Your Recruiter Sourced 340 Candidates and the Hiring Manager Reviewed 14

It is Thursday, September 3, 3:14 PM. Your VP Engineering opens the ATS. The staff engineer requisition posted July 14. 340 candidates sit in the pipeline. The recruiter reached out to 187 on LinkedIn, 92 replied, 41 booked an initial screen. The hiring manager reviewed 14 profiles across eight weeks. Twelve sent a take-home. Four made it to the onsite loop. Two got offers. One accepted, ghosted week two of the start ramp. Role reposted last Monday.

The requisition is 61 days old. The eng org needs the seat by October to unblock the payments rewrite. The recruiter closed two other roles in the same window and worked a director search in parallel. The candidates who replied on the second week of the sequence never got a follow-up because the recruiter was mid-negotiation on a data role. The one who ghosted the ramp asked twice for equity clarity in week one and got a Notion doc last edited in April.

Pull the trailing four quarters of open roles. Count the median time-to-close. Most Series B books count 74 to 118 days for a senior IC and 140 to 210 for a Head-of. Count the offers that closed above the top of band. Most count one in three. Count the candidates in the pipeline the hiring manager never opened. Most count 240 to 480 for a staff-level requisition. Talent is not a sourcing problem. Talent is a queue.

The 340 candidates nobody screened

Walk the pipeline. 187 sourced through LinkedIn Recruiter, 92 replied, 41 booked a screen. 82 came inbound from the careers page, 24 got a rejection auto-email in week one, 18 got a screen, 40 sat in the "new applicant" bucket seven weeks. 71 came from the referral portal, 34 got a screen, 19 moved to the tech round, six made an onsite loop. The 340 profiles carry a trailing signal on every stage nobody read.

The recruiter shipped the first sourcing sequence on July 15 and moved to the director search on July 22. The follow-up template goes silent after touch three. The hiring manager gets a Slack ping when a candidate books the screen and reads the profile at 8:47 AM the morning of the call. The tech screen scorecard lives in a Google Sheet the recruiter opens twice a week. The offer letter template still points to the old option pool.

The team that should own this knows it is broken. The recruiter runs two other requisitions in parallel and cannot refresh every sourcing sequence when a candidate replies on the third touch. The hiring manager reviews the profile the morning of the screen because the ATS notification lands the day before. The Head of People builds the offer package the day the loop closes because the compensation model refreshed in June and got hardcoded on August 4. The 61-day-old requisition is a $260K to $410K sunk cost on the payments rewrite the seat exists to unblock.

Hiring a Head of Talent is the slow answer

The textbook fix is a Head of Talent, a Senior Technical Recruiter, or a talent partner with a time-to-close number. Loaded comp in the US runs $160K to $240K a year for the senior recruiter, $220K to $320K for the head-of, with an agency search on the same role costing 22 to 30 percent of first-year cash. Months one through three go to rebuilding the sourcing sequences, refreshing the scorecard library, and killing the 40-profile "new applicant" bucket sitting seven weeks stale. Months four through nine are when time-to-close moves from 96 days to 58 on senior IC roles, reply rate on outbound moves from 22 percent to 41, and the ghost-in-week-two number drops toward zero.

The fractional version is faster and stops at the same wall. Eight to twelve thousand a month buys twelve to sixteen hours a week of senior recruiting work. The first month rebuilds the sourcing template and clears the inbound backlog. The 340-profile pipeline keeps drifting because a fractional recruiter cannot refresh every sequence on the day a candidate replies, brief every hiring manager on the profile before the screen, chase every reference call and offer clarification the same afternoon, refresh the compensation model against every posted role on the current pool, and personalize every ghost-recovery email against the specific reason the candidate walked.

Both versions assume the work is a person opening an ATS on a cadence. The work itself is sourcing on the new job spec by Monday, refreshing the outbound sequence against every candidate reply within two hours, screening the inbound stack every morning against the trailing four hires and the current scorecard, briefing the hiring manager 30 minutes before every screen with the profile, the trailing role progression, the referral trail, and the two questions the candidate flagged, drafting the offer package on the day the loop closes against the current pool and the current stock price, and running a start-week check-in on every accepted offer. On four open roles and a 340-profile pipeline that is 45 to 60 hours a week of senior recruiting work. No single hire clears the pile and holds the time-to-close number at the same time.

What a fractional AI talent function owns

Hand the ATS, the LinkedIn Recruiter seat, the referral portal, the sourcing sequence library, the scorecard library, the compensation model, the option pool cap table, the trailing four quarters of closed roles, and the current requisition set to a fractional AI agent. The agent does the work a Head of Talent, a Senior Recruiter, and a Head of People pod would do together. The cadence is per-reply on outbound, per-morning on inbound, per-screen on the hiring manager brief, per-loop on the offer package, per-week on the pipeline health digest.

Every reply routed within two hours. A candidate replies on touch three at 11:47 AM. By 1:34 PM the sequence sends a personalized follow-up naming the specific project the candidate would own, the trailing scoping doc from the last standup, and three time slots on the hiring manager's calendar. Reply rate on the outbound sequence stops depending on which requisition the recruiter opened this morning.

Every inbound triaged before 10:00 AM. The careers page ships 47 applications overnight. By 9:14 AM the queue names the eight profiles matching the current scorecard on the staff engineer role, the four matching the senior product manager role, and the 35 that hit the auto-reject on years-of-experience or location. The 40-profile stale bucket stops existing.

Every hiring manager briefed 30 minutes before the screen. The screen books for 2:00 PM Thursday. By 1:30 PM the brief lands in Slack with the profile summary, the trailing role progression, the referral trail, the trailing two GitHub or portfolio artifacts on the stack, and the two questions the candidate flagged in the reply. The hiring manager stops reading the profile the morning of the call.

Every offer package built the day the loop closes. The onsite loop closes Friday at 4:00 PM. By Friday at 5:47 PM the offer package refreshes against the current option pool, the trailing 90 days of comparable offers, the strike price on the last 409A, and the vesting cliff on the current template. The Head of People stops rebuilding the model on Monday.

Every Monday a pipeline health digest. Monday, 9:00 AM. The digest names the six roles open, the median days-in-stage on each, the three candidates at risk of ghosting on week two of the sequence, and the two references the recruiter should call today.

The unit economics of a role that stayed open six months

A Series B company at $28M ARR running four open roles against a 340-profile pipeline is burning three specific things. The VP Engineering, the recruiter, and the Head of People spend a combined 18 to 28 hours a week on sourcing reviews, scorecard rewrites, and offer package rebuilds against a fully loaded hour of $180 to $290. That is $13K to $32K a month of senior time on work a live routing engine clears. The recruiter bench gets six to nine hours a week back inside the first sprint. Read the case studies for the shape of that trade on a similar bench.

The requisition line is the second one. A staff engineer seat open 118 days on the payments rewrite delays the launch by a quarter, defers $340K to $620K of infrastructure savings, and holds the sales team on the old rate card. Moving time-to-close from 96 days to 58 across two quarters compounds against the operating plan on four open roles and a $260K to $410K loaded cost per seat delayed. One rewrite that ships in October instead of January pays for the sprint and the ongoing spend twice over.

The candidate line is the third one. A hire who ghosted week two of the ramp on an unclear equity story costs the org the $52K signing bonus, the recruiter fee if the search was agency, and the six weeks of engineering time the hiring manager spent in the loop. A 14-day sprint to stand up the agent runs in the low to mid five figures. Ongoing cost lands closer to two Gem seats than a Head of Talent hire. The outbound reply routing and the inbound triage ship in week one. The hiring manager briefing engine and the offer package builder ship in week two. The Monday pipeline health digest lands off a live feed before the sprint closes.

What changes after the sprint

Picture the same Thursday, 3:14 PM moment, thirty days after the sprint ships. Your VP Engineering is not scrolling the ATS. The staff engineer role closed on day 58 with a candidate the outbound sequence reached on touch three and briefed against the payments rewrite scoping doc. The offer package landed on the candidate's inbox Friday at 6:00 PM with the option pool math, the vesting schedule, and a Loom from the founder attached. The candidate signed Sunday. Start date is October 6.

By Monday the digest reads four open roles, median 54 days-in-stage, one candidate at risk on week two of a director loop, and two references to call today. The recruiter opens the ATS on Tuesday and reads a queue of eight profiles the agent triaged overnight instead of the 47-profile inbound bucket. The Head of People closes the option pool refresh loop by comparing the trailing 90 days of offers against the current 409A instead of rebuilding the model on Monday.

If your ATS currently holds a 340-profile pipeline, four requisitions past 61 days, and one hire who ghosted week two on an equity story that never got clarified, the version where every outbound reply routes in two hours and every hiring manager reads the brief 30 minutes before the screen is fourteen days away. Talent is a function. You can hire against it, you can retain a fractional partner for it, or you can scope a sprint and have it running this month. The 340 profiles are already in the ATS. The math is whether the recruiter opens each one on the day the candidate replies or the day the requisition reposts.