Source Outbound When You Need Evidence, Not When You Need More People

Which do you need, outbound or inbound?

Whichever one matches the shortage you actually have. Being short of applicants and being short of information about the applicants already in front of you look identical on a stalled requisition and have opposite remedies. Outbound adds people at the top. It does nothing about the middle, where a recruiter is now reading roughly three times the volume per hire that they read in early 2021 and interviewing a smaller share of it ¹.

Two numbers size that middle. Ashby's analysis of over 109 million applications and 247,000 jobs puts the average at 291 applications per hire against roughly 100 in early 2021, with the share of applications that reach an interview falling from about 7 to 8% down to between 3.6% and 4.7% depending on role type ¹. Read the caveats with the number: that is one applicant tracking vendor's customer base, weighted toward venture-backed technology employers hiring knowledge roles, and the count includes duplicate, automated and fraudulent submissions, which the vendor names as a driver without sizing. It describes nothing about hourly, clinical, field or public-sector hiring, where a shortage of people is often exactly what is happening.

So the decision rule is local, and you can run it this week:

• Applications arriving, and you cannot separate them. Outbound will not touch this. The constraint is evidence, and the money belongs in a stage that produces some.
• Applications not arriving at all. Check the form and the pay before buying outreach, then source. Both are cheaper than a seat.
• Applications arriving from outside the population you need. This is the genuine outbound case: the people who can do the work exist, they are employed, and nothing you post will reach them.

Why do sourced candidates convert better?

Because a human already screened them. An outbound candidate was read, chosen and approached by a recruiter before entering the funnel; an inbound candidate was not. Comparing the two on funnel conversion measures that first act of selection and then credits it to the channel. The same gap would show up if the outreach message were blank, which is the test no vendor page runs.

Stage conversion is a treacherous number generally, and the shape of the treachery is visible in the benchmarks. Recruiter screens pass roughly 35% of the candidates who reach them, while 95% pass through post-onsite and 81% pass through the offer stage ². Take that 95% as a timing fact: by the time an onsite ends the decision has effectively been made, so the stage does very little independent filtering. The heaviest filtering happens earlier still, at application review, which is not one of those numbers at all ².

Read that way, a high conversion rate marks where a decision has already been taken. It is a fact about the stage. Any comparison that puts one channel against another has to be measured at a point where both are treated identically, which no stage before your first evidence stage manages, because everything before it is somebody's reading of a document. The volume problem underneath that is worked through in why applications per opening tripled and how many are real.

A conversion multiple moves with the screening in front of it, so it cannot carry a budget decision. Approve a sourcing seat on a shortage you have named, with a number attached that would tell you the shortage had ended.

What does outbound actually buy?

A conversation before the funnel starts. Outreach is the one moment in hiring where a candidate can ask what the band really is, how many rounds there are, who they would report to and when a decision gets made, and get an answer from somebody allowed to give one. That exchange is the product. Access to people you could not otherwise reach is the half that got cheap.

That reframing changes what a good outreach message contains. Three things earn a reply, and a pasted job description is none of them: the specific piece of the person's work you actually read, the one sentence of the role that would matter to somebody already employed, and the number of rounds with the pay range stated up front. Everything else is filler, and the filler is why response rates are what they are.

It also changes who should do it. A conversation about comp, level and timeline is a conversation the hiring manager or a recruiter with authority can have. A coordinator sending templated messages is buying the cheap half and none of the expensive half.

Before any of that, though, check the application form. Appcast's 2025 benchmark, built on over 281 million clicks and 25.6 million applies across more than 1,300 United States employers, puts the median apply rate on long-apply ATS ads at 6.1% against a 19.37% average on one-click easy-apply ads ³. Those two rates are not a controlled comparison: different boards, different traffic and different intent, so the roughly threefold gap is an upper bound on what shortening a form would buy. Both are apply rates on paid job ads, which is not the rate on your own careers page. It is still the cheapest thing on the list, and it is the one nobody checks before approving outreach. What replaces a keyword filter once the volume arrives is a different question, taken up in what replaces an ATS keyword filter when every resume matches.

Run one pass rate across both channels

Measure both on the share of candidates who clear your first evidence stage, not on the share who apply, reply or accept a screen. It is the only number where inbound and outbound are treated identically, because by that point both have done the same work under the same rules and a reviewer is reading what each of them produced. Compute it for the last two requisitions before the next renewal.

Three results, three decisions:

1. The rates are the same. Outbound is a volume tactic at your company. Fund it or cut it on that basis alone. 2. Outbound passes higher. The recruiter's pre-selection is adding real information. Keep the seat, and then do the interesting thing: write down what that recruiter is selecting on and apply the same criteria to inbound review, where it is currently free. 3. Outbound passes lower. The outreach is chasing titles and current employers. This is common and fixable: rebuild the search string out of the tasks the role actually does.

The number needs a denominator you can defend, so count candidates who were invited to the stage, whether or not they completed it. Completion rates differ by channel for reasons that have nothing to do with capability: a candidate approached about a role they were not looking for drops out more readily than one who came looking.

And keep the comparison running. Channel advantage moves with the market, with the role and with whoever is doing the outreach this quarter, so a single reading tells you about one requisition.

See a sample report >