The Unit Economics of Outbound Recruiting: A Model You Can Actually Run

Almost every published cost model for outbound recruiting makes the same mistake: it treats the tooling line item as the interesting variable. In practice, tool spend is usually the smallest term in the equation, and optimising it is close to irrelevant. The dominant term is loaded recruiter time — and the variable that controls loaded recruiter time is shortlist precision.

This article builds the model explicitly. Every number below is a stated assumption, not a research finding. The point is not to hand you benchmarks; it is to hand you a structure you can populate with your own data, because the shape of the model is stable even when the inputs are not.

The funnel, stated as variables

Outbound recruiting is a six-stage funnel. Naming the stages precisely matters, because most disagreements about recruiting efficiency turn out to be disagreements about stage definitions.

Stage	Variable	Definition
Sourced	S	Profiles retrieved by a search, before human judgement is applied
Qualified	Q	Profiles that survive review against the real hiring criteria
Contacted	C	Qualified profiles that actually receive outreach
Replied	R	Contacts producing a substantive response, positive or negative
Screened	N	Replies converting to a first conversation
Hired	H	Screens converting through the process to an accepted offer

The ratio that does the most work in this model is precision, defined as p = Q / S. It is the share of what your search returns that a competent recruiter would actually contact. Precision is not a vanity metric; it is a direct multiplier on review labour, which we will see shortly.

Why "qualified" has to mean something specific

A profile is qualified only if it satisfies the criteria that actually drive the hiring decision — including the ones that are not expressible as database filters. Seniority, location and current title are filterable. "Has operated at this scale before", "has shipped in a regulated environment", "has done this without a large existing team" are not, yet they are frequently the criteria that determine whether a hiring manager takes the call.

This distinction is the reason precision is usually far lower than teams expect. A search that filters cleanly on structured fields can still return a majority of profiles that fail on the unstructured criteria — and those failures are invisible until a human reads the profile.

The cost equation

Total cost of an outbound hire decomposes into three terms:

Total cost = Sourcing cost + Review cost + Engagement cost

Sourcing cost   = S × (cost per sourced profile)
Review cost     = S × (minutes per profile review) × (loaded hourly rate / 60)
Engagement cost = C × (minutes per outreach + follow-up) × (loaded hourly rate / 60)
                + N × (minutes per screen) × (loaded hourly rate / 60)

Note what appears in the review term: S, not Q. You pay review cost on everything the search returns, including everything you reject. This is the structural reason low precision is expensive — rejection is not free, it is the main product of a low-precision search.

Worked example

The following inputs are illustrative assumptions for a mid-market senior engineering role. Substitute your own.

Input	Assumed value
Loaded recruiter cost	$60 / hour
Cost per sourced profile	$0.02
Review time per profile	2 minutes
Outreach + follow-up time per contact	6 minutes
Screen length	30 minutes
Precision (Q / S)	20%
Contact rate (C / Q)	90%
Reply rate (R / C)	15%
Screen rate (N / R)	40%
Screen-to-hire (H / N)	12%

Working backwards from one hire: one hire requires about 8.3 screens, which requires about 21 replies, which requires about 139 contacts, which requires about 154 qualified profiles, which at 20% precision requires about 772 sourced profiles.

Cost component	Calculation	Cost per hire
Sourcing	772 × $0.02	$15
Review	772 × 2 min × $1/min	$1,544
Outreach	139 × 6 min × $1/min	$834
Screening	8.3 × 30 min × $1/min	$249
Total		$2,642

Sourcing spend is 0.6% of the total. Review labour is 58%. Any conversation about outbound efficiency that focuses on the first number and ignores the second is optimising the wrong variable.

Sensitivity: which levers actually move the number

Holding everything else constant, here is what a meaningful improvement in each variable does to cost per hire in this model.

Change	New cost per hire	Delta
Baseline	$2,642	—
Sourcing cost per profile falls to zero	$2,627	−0.6%
Review time drops from 2 min to 1.5 min	$2,256	−14.6%
Reply rate rises from 15% to 18%	$2,437	−7.8%
Precision rises from 20% to 40%	$1,870	−29.2%
Precision to 40% and reply rate to 18%	$1,706	−35.4%

Two observations follow directly.

First, eliminating sourcing cost entirely is worth less than a rounding error. Negotiating your data vendor down is not a lever worth much executive attention.

Second, precision is the only single variable that produces a step change. Doubling precision halves the volume of profiles requiring review while holding qualified output constant. The review term collapses accordingly. Precision has a twin that this model does not capture — recall, the share of qualified people your search never surfaced — and the two behave very differently; we treat that separately in precision and recall in shortlists.

The compounding effect that the table understates

The sensitivity table treats precision and reply rate as independent, which understates the real effect. In practice they are correlated: a higher-precision shortlist means outreach goes to people who genuinely fit, which supports more specific messaging, which tends to raise reply rates. The last row of the table — precision and reply rate improving together — is the more realistic outcome of a precision improvement, not an additional separate win.

There is a third-order effect as well. Fewer, better-targeted conversations reduce time to hire, and time to hire carries its own cost in the form of unfilled-role productivity loss. That cost sits outside this model but frequently exceeds every term inside it.

The failure mode: buying volume and calling it coverage

The most common way teams make outbound economics worse is by adding sourcing volume without adding precision. It feels like progress — the top of the funnel grows, dashboards show more activity — but the model shows what actually happens. Review cost scales with S. If precision is unchanged, doubling S doubles review cost and doubles qualified output, leaving cost per hire flat while consuming twice the recruiter capacity.

Worse, review quality degrades under volume. A recruiter working through 1,500 profiles applies less judgement per profile than one working through 400, which pushes effective precision down. Volume-without-precision is not neutral; it is often negative.

Instrumenting this in your own team

You need six counters and one rate. Most teams already have four of them in an ATS and are missing exactly the two that matter.

1. Sourced (S) — how many profiles the search returned. Frequently untracked, because searches are run ad hoc and results are never counted.
2. Qualified (Q) — how many survived review. Requires recruiters to record rejections, which is the single most commonly skipped step.
3. Contacted, Replied, Screened, Hired — usually available from the ATS or sequencing tool.
4. Loaded hourly rate — fully loaded cost, including benefits and overhead, not base salary divided by 2,080.

If you track nothing else, track S and Q. Precision is the variable with the highest leverage and the lowest measurement cost, and almost nobody measures it.

If you are a team of one to three, the loaded-rate framing above is more apparatus than you need. The same conclusion falls out of a much simpler worksheet in what sourcing actually costs a small team.

A caution on benchmarking

Resist the urge to compare your precision figure to anyone else's. Precision depends on how strictly you define "qualified", and that definition is not standardised across teams. A team with 15% precision and a strict definition may be running a healthier funnel than a team reporting 45% with a loose one. The number is useful as a time series within one team and largely meaningless across teams.

What this implies for tooling decisions

The model gives a clean evaluation criterion. A sourcing tool is worth adopting if it reduces total cost, which in practice means one of two things:

• It raises precision — the same qualified output from fewer reviewed profiles.
• It reduces review minutes per profile — for instance by surfacing the evidence for each criterion so a recruiter can verify in seconds rather than reading a full profile.

Tools that do neither, but return more profiles faster, are increasing the input to your most expensive stage. The sticker price is not the relevant number; the effect on review hours is.

This is the design constraint Hiris is built around: turning hiring criteria — including the ones that are not filterable — into a shortlist where each profile arrives with the reasoning and proof links attached, so verification is a matter of seconds rather than a full profile read. The economics above are the reason that specific problem is worth solving rather than simply returning more results.

Summary

Run the model with your own numbers before you run a tooling process. In most funnels you will find that sourcing spend is a rounding error, review labour is the dominant cost, and precision is the variable with by far the highest leverage. That conclusion is robust to a wide range of inputs — which is precisely why it is worth checking against yours.