Call every candidate in flight before you post anything. The recruiter seat takes about eight weeks to fill and a month to ramp, and the candidates go cold in days, so week one belongs to them. Rank the six roles by what an empty week costs, send the two that cannot wait outside, and decide the seat in week two.

Our only recruiter just quit

That was the whole question, so here is the answer in the order I would do it. Nothing about the six roles is on fire in the way it feels this morning. What is on fire is smaller and more specific: the people the recruiter was mid-conversation with, and the fact that the pipeline lives in their inbox.

Here is the example I will work through. Six open roles at a base around $160,000: a senior backend engineer on the payments launch, a platform engineer, a product manager, an account executive, a product designer and a data engineer. Fourteen candidates somewhere between a first call and an offer. Hiring managers who have interviewed plenty and never once run the process.

Business day What happens Who
1 Get into the recruiter's inbox, calendar and tracker; list every live candidate with stage and last contact You
1 to 2 Call all 14, offers and final rounds first; give each a name and a date Hiring managers
3 Move the pipeline out of the inbox into one tracker with stages per role You
4 to 5 Rank the six roles by the cost of an empty week and by live candidates You and the managers
6 to 7 Write the brief for the two roles going outside, post them, set the fee You
8 to 10 Weekly cadence on the other four; decide whether the seat gets backfilled You

Do not post the recruiter's role first

The standard advice, and the instinct, is to post the recruiter's role the same afternoon and backfill the seat before touching anything else, because everything else depends on having a recruiter. I understand the instinct. I think it is backwards, and the calendar is why.

Filling a recruiter seat is a search like the six you already have. Post it this week and, in my experience, an accepted offer is six to eight weeks out, the notice period adds two to four, and the first month is ramp: learning what senior means here, rebuilding six pipelines from a tracker someone else kept. The earliest the new hire moves any of the six roles is around week 12.

Recruiters are available this year, which helps: ERE's 2026 report on talent acquisition careers notes that the BLS category it uses for recruiters shrank in 2025 for the first time, by 0.5%, so the seat is fillable. It is just not fillable fast.

The 14 candidates decay at a different speed. A candidate who had a final round on the calendar and then hears nothing for a week assumes the worst and takes the other offer. And every week the seat stays empty on the two roles that cannot wait costs real money, which I put numbers on below, and the numbers are bigger than a fee.

So the order is the candidates, the inbox, the ranking, the two roles, and the seat in week two, once you can see whether the six roles are a steady state or a one-time push. The signal is the plan: eight or more hires in the next 12 months, and again the year after, and you backfill the seat; if these six are the year, rent hours or pay per hire, and do not rebuild a function for a push.

The first two days are for candidates in flight

Get into the recruiter's inbox and calendar on day one. If they left on good terms, a 30-minute handover call is worth more than anything else you will do this week, and it is worth asking for even if they did not.

List every live candidate: name, role, stage, last contact, next promised step. In the example that is 14 people, and the ones who matter most are the product manager at offer stage, the payments engineer at final round, and anyone who was told "we'll get back to you Friday".

Then call them. Not email; call. Fourteen calls at 20 minutes is about five hours, split across the hiring managers, and each call says three things: who their contact is now, what the next step is, and the date it happens. Offers first, final rounds second, the rest by last contact.

A candidate who hears from a hiring manager within two days of the recruiter leaving mostly stays. One who hears nothing for a week is mostly gone, and you will not know which until you call.

These are also the cheapest hires you will make this quarter. SHRM's 2022 write-up on the real costs of recruitment put the average cost per hire near $4,700, and the sourcing share of that is already spent on the 14. A pipeline you let go cold is money spent twice.

While you have each of them on the phone, ask what they were told about comp, start date and the loop, because a recruiter's promises leave when the recruiter does, unless someone wrote them down. Write them down.

Getting the pipeline out of the recruiter's inbox

If the tracker was the recruiter's inbox, a spreadsheet, or an ATS only they logged into, day three is for moving it somewhere the hiring managers can see. One place, one row per candidate, stages per role, and the people already rejected marked as rejected so nobody re-contacts them next month. It is boring. It is also the step that stops the next resignation from being the next disaster.

If you have no ATS, or yours was the recruiter's own subscription, the ATS on Chosen is free and takes a spreadsheet as a bulk import. A person at Chosen does the migration from another ATS while the old one keeps running, so your candidates, roles and stages come across and nothing restarts from zero. That is the plumbing. It has to be there before outside recruiters can submit into it, and it has to be there anyway.

Rank the six roles by cost of delay

With the candidates called and the pipeline visible, rank the roles. Two numbers per role: how many live candidates are past a first call, and what a week with the seat empty costs. The second number is rough by nature, and a rough number beats a feeling.

Role Live candidates An empty week costs Decision
Senior backend, payments launch 4, one at final round About $11,500 of delayed launch revenue Send outside
Account executive 2, both early About $17,300 of quota Send outside
Product manager 3, one at offer Hard to price; the offer is the work Close it yourself
Data engineer 2 Reports slip; nothing sold slips Keep in-house
Platform engineer 2 Roadmap, not revenue Keep in-house
Product designer 1 A contractor covers it Pause for a quarter

Here is the arithmetic behind the first two rows. The account executive carries a $900,000 quota; divided by 52, an empty week is about $17,300 of bookings nobody attempted, and even after discounting a ramp quarter the number stays large. The payments launch is tied to about $50,000 a month of new revenue, so a week of slip is about $11,500.

A placement fee at $160,000 base and 20% is $32,000. Outside help does not fill a role overnight, but it saves the eight to twelve weeks the backfill route costs, and eight weeks of either of those two roles empty is $92,000 to 138,000 of delay against a $32,000 fee that is only paid if someone starts.

The rule that falls out: send a role outside when a week of delay costs more than the fee divided by the weeks you would save, which here is $32,000 over eight weeks, or $4,000 a week. Both of the top two clear it by a wide margin. The product manager stays in-house for a different reason: there is an offer on the table, and the work now is closing, which is yours.

The two roles to send outside

The brief is what the recruiter had in their head, and it has to be on paper before anyone outside can work the role. That means the five to seven requirements that actually matter, ranked, with the must-haves marked; the comp band; the companies you would love to hire from and the ones you would not; what the last two hires looked like; the interview loop and who runs each step.

Two hours per role, with the hiring manager in the room. A brief that is a wish list gets you 20 resumes that are all a little wrong.

Then choose the shape of the help. An agency quotes you a percent of first-year base and its own terms, and works the role from its own database. A recruiter by the hour bills whether or not anyone starts, which is fine for a long list of roles and expensive for two hard ones.

On Chosen you post both roles and set a fee on each, owed when a hire starts; the wizard suggests 20%. A person at Chosen approves recruiters onto each role and reviews every packet before you do; you commit to a decision on each within 5 business days, the one habit that falls to the hiring managers. It is small, 13 approved recruiters in September 2026; keep your pipeline moving. Chosen does not guarantee that any role will be filled.

Get into the inbox today, make the 14 calls in the next two days, and leave the recruiter's role unposted until week two, when you will know whether the six are a one-time push or a steady state.