Agency Elimination Economics: The Math Nobody Does
The conversation that kills agency budgets usually starts the same way. A VP of Engineering mentions offhand that the last six hires came in over headcount because the agency fees pushed total cost above the approved band. Finance starts asking questions. Someone pulls up the invoices.
What they find is predictable: 20 to 25 percent of first-year salary, per hire, per placement. On a $180,000 software engineer, that is $36,000 to $45,000 per head. If the company made ten hires through agencies in a quarter, the bill is between $360,000 and $450,000. For a single quarter.
Nobody ran that math when they signed the contingency agreement. The agency framed it as "no cost unless we find someone," and that felt low-risk. It is not. The cost is simply invisible until the invoices land.
I have spent my career running the alternative model. Here is what the math actually looks like, and when the agency model makes sense despite the cost.
What Agencies Actually Cost
The standard contingency fee range is 18 to 25 percent of first-year base salary. Retained search for senior roles runs 30 to 33 percent. Most companies blend both depending on the role and urgency, which makes the blended cost hard to see until you aggregate it.
The calculation that should live in every talent budget is simple: multiply your planned annual hire count by the average salary for those roles, then apply the agency fee percentage. For a company hiring 40 people at an average base of $130,000 through agencies, that is $1,040,000 to $1,300,000 per year in placement fees alone, on top of any internal recruiting costs.
Most companies do not run this number explicitly. It gets absorbed across hiring manager budgets, buried in "personnel costs," or treated as a cost of doing business. That framing is the problem.
What a Sole Recruiter Actually Costs
A senior in-house TA leader in the $130,000 to $165,000 base range, with benefits and tools, runs approximately $180,000 to $210,000 in total annual cost. Call it $200,000 all-in as a working number.
At that cost, you need to save two agency fees on average-salary roles to break even in year one. Two hires. Everything after that is margin.
At Fisker, I scaled the team from 300 to over 1,700 employees, averaging 17 hires per month with a peak of 51 in a single month. At Sony Honda Mobility's AFEELA U.S. launch, I was the sole recruiter: 48 hires across 9 functions in 12 months, 98% offer acceptance (48 accepted, 1 declined), 42-day average time to fill against an industry average of 60 to 75 days.
The agency savings generated across those engagements exceeded $1.5 million annually. That is not a hypothetical. It is what direct sourcing at volume produces when you have someone who knows the market and owns the full pipeline.
The break-even math: If a sole recruiter costs $200,000 all-in and saves two agency fees per month at a 20% rate on $150,000 average salaries, the annual savings is $720,000 net of recruiter cost. Scale the hire volume or salary level and the savings compound quickly.
Where the Hidden Costs Are in the Agency Model
The fee percentage is the obvious cost. The less visible costs are what make the agency model genuinely expensive for companies in a growth phase.
Candidate quality variance. Agencies are optimized to close placements, not to find the right person. The incentive structure rewards submitting candidates fast and getting an offer accepted, not matching candidates to culture or predicting retention. When a $40,000 placement leaves in eight months, you pay the fee again.
No institutional knowledge.** When a search closes, the agency's knowledge of your company walks out with them. The next search starts cold. An internal recruiter builds context about your interview process, your offer dynamics, your EVP, your hiring managers' real preferences. That context compounds into efficiency over time.
Offer acceptance rates. The industry average offer acceptance rate is 85 to 90 percent. Declining offers is most common in contingency searches because candidates are often in multiple agency pipelines simultaneously and have other options in play by the time an offer arrives. My 98% acceptance rate on 49 offers comes from owning the candidate relationship directly, aligning on compensation early, and managing the close the way a recruiter who is not incentivized to rush should manage it.
Time to fill illusion. Agencies often promise faster time to fill. In practice, the best candidates in any market are not sitting in agency databases. They are sourced directly, through referrals, LinkedIn outreach, and networks built over years. An internal recruiter with deep vertical knowledge often fills roles faster because they are sourcing the right people, not recycling the same candidate pool every agency has already called.
When Agencies Are Still Worth It
I am not arguing that agencies are never justified. There are three situations where the math works in their favor.
First, pure volume spikes. If you need 30 hires in 60 days and your internal team cannot scale that fast, agencies can add capacity quickly. The fee is a capacity cost, not a sourcing failure.
Second, specialized niche searches. For a board director search, a C-suite role requiring a narrow candidate universe, or a highly technical specialty with a small global pool, a retained search firm with deep relationships in that community may legitimately find candidates you cannot. The fee is for the network access, not just the recruiting function.
Third, early-stage companies with no HR function. Before you have an internal recruiter, agencies get you moving. The cost is acceptable when the alternative is zero hiring capacity.
Outside those situations, the math tilts toward direct sourcing, particularly for companies in the 50 to 500 employee range hiring across multiple functions simultaneously.
What "Sole Recruiter, Full Function" Actually Requires
The model only works if the recruiter can cover all the functions the company needs. At AFEELA, that was engineering, product, operations, legal, finance, communications, sales, business development, and HR. Nine functions, one recruiter, 48 hires, zero agencies.
That breadth requires someone who can source both a director of regulatory affairs and a staff software engineer in the same week, adjust their sourcing strategy for each, and hold the candidate relationship from first outreach to day one. It is not a common profile.
It also requires the tools to support direct sourcing at volume: LinkedIn Recruiter or a comparable sourcing platform, an ATS that does not create friction, and a hiring manager culture that trusts the recruiter to filter before presenting. When those conditions exist, a single strong recruiter can handle what most companies assume requires a team and an agency roster.
The Board as a Sourcing Layer
For mobility companies specifically, the talent pool for direct sourcing is larger and more concentrated than it appears. Engineers, manufacturing leaders, supply chain specialists, and operations talent from EV, AV, eVTOL, electric marine, and autonomous delivery companies are an identifiable, reachable population. I track over 5,400 open roles across those verticals on my free Mobility Jobs Board, which surfaces which companies are actively scaling and in which functions. The companies hiring at volume are the ones producing candidates when those hires move on. That market intelligence is what direct sourcing runs on.
The point is not the board itself. The point is that a recruiter who understands the mobility talent landscape and tracks it systematically is doing something agencies cannot: building durable sourcing infrastructure that gets cheaper per hire over time, not more expensive.
The Metric Your Board Should Be Asking For
Cost per hire is the number that makes the agency model visible. It is not commonly tracked at the granularity that would trigger action, but it should be.
At Gurley Leep, I managed a high-volume recruiting function that averaged $243 in cost per hire across 782 hires over 28 months. That number includes all recruiting infrastructure costs divided by total hires. It is a different category from what most companies pay per agency placement on a single senior role.
If your company does not know its cost per hire by channel, by function, and by source, the agency spend is invisible by design. Making it visible is the first move toward eliminating it.
Start here: Pull your last 12 months of agency invoices. Divide total fees by total agency placements. Compare that number to your average total compensation package. That ratio tells you what your current model costs per head, and what the break-even point is for an internal hire.
What to Do With This
If you are building a team in EV, AV, eVTOL, or an adjacent vertical right now, the talent market is deep enough and specialized enough that direct sourcing compounds. Every hire you make through an internal recruiter builds pipeline, process knowledge, and candidate relationships that reduce the cost of the next hire. Every agency placement resets to zero.
The question is not whether to eliminate agencies. The question is at what hire volume and company stage the math tips, and whether you have the right internal recruiter to execute the model. Most mobility companies I have seen would clear that threshold faster than they expect.
See who is hiring right now. My free Mobility Jobs Board tracks open roles at EV, AV, eVTOL, electric marine, and autonomous delivery companies nightly, pulled from their career systems. Browse it at larrysherwoodjr.com/jobs/. No signup required.
Building something ambitious?
I build recruiting functions from scratch as a sole recruiter. 48 hires for the AFEELA U.S. launch, 98% offer acceptance, $1.5M+ in annual agency savings. Currently open to senior TA leadership roles, remote.
Prefer email? Send a message instead.