How to start a recruiting agency in the US: a 2026 playbook for solo recruiters

You have placed other people for years. Now you want to keep the commission. Good. The math is obvious: a single senior engineer placement at a 20% success fee on a $180K base is $36K in one invoice. Four of those a year and you are past what most in-house recruiters earn.

The part nobody tells you is that the work of building the agency has almost nothing to do with the work of recruiting. In the first ninety days you will spend more time on entity paperwork, licensing, and invoicing than on sourcing. This guide walks you through the exact order: what to decide, what to buy, what to skip, and where the money actually comes in.

Everything below assumes you are going solo, starting from zero clients, and want to be billing by the end of the first quarter.

Five decisions that matter more than the rest

Before you file a single form, lock these five answers. Every later step depends on them.

Legal home. Register in the state where you live and work. Delaware and Wyoming sound clever in LinkedIn posts and will cost you extra without giving you anything. You will still have to register as a "foreign LLC" in your home state because that is where you actually do business.

Niche. One function, one seniority band, one or two industries. "Tech recruiter" is not a niche. "Senior backend engineers for Series A-C fintech in New York" is a niche. Clients pay premium fees for depth, not breadth.

Fee model. Contingency (15-25% of base salary, paid on hire), retained (25-33%, paid in thirds), or flat fee ($3K-10K per hire). Pick one for the first year. Mixing models early creates pricing chaos and slows closes.

Starting budget. The realistic MVP is under $500 out of pocket: state filing fee, a domain, email, and a phone line. Anything you add in month one is a guess. Buy it in month three when you know what you actually need.

Your first five clients. Not your first fifty. Name the five companies you already have warm lines into. If you cannot list five by name today, the problem is not tools or pricing. It is demand.

Lock these five and the rest of this guide becomes an execution checklist.

Step 1: pick your state and register the entity

Most solo recruiters should form an LLC in their home state. Not Delaware. Not Wyoming. The savings from a "tax-friendly" state get eaten by the foreign-LLC registration fee in the state where you actually work, plus a registered agent you now need to pay forever.

LLC filing fees by state

The numbers vary widely. A few examples from 2026 Secretary of State schedules:

Six states require no annual report at all: Arizona, Mississippi, Missouri, New Mexico, Ohio, and South Carolina. If you live in one, your ongoing compliance cost is effectively zero.

The steps

After you have decided where to file, the sequence is tight.

First, run a business-name search on your Secretary of State's website. The name must include "LLC" or "Limited Liability Company." Check that the matching .com domain is available before you commit.

Second, file your Articles of Organization online. In most states this takes twenty minutes and confirms within 1-3 business days.

Third, get an EIN from the IRS. It is free, takes ten minutes on IRS.gov, and you need it to open a business bank account. Do not pay any third-party service for this.

Fourth, write a simple operating agreement. Even as a single-member LLC you want one in writing. A one-page template works for year one.

Fifth, open a business checking account. Chase, Mercury, and Relay all offer free business accounts with quick online opening. Keep personal and business money separate from day one.

S-Corp election: wait

You will see advice to elect S-Corp status to cut self-employment tax. It is real, but the math only starts to work once net profit passes roughly $60K-80K. Below that, the added payroll, bookkeeping, and filing costs exceed the tax savings. Stay as a default LLC (taxed as a sole proprietor) in year one and revisit at tax time.

Step 2: check the employment agency license

This is the step most new agencies miss and the one that creates the worst back-office surprises. Twenty-two US states plus DC require a dedicated license to operate as an employment or staffing agency, separate from your LLC. California does not require a license but requires a $25K-50K surety bond that you still have to post before you bill a client.

Check this before you sign your first contract. Operating without a required license can void placements, invite state labor department penalties, and force you to refund fees already collected.

States that require an employment agency license in 2026

Source: Harbor Compliance state-by-state employment agency table, revised March 2024 and cross-checked against each state labor department in April 2026.

States without a state-level license requirement

Alabama, Arizona, Colorado, Delaware, Florida, Georgia, Idaho, Kentucky, Maine, Maryland, Minnesota, Mississippi, Missouri, Montana, Nebraska, New Hampshire, New Mexico, Ohio, Oklahoma, Oregon, Pennsylvania, Rhode Island, South Dakota, Tennessee, Texas, Utah, Vermont, and Virginia do not require a state-level employment agency license in 2026.

California, while not requiring a license, requires a surety bond of $25,000-$50,000 posted with the Department of Industrial Relations before you can legally operate a direct-hire placement business.

What a surety bond actually is

A surety bond is not insurance. It is a guarantee to the state that if you misuse client funds or violate your license, the state can collect damages. You pay an annual premium (typically 1-3% of the bond amount) to a surety company. On a $10,000 New York bond, that is $100-$300 per year. You get the bond, file it with the labor department, and renew annually.

If your state requires one, price it with two or three surety providers. Don't buy from the first one your lawyer suggests. The underwriting criteria are standard and rates vary.

Step 3: pick a niche that pays

This is where most new agencies fall over. They try to be "a tech recruiter" or "a healthcare recruiter" and end up with no pricing power. A real niche is narrow enough that you can list every hiring manager who matters in it.

A useful test: can you name ten companies, five hiring managers at each, and one active req for each manager? If not, you are not niched down enough to sell a premium fee.

Five niches that pay well in 2026 and are accessible to a solo recruiter:

Senior and staff engineering at Series A-C startups. Fees range $30K-60K per placement. Competition is fierce but so is demand. The Block and Oracle layoff waves in Q1 2026 pushed a lot of senior talent into the market, and startups are hiring the best of them fast. See our Block AI layoff coverage and Oracle layoffs analysis for supply-side context.

Healthcare (traveling nurses, specialized clinicians). Fees are lower per head but volume is high and retention contracts are standard. Aya, Medical Solutions, and AMN dominate, but regional specialty firms still win by working narrow specialties.

AI and ML roles. Fee premium is real: 20-25% of base instead of 15-20%, and base salaries are $200K-400K. The catch is that many of the strongest candidates are already in play with big platforms. You have to be faster, not cheaper.

Sales and GTM for mid-market B2B SaaS. AEs, SDRs, CS leads, RevOps. Fees $20K-40K. Demand is consistent and the hiring bar is legible (quota attainment, ACV, deal size).

Executive search in a specific vertical. VP Engineering for e-commerce. CFO for DTC brands. Head of People for fintech. Retainers $60K-150K. The ramp is slow (it takes a year to build the network), but once you are in the network the margins are the best in the business.

Pick one. Add a second only after the first is producing consistent revenue.

Step 4: choose your pricing model

Pricing is strategy, not math. The number you charge tells clients whether you are a commodity or an expert.

Contingency. 15-25% of first-year base salary. Paid only when the candidate starts. Low friction to sign, but your effective hourly rate is brutal until you close. Most solo agencies start here.

Retained. 25-33% of base, paid in thirds (signing, shortlist, hire). Client is committed, you are committed. Works for senior and executive roles or when you have an existing relationship.

Flat fee. $3,000-$10,000 per hire, often with a tiered guarantee. Easy to sell to first-time hiring managers and small companies. Margins are thinner per role but pipeline velocity is higher.

Hourly. $75-$150/hour for sourcing or on-demand recruiting support. Good as a bridge revenue stream in months one through three while placements ramp.

Recruiter-on-demand retainer. $4,000-$8,000 per month for a fixed number of sourcing hours or placements. Popular with Series A startups that do not need a full in-house recruiter.

For the full math on how these fee models compare to agency employee salaries and what solo recruiters actually take home, see the 2026 recruiter salary analysis. It breaks down the commission structures, geographic premiums, and remote-vs-office pay gap using a mix of public data and our Intelligence compensation dataset.

The short version: a solo recruiter closing four mid-market tech roles a year at 18% on a $150K average base grosses $108K before expenses. That is competitive with a senior in-house recruiter, without the manager overhead.

Step 5: the minimum viable stack

You do not need the full enterprise tool stack to bill your first invoice. Most of what you see on recruiting vendor landing pages is optimized for teams of twenty, not for you. Start minimal and add based on what your actual workflow reveals.

Here is what you actually need month one:

Candidate sourcing. A sourcing tool that can search beyond LinkedIn alone and give you compensation context. Glozo covers this: intent-based search, 30+ sources aggregated, and a compensation estimate on every profile so you can filter out candidates outside the budget before you spend time reaching out. Free to try, pay as you scale.

ATS. You do not need one in month one. A Notion database or Airtable with columns for candidate, stage, client, and next action is enough for your first dozen placements. Move to a real ATS (Manatal, Recruit CRM, Bullhorn) when you break ten active roles.

Outreach. A sequencer like Lemlist, Smartlead, or the one inside your ATS. Keep email warm-up turned on. Send no more than 20-30 outreach emails per day from a new domain.

Client CRM. HubSpot Free works. So does a Notion board. Separate from candidate tracking.

Job description writing. Bad JDs kill response rates. Use a tool like jd.glozo.com to tighten descriptions against market data before you post.

Contracts. A standard services agreement template from a lawyer in your state. Budget $500-$1,500 for a good one. Reuse it forever.

Accounting. QuickBooks Self-Employed or Wave. Do not reinvent this.

If you want a wider view of free tools that solo recruiters use, the 22 free tools guide covers the adjacent space.

Step 6: land your first five clients

If you cannot name five warm leads today, everything above is premature. The agencies that fail in year one fail here, not on tooling or legal structure.

Three tactics that work for solo recruiters in 2026.

Warm reactivation. Every hiring manager you have ever placed a candidate with is a lead. Message each one: "I started my own shop. Here is what I am focused on. Who at your company is hiring in [your niche] right now?" Sixty of these messages in week one will produce 3-5 conversations and 1-2 first contracts.

Niche content and presence. One LinkedIn post per week about market data in your niche. Specific numbers. No generic "the war for talent" content. The goal is to be the person hiring managers think of when a req opens. This compounds: week one nothing happens, week twelve you start getting inbound.

Founder-direct outreach. Pick 50 Series A-C companies in your niche. Email the founder or hiring manager directly. Not "are you hiring?" but "I noticed you are hiring a [role]. I have three candidates who have done this at [peer company]. Worth a 15-minute call?" The conversion rate on a tight, specific email with named candidates is 3-5x what a generic pitch gets.

One thing that does not work: waiting for clients to find you through SEO or paid ads. Agency SEO works for established shops with case studies. In year one, you have no case studies. Outreach and warm network are the entire game.

The real cost to start (under $500)

Forget the LinkedIn posts about $50K "investments." For a solo US recruiting agency in 2026, the actual MVP budget looks like this:

Total realistic month-one out-of-pocket: $300-$700 depending on your state. Everything else waits until you are billing.

The money people lose in year one is not on tools. It is on opportunity cost (working free for slow clients) and on bad contracts (missing replacement clauses, weak net-30 language, no retainer protection).

Step 7: first-year revenue reality check

Two honest numbers before you quit your job.

A solo recruiter in year one, working full time, placing one role per month at an average 18% success fee on a $120K base, grosses roughly $260K annually. That is the ceiling, not the median. The realistic year-one number for someone starting from a warm network is closer to $80K-$150K, with heavy variance by quarter.

A solo recruiter in year one without a warm network is often under $60K. Many never reach profitability and return to in-house work in month nine. This is not a judgment; it is the base rate. The agencies that survive are the ones where the founder had concrete demand before they filed the LLC.

If you want the full breakdown of freelance fee models vs. agency employee salaries, geographic premiums, and the remote pay inversion for senior roles, the 2026 recruiter salary analysis has the tables.

Expect 90 days before the first invoice clears. Plan cash for six months.

Three mistakes that kill first-year agencies

Three patterns show up over and over among agencies that fold in the first twelve months.

Signing every client who says yes. In the first six months you have no pipeline to fall back on, so the temptation is to say yes to any signed agreement. The damage is that bad clients eat your calendar. A hiring manager who sends three unclear reqs, ghosts for two weeks, then rejects three shortlists in a row consumes the capacity you needed for good clients. Qualify hard. Fire fast.

Contingency on everything. Contingency feels like safe risk-sharing. It is actually you financing the client's hiring process with unpaid labor. A 90-day contingent search with no hire is 80-120 hours of work at a $0 hourly rate. Move at least some of your volume to retained or flat fee by month four.

Building an ops stack before having clients. No ATS, no career site, no brand in month one. These are month-six problems. The failure mode is spending month-one energy on vendor calls and branding when you should be on client calls.

None of these are secret. They are just easy to make when you are excited about building the business instead of selling placements.

Ready to start billing?

Filing the LLC is the easy part. Closing the first five placements is the part that makes the agency real.

Glozo gives you intent-based sourcing, compensation context on every profile, and an "Open to Offers" signal that flags passive candidates most likely to respond, before you spend a single outreach credit. It is what we would use if we were starting a solo agency in 2026.

Start sourcing with Glozo → free to try, no credit card required for the first project.