Founders usually hire their first recruiter about two quarters later than they should. The signal is not headcount. It is that hiring has become a permanent, unmanaged draw on the founding team's calendar.
The three signals that say "now"
Volume with continuity. Eight or more open roles per year, repeating rather than one-off. Agencies are efficient for spikes; they are expensive for a steady state.
Founder time. If two founders spend six or more hours a week on sourcing, screening and scheduling, an in-house recruiter is cheaper than the alternative use of that time — and better, because the pipeline knowledge stays.
Candidate experience is slipping. Delayed feedback, inconsistent interview loops and offers that go stale. That damage compounds in a small market.
Run the arithmetic honestly: agency fees at 8–12% of annual salary across the roles you expect to fill in twelve months, against a recruiter's fully loaded cost plus tooling. Somewhere around six to eight hires a year, in-house usually wins — and wins more once you count process quality.
What to hire for at this stage
Do not hire a recruitment coordinator and expect a hiring engine. Do not hire a TA director and expect them to source at midnight. At your first recruiter, look for a builder-operator: someone who will run the requisition, source directly, and also write your interview kits.
A usable scorecard has four sections.
Outcomes. Close 18–24 roles in the first year across engineering, sales and operations. Median time to fill under 40 days. Offer acceptance above 80%. A working ATS with clean data.
Competencies. Direct sourcing skill (they can build a Boolean string live), structured interviewing, stakeholder management with opinionated founders, written communication, and comfort with an ambiguous employer brand.
Signals to interview for. Ask them to source live against a real role and narrate their reasoning. Ask what they changed in a hiring process and what happened. Ask for the last role they failed to close and why — the honest answer is diagnostic.
Anti-signals. Only ever worked with inbound applicants. Cannot name their own metrics. Describes hiring managers as an obstacle rather than a partner. Volume claims with no quality context.
Interview loop we recommend
Four stages, three hours total: recruiter screen for motivation and market knowledge; a 45-minute live sourcing and role-intake exercise; a hiring-manager conversation on partnership and pushback; and a founder conversation on judgement and bar-setting. Score against the scorecard immediately after each stage, independently, before discussion.
The first 90 days
Days 1–30: build the base. Sit in on three functions' work to understand what they do. Write the intake template. Set up or clean the ATS with stages that reflect your real process. Define one metric set and publish it weekly.
Days 31–60: fix the funnel. Structured scorecards for every stage of your top three roles. Interviewer briefings — most of your interview quality problem is untrained interviewers, not bad candidates. Two sourcing channels tested with numbers attached.
Days 61–90: prove the engine. First hires closed through the new process. A referral mechanism that runs without nagging. A quarterly hiring plan agreed with each function, with a named hiring manager per role.
How to know it is working
Fewer founder hours on hiring, not more meetings about hiring. A candidate pipeline you can read without asking anyone. Interview feedback within 48 hours as a norm. And, by the end of quarter two, a hiring manager saying that a role got filled better than they expected — which is the only endorsement that matters.
If the first ninety days produce only forwarded resumes, you did not hire the wrong person; you hired the right person into an undefined job. Write the scorecard before the job post, and both problems disappear.