A different way to look at the cost of an open role

Applications are up 412%. Your hiring signal isn't.

Most conversations about hiring start with the candidate. This one starts with what the open seat is actually costing you. It takes about two minutes, and there's a calculator below.

Every one of those applications has to be screened, reviewed, interviewed or turned down by somebody on your team. It costs the candidate almost nothing to apply, and these days an AI agent can fire off a resume to four hundred companies in an afternoon. Your team is the one who eats the cost, on every single one.

Your team is spending expensive hours filtering cheap applications. That's the real cost, and it's not even the expensive part yet.

412%
increase in applications per recruiter since 2022. Our analysis, published April 2026
254
applicants on the average technical posting. Some roles clear 700
250+
applications can enter the funnel for one role. Your team still has to find the few worth pursuing
0
matching rise in hiring. Hiring did not rise with application volume
Why more applicants made it worse

The applications aren't the real problem.

The problem is what they force your organization to do. Volume arrives as work, and it gets more expensive the further down the funnel it travels. More applicants don't create more talent. They create more decisions.

The cheap one

The bad applicant.

Wrong stack, wrong level, wrong country. Costs you two minutes and a rejection. Now multiply it by several hundred.

The costly one

The plausible applicant.

Passes the filter, earns a screen, sometimes a full panel. Your engineers pay for this one in hours rather than minutes.

The expensive one

The one you make an offer to.

Qualified, impressive, and never going to join you. Costs a full loop, comp approval, and four weeks you don't get back.

Many of the people you would most want are not applying at all. They have to be recruited.

This isn't an argument that your talent team is doing something wrong. The cost of applying collapsed. The cost of filtering shifted to them.

What a large posting actually contains
One senior posting, roughly 750 applications, illustrative
310Submitted by an AI agent. Never read the description.
190Wrong stack or wrong seniority.
95Not eligible for the location.
95Qualified, but applied to you as a hedge. Wants a household name, a public-company package or fully remote.
35Qualified, available, authorized to work without sponsorship. Now ask how many are still on the market by the time you call, will accept your band, and are the person you would actually choose.
25Qualified, but on an H-1B that has to transfer to you. Roughly $10,500 to $13,000 per hire in government and legal fees with premium processing, no lottery. If their green card has not cleared the I-140 stage, the PERM process restarts under your name: $10,000 to $20,000 over several years, and they can walk during it.
Fees as of August 2026. Premium processing $2,965. H-1B transfers are exempt from the $100,000 proclamation fee, which is in any case blocked pending appeal.

The proportions move with the role, the level and the location. The shape holds. Your organization has to process the entire pile to reach the last two bands, and what is left is smaller, slower and more expensive than the headline count suggests. The four or five people you would most want are not in the pile at all.

Model your signal tax

You already know hiring is expensive. The question is how much of it is wasted.

Your facts on the left, our assumptions underneath. Change anything you disagree with; nothing is sent anywhere.

Model your signal taxOne year, one organization
Your organization
Model assumptions Adjust anything that looks wrong
Capacity consumed
17 workweeks
of recruiting and engineering capacity, across 12 roles.
Seventeen of the fifty-two weeks in your year.
Internal labor, under these assumptions
$83,380
People time only. Enable a vacancy cost to include the empty seat.
Applications processed3,000
Screening hours100
Engineering interview hours576
Offers made per hire1.54
Days a role stays open55
Your assumptions, your arithmetic. Decline delay is a separate input so it is never counted twice inside time to fill.
See the math

Screening hours are roles multiplied by applications multiplied by minutes, divided by sixty, costed at the recruiting rate. Interview hours are roles multiplied by loops multiplied by engineering hours per loop, costed at the engineering rate.

Offers per hire is one divided by one minus your decline rate. Days open is your base figure plus the extra offer cycles multiplied by the days a decline adds. The base is deliberately defined as time to fill when the first offer is accepted, so the decline penalty is never already baked in.

Vacancy cost is roles multiplied by days open multiplied by the daily cost, and it is off by default because most teams cannot defend a number for it. The result stands without it.

The part that isn't on the invoice

But the money isn't the part I'd worry about.

$83,380 is absorbable. Most companies your size already absorb it, inside a budget nobody audits by the hour. 17 workweeks is finite capacity you don't get back, and that is where those hours came from.

1
The trade you didn't agree to

What didn't get done while your best people were hiring people you didn't hire?

You didn't budget for this. It accumulated one résumé, one screen, one interview panel and one declined offer at a time.

An empty seat isn't a recruiting problem. It's an execution gap.

The recruiting hours are the smaller number. The larger one is what this person was hired to accomplish and hasn't.

2
The reason the req exists

What was this person supposed to accomplish?

Who picked up the work?

There isn't a good answer here. That is the point of the question.

Your best engineer did.

Two jobs now, and one isn't the job they were hired for. Also the person most able to leave.

You did.

You're solving execution problems instead of leading the organization. Your calendar shows it.

The team shared it.

Everyone is more fragmented, nothing has a clear owner, and velocity drops with no single visible cause.

Nobody did.

Then the work simply didn't happen, and the plan quietly changed without anyone deciding to change it.

Six months, side by side

Your empty seat doesn't exist in a vacuum.

Your competitors are still hiring, shipping and iterating. The vacancy is only half the story. The other half is what happened across the market while it stayed open.

Month 1
You reopen the search and rewrite the job description.
They onboard
Month 2
You interview. Panels are booked three weeks out.
They build
Month 3
Your finalist declines. The loop restarts from screening.
They ship V1
Month 4
You restart, with a wider brief and a tired panel.
Customer feedback
Month 5
You make another offer and wait on a counteroffer.
They ship V2
Month 6
Your new hire starts, on day one, learning the codebase.
6 months in

You didn't lose six months of recruiting.
You gave your competitor a six-month head start.

Execution compounds. So does delay. Your competitor doesn't just ship first, they learn first, and the next decision they make is better informed than yours. The market doesn't care why you couldn't hire them. It only sees what your competitor shipped while you were looking.

Work redistributes to the strongest people. Their load rises. Focus fragments. Burnout risk grows. Eventually the hiring problem becomes a retention problem.

One vacancy can create the next.

The second tax

The most expensive candidate isn't unqualified. It's the one who looks perfect.

Week 1
They pass the screen. The pipeline finally looks healthy.
Promising
Week 2
They impress the panel. Your staff engineer says hire.
Strong
Week 3
Final stage. You get compensation approved above band.
Committed
Week 4
"I've decided to go in another direction."
Gone

Your team didn't lose an offer. They lost the last four weeks.

Qualified is not the same as likely.

Answered by your process

Can they do the job?

  • Technical depth and system design
  • Judgement under real constraints
  • How they work with your team
Answered by nobody

Will they choose your job?

  • Why they would leave at all
  • What they want next, and the company stage that fits
  • Compensation expectations, and what it would take
  • Location and work model
  • Competing opportunities, and how far along
  • Counteroffer risk, and what would make them stay
  • The criteria that will actually decide it

We don't start closing when the offer arrives.

By offer stage there should be very few surprises. Not none; anyone promising certainty is overselling. Very few.

If nothing changes

What happens next year?

More filters. More recruiter headcount. More engineer interview time. Another ninety-day vacancy. Another late decline. Another roadmap push.

How long are you willing to keep doing that?

A different starting point

Stop filtering whoever applied. Recruit who should be there.

We don't sell a better filter. We change what enters the funnel. We create an additional precision pipeline alongside the team already doing the work.

Reactive

Who applied?

  • Advertise
  • Receive
  • Filter
  • Screen
  • Interview
  • Convince
  • Hope
Intentional

Who should we recruit?

  • Define the bar
  • Map the market
  • Approach directly
  • Qualify skill
  • Qualify intent
  • Curate
  • Close
The access advantage

How do you recruit the people who never applied?

You stop treating your ATS and your LinkedIn results as the whole market. LinkedIn is useful. It just isn't the market. Most searches lean on applicants, résumé databases and the obvious profiles. We don't just search harder. We search a larger talent surface.

What they invented

Patents and filings.

Filings and named inventors expose the engineers actually creating technology in a field, not the people whose résumés contain the right words.

What they shipped

Code and contributions.

Maintainers and contributors in the ecosystems that matter to your stack. One source among several; plenty of excellent engineers work behind closed source.

What they know

Research and communities.

Published work, conference programs, standards bodies and technical communities reveal who is working at the frontier of a discipline.

Where they fit

Career patterns.

Experience structured across skills, companies, projects and adjacent work, so the right person surfaces even when a keyword search never would.

We identify people by evidence of what they've built, not simply what a résumé claims. That matters more now because AI can rewrite a résumé in seconds. It cannot retroactively create the product someone shipped, the patent they filed or the research they published.

The obvious search

Who matches these keywords?

  • LinkedIn search
  • ATS database
  • Inbound applicants
  • Résumé keywords
  • Familiar names
The market map

Who has actually done this work?

  • Relevant companies and adjacent teams
  • Patents, shipped code, research
  • Technical communities
  • Career patterns
  • Direct outreach

The objective isn't a bigger candidate pool. It's a better starting point.

The data expands the market we can see. The recruiter still has to earn the conversation. Discovery is technology-assisted. Headhunting is still human: someone experienced has to get the person to engage, qualify them, and close them.

This is why the model produces fewer submissions. We spend more effort deciding who deserves to enter your process, so your team spends less deciding who should leave it. Your hiring team should be making fewer decisions, not more.

How our intelligence layer works

Traditional search depends heavily on exact titles and keywords. We structure professional experience across roles, skills, technologies, companies, projects and career patterns, then use semantic retrieval and model-based ranking to identify adjacent and non-obvious matches: matching by meaning rather than by literal string. The public evidence sources above (filings, code, research, communities) are joined to that structure so a search can start from what someone has demonstrably done.

None of it replaces the conversation. It decides who is worth having one with.

Evidence

What this model has produced.

Trailing figures across senior technical search. Historical results, not a forecast for your role.

Proof 01 · Precision

8:1

Submittal-to-offer ratio. The evaluation is done before you meet them.

Proof 02 · Conversion

85%+

Offer acceptance, frequently above 90%. Prepared candidates convert.

Proof 03 · Outcome

35 hires · 5.5 months

Built an engineering organization from zero, including six executive searches.

Before you hand us the role

The reasons you might still say no.

Fair questions. Here are our answers.

We already have a strong internal talent team. Why would we need you?

Good. We work best alongside strong internal teams.

This isn't us versus them. On difficult searches, we create a parallel pipeline that gives the hiring team another view of the market.

One of two things happens.

Your internal team is already surfacing the strongest available talent, and our work confirms it.

Or we introduce people your existing process did not reach.

Both outcomes are useful.

Your internal team stays at the center of the hiring operation. We become additional search capacity when specialization, bandwidth, urgency or access to a difficult talent market becomes the constraint.

Think of it as compare and contrast, not replacement.

We already have other recruiting firms working on this role.

Then you probably don't need another firm doing the same thing.

The question isn't how many recruiters are working on the search. The question is whether each one is expanding the talent market you can see—or simply producing more names from the same visible pool.

We would rather take responsibility for one difficult search than become the fourth firm generating activity across five.

Give us the role that is still open despite the existing coverage.

If our shortlist looks exactly like everyone else's, you have your answer.

If it doesn't, you just expanded the market.

We don't use outside recruiters.

There is usually a reason for that policy.

Maybe a previous firm charged a significant fee and produced mediocre results. Maybe hiring managers were overwhelmed with resumes. Maybe outside recruiters created more work than they removed.

Those are reasonable reasons to be skeptical.

We're not asking you to change your entire recruiting model.

We're asking you to test the assumption on one difficult role.

On a contingency search, there is no search fee unless you hire someone we introduce. Compare our pipeline against what you already have.

If we don't improve the market you can see, don't continue.

If we introduce the person you ultimately hire, the exception answered the question.

Why pay a recruiting fee when we already employ recruiters?

Because the fee isn't the real comparison.

The comparison is the fee versus another 30, 60 or 90 days of an empty seat.

Another month of engineering interviews.

Another finalist who declines.

Another roadmap item that moves right.

More work absorbed by the people already carrying it.

And another month in which your competitors continue hiring, shipping, learning and iterating.

If your internal team can produce the right hire faster, use them.

But if a specialized search partner changes the outcome, measure the fee against the cost of leaving the problem unresolved—not against zero.

The cheapest recruiting process is not always the one with the lowest fee.

It's the one that gets the right person productive fastest with the least organizational waste.

Can't our recruiters find the same people on LinkedIn?

Sometimes.

LinkedIn is useful. It just isn't the entire talent market.

We start with the companies, teams and environments where the capability actually lives. Then, where relevant, we expand the evidence surface through patents, open-source contribution, published research, conference participation, technical communities, professional history and adjacent career patterns.

The objective isn't to find obscure candidates simply so our list looks different.

It is to answer a better question:

Who has actually done this work—whether or not they used the keywords we searched for, appeared in the obvious search results, or ever applied to the job?

Then we recruit those people directly.

That broader search philosophy is already supported by Verticalmove's sourcing model, which uses market mapping and multiple evidence sources rather than relying on a single database or search interface.

How do you know the candidate will actually accept our offer?

We don't know with certainty.

Anyone who tells you they do is overselling.

But an offer decline should rarely reveal something nobody knew.

Before we introduce someone, we are already trying to understand why they would move, what they want next, compensation expectations, company stage, location, competing opportunities and the criteria that will ultimately determine their decision.

And that conversation continues throughout the interview process.

We don't start closing the candidate when the offer arrives.

By the time you are ready to make one, we should understand what they are likely to do, why they are likely to do it, and what could still change the outcome.

The goal isn't certainty.

It's very few surprises.

What if the person we're looking for genuinely doesn't exist?

Then finding that out quickly is valuable.

A serious search should occasionally tell you that the specification is wrong.

Maybe the compensation doesn't support the level.

Maybe the geography removes too much of the market.

Maybe the role combines two jobs that rarely exist in one person.

Maybe the experience requirements eliminate people who could actually deliver the outcome.

Or maybe the talent pool is simply much smaller than expected.

If that is what the market tells us, we'll show you the evidence and help you decide which variable to change.

Discovering that the role is unfillable as written in week two is useful. Discovering it in month nine is expensive.

A successful search does not always begin by proving the job description right.

Sometimes it begins by proving what the market will actually support.

Don't give us your easiest role.
Give us the one your team is tired of talking about.

One difficult role.
Compare what we produce against what you already have.
Then decide whether the difference is worth continuing.

Put One Role to the Test
hello@verticalmove.com