen. Take the screen yourself as an MA applicant on a phone at 10 p.m. If it feels like a form that never ends, your candidates will feel the same.
- What the price assumes. Confirm whether pricing scales by req, by seat or by applicant, and model it against your next 12 months of openings.
A longer buyer checklist lives in what to look for in a talent sourcing platform. For the matching side of the same problem, see how resume matching holds up when nurse applicant volume spikes.
Where Eximius fits in a clinic pilot
Eximius is built to be measured this way. It posts to the boards clinicians use, including Indeed, ZipRecruiter and Vivian for healthcare roles, and captures applications from your career page through a webhook, forward-to email or application URL. Resume matching ranks applicants and past candidates against each req using semantic and keyword signals, so the recruiter reviews the strongest fits first.
Sia, the Eximius screening agent, runs a structured screening conversation over chat, voice or video with every applicant, collecting answers against the criteria the clinic sets: license, certifications, shift availability, commute. The recruiter reviews those answers and decides who moves forward. Scheduling removes the email back-and-forth for interviews, and shortlisted candidates push into your ATS (Greenhouse today) or into the barebones ATS Eximius provides for clinics that do not have one.
What Eximius does not do is decide who to hire. It gives the team better inputs, sooner, so the people who know the clinic make the call while the candidate is still available.
What to do this week
- Write the qualified definition for each clinical role you have open: license, certification, shift and commute. One line per role.
- Pull last quarter's baseline for qualified screens per req, median time to first contact and agency escalations, even if the numbers are rough.
- Pair your open reqs into two comparable arms by role and location, and confirm hiring managers in both arms can interview within five days.
- Copy the scorecard above into a shared sheet and set pass thresholds now, before any vendor sees your data.
- Price one agency escalation using your actual bill rate and direct cost, so the pilot result translates into dollars the founder or CFO will recognize.
The decision a clinic group faces is not whether to buy sourcing technology. It's whether a specific platform lowers the cost of a filled clinical role on your reqs, in your market, at your response speed. Set the thresholds first, run both arms for 30 days, and let the agency line on next quarter's P&L settle the argument.
Want to see how structured screening changes time to first contact on your clinical reqs? Book a demo and we'll run your next role through the Eximius workflow.




