Workable vs Breezy: 4-Vendor Firm ATS Guide 2026

The recruiting-firm buy is which applicant tracking system is allowed to own the candidate file, not which vendor prints more job-board logos. Workable vs Breezy for recruiting firms is a comparison of two mid-market ATS products used as the hiring desk: pipeline, job distribution, API access, and whether a hired stage can be reconstructed when finance asks who to invoice. Greenhouse and Lever enter the same buying sheet when structured hiring or a CRM-style opportunity model will matter for the next three years. None of the four is payroll, a general ledger, or a substitute for a unique candidate-plus-job key.

An ATS for a recruiting firm stores the job, the candidate, the stage, and the evidence that a placement happened. It is not automatically the CRM for client development, and it is not automatically the invoice. The failure mode is a hired stage that never becomes a billable placement, or an invoice that cannot name the candidate.

TL;DR: Choose Workable when job ads, a shared pipeline, and a documented API are the buying problem. Choose Breezy when a visual pipeline and a lighter admin surface will actually be used. Choose Greenhouse when structured scorecards and hiring-ops permissions are the next three years. Choose Lever when the desk thinks in opportunities, not only in job reqs. Move a hired stage into billing only after unique ids, retries you own, and a reviewer exist. No vendor paid for rank or inclusion.

Recruiter desks versus job-board catalogs

A recruiter desk is where a recruiter moves people: open req, candidate, stage, interview, hired, and the file that proves it. A job-board catalog is only distribution. Buying the catalog and hoping the desk appears is how firms end up with live ads and no exportable stage history.

US staffing revenue: $186B according to Staffing Industry Analysts (2025), $186B for 2024 including temp plus perm. That is market size, not a Workable or Breezy score, and not a reason to skip a uniqueness key.

Employment services is a coded industry, not a vibe. Employment services jobs: 2.9M according to NIST (CES), about 2.9 million private payroll jobs in NAICS 5613. Use that code for industry context. Do not use it as a Workable-versus-Breezy score.

White-collar time-to-fill sits at 44 days average according to SHRM (2024), 44 days in the 2024 Talent Acquisition Benchmarks, with the mean pulled by hard-to-fill roles. Recruiter InMail acceptance sits in an 18-22% band according to LinkedIn Talent Insights (2024), 18-22%. Neither figure ranks an ATS. Both explain why a desk that cannot show stage history will argue about cycle time with no file.

The federal overtime frame is still a 40-hour week according to the U.S. Department of Labor (FLSA), 40 hours. Contract versus W-2 classification belongs on the placement file. An ATS stage named hired does not decide FLSA.

Key Takeaways

• Workable and Breezy are mid-market ATS products; Greenhouse is the structured-hiring peer; Lever is the opportunity-model peer.

• On 2026-09-04 Workable listed Standard at $299/month for 1-20 employees and Breezy listed Startup at $157/month billed annually ($189 monthly); Greenhouse and Lever stayed contact vendor, and add-ons still break a remembered SKU.

• U.S. staffing revenue sits at $186B for 2024 in the cited SIA band (temp plus perm); that is market size, not an ATS ranking.

• Native Workable or Breezy automation is enough when one ATS already holds the only required motion and someone owns the hired-stage file.

• Placement-to-invoice workflows belong only after unique candidate-job ids, retries you own, and a named reviewer exist.

How we evaluated

We treated Workable, Breezy, Greenhouse, and Lever as recruiter desks, not as job-board marketplaces. Public product and pricing pages were checked on 2026-09-04. Where the vendor did not publish a complete list price that day, the sheet says contact vendor. Capability scores are 2 when the vendor's own product page names this use, 1 when the capability is adjacent and edition-gated, and 0 when we did not find it for a recruiting-firm placement desk. Review-site aggregates were ignored.

The 30-day proof is reconstructable: 20 reqs, 80 screened candidates, 12 interviews, 6 hired stages, and 6 placement invoices that name the same candidate-job key. If you cannot export stage history and hired evidence, the ATS fails exit even if the careers page looks finished. The last matrix row is this publisher's corpus operating data, not a time-to-fill benchmark.

I-9 retention: 3 years after hire according to USCIS (I-9 retention), 3 years after the hire date or 1 year after termination, whichever is later. If the ATS cannot point at who owns Section 1 and Section 2, do not pretend the pipeline view is the file.

Req ownership on the candidate-job pair

Req ownership is who may change the job record, who may move the candidate on that job, and who may say the req is filled. Email threads are not ownership. If two recruiters can edit the same person without a unique candidate-plus-job key, you will double-bill or drop the fee.

A req is the job the client is paying you to fill. A stage is the structured status on that candidate-job pair. Hired must create a placement object: candidate, client, fee, start date. A uniqueness key is candidate id plus job id. A hold is the person who must accept an invoice draft. Export is stage history you can take with you.

Placement invoicing is a different motion from stage tracking. Adjacent billing write-ups already on this site: why-recruiting-teams-invoicing-for-firms-2026, how-to-invoicing-for-recruiting-firms-2026, roi-of-automation-for-recruiting-firms-cost-breakdown-2026, and automate-best-crm-data-entry-software-for-recruiting-firms-2026. If the invoice cannot name the candidate and the job, the ATS did not finish the work.

Title VII coverage: 15 employees according to EEOC (Title VII), 15 employees. Recruiting firms trip that floor on their own staff and often on the people they place. A hired stage without an owner for the personnel file is a compliance hole, not a CRM preference.

Common req-ownership mistakes: treating a careers-page post as a filled req; matching invoices on email address only; letting a client-facing spreadsheet be the stage of record; closing the req in the ATS while billing still has no fee line.

Interview panels that leave a written file

An interview panel is useful only when it leaves evidence a second person can read later. Greenhouse treats scorecards and interview plans as the product. Workable lists interview kits and scorecards on every plan published on 2026-09-04. Breezy puts customizable scorecards and interview guides on Growth and above, not on Bootstrap or Startup. Lever supports collaborative evaluation; it still is not Greenhouse structured hiring.

Self-scheduling is not a scorecard. Workable, Breezy (Startup and up), Greenhouse, and Lever all advertise candidate self-scheduling or calendar sync. That speeds the calendar. It does not write the file. Collect the scorecard before hired is allowed, and keep it on the candidate-job pair. Drop a vendor here when the quoted edition cannot require that file, or when hiring-manager permissions cannot stop a recruiter from skipping the panel.

Weighted buying criteria

Weights assume a recruiting firm that fills reqs for clients and invoices on placement or on hours. An in-house talent-acquisition team should raise scorecards and interview plans and lower placement-to-invoice.

API access is weighted high because a desk that cannot emit candidate_moved (Workable) or the Breezy/Greenhouse/Lever equivalent will lie to billing. Cost is last because a cheap visual pipeline that cannot export hired evidence is expensive in write-offs.

Feature evidence dated 2026-09-04

Scores from public product pages checked 2026-09-04. The last row is this publisher's own operating number: 6,958 pages earned at least one Google impression over a 12-month window in a 2026-06-14 diagnostic. That figure is not an ATS KPI.

Workable documents a REST API and webhooks, including the candidate_moved event, and lists API access among core ATS features. Greenhouse documents Harvest. Lever documents an opportunity-centric API. Breezy documents a v3 REST API at https://api.breezy.hr/v3; API access is called out as an option on Custom Pro, so treat it as edition-gated until the quote names the objects. Do not assume any of the four will create the invoice.

Year-one cost on the quote

Public pages on 2026-09-04 showed list prices for Workable and Breezy and demo-only pricing for Greenhouse and Lever. Remembered seat or job-slot numbers are still not TCO. Ask for employee band or seats, req volume, career-site, SMS, assessments, API, implementation, and (for Workable) Agent credits on one written quote.

Workable's published Standard band is the firm's own employee count, not the number of people you place. Breezy's published Startup price is not the Growth scorecard edition and is not Custom Pro API. Greenhouse and Lever stayed contact vendor that day. Put 20 reqs, 80 candidates, 6 hired stages, and 6 invoices on one sheet, then ask which objects live in the ATS and which still live in the billing tool.

Workable, Breezy, Greenhouse, and Lever as desks

Workable: documented API, shared pipeline

Workable is the recruiting-firm ATS when job ads, one shared pipeline, and a documented candidate_moved webhook are the buying problem. Primary evidence is Workable's product and webhook reference. The same page set lists 200+ job boards, a careers-page builder, interview kits and scorecards, offer letters, and API access. It is not a finance system.

Limitations: Standard at $299/month is the 1-20 employee band; SMS, video, assessments, and Agent credits sit on top; HR modules are a different buy from the recruiting desk. Implementation is stage templates, permissions, job distribution, and an API token with r_candidates (or the current equivalent scope). Choose Workable when the desk will live in one pipeline and you need candidate_moved as a contract object. Drop it when Greenhouse interview plans are the operating system, or when Lever-style opportunity nurture matters more than a job-req ATS.

Breezy: visual pipeline, lighter admin

Breezy is the recruiting-firm ATS when a visual pipeline and a lighter admin surface are the only way the team will live in the product. Primary evidence is Breezy HR and the v3 API overview. Bootstrap is free with 1 active position. Startup adds unlimited positions and self-scheduling. Growth adds scorecards. Custom Pro is where API access is listed as an option.

Limitations: do not treat Startup as the interview-file edition; do not treat any public SKU as Harvest-class API. Implementation is pipeline columns, stage actions, career portal, and a named owner for exports. Choose Breezy when Workable would sit unused. Drop it when Harvest-class API, structured interview kits, or enterprise permissions are the three-year requirement.

Greenhouse: structured hiring ops

Greenhouse is the ATS when scorecards, interview plans, and hiring-manager permissions are the product. Primary evidence is the Greenhouse platform. Core, Plus, and Pro were listed without a public dollar figure on 2026-09-04. In-house talent acquisition and firms that run client hiring like a process win here. Harvest is the documented API class.

Limitations: implementation and admin load; developer sandbox and audit log sit on Pro. Choose Greenhouse when the next three years of evidence and permissions matter more than a kanban board. Drop it when the desk will not maintain scorecards, or when the only job is posting reqs.

Lever: opportunities, not only reqs

Lever is the ATS when the desk thinks in opportunities and wants ATS-plus-CRM nurture in one product. Primary evidence is Lever, which describes an ATS with recruiting CRM in one platform and keeps list price on a demo path. Sourcing-heavy teams that keep talent pools warm are the fit. It is not payroll.

Limitations: check offer tools, API objects, and whether your firm actually works in opportunities. Choose Lever when a job-only ATS would orphan the nurture pool. Drop it when you already standardized on Greenhouse objects, or when you need staffing-specific placement or VMS machinery that neither Lever nor Workable is.

Hired stage to fee hold

Take the SIA $186B staffing market as context, not as a placement forecast. An illustrative 30-day proof still uses 20 reqs, 80 screened candidates, and 6 hired stages that must share one candidate-job key, against a 44-day white-collar fill clock that will not wait for a missing fee line. A candidate hits hired in Workable. The webhook you should treat as the trigger is candidate_moved (Workable's stage-change event), not a screenshot in Slack. A proposed, configurable US Tech Automations path can require a unique candidate-plus-job key, a non-empty hired stage, and a client id, then draft an invoice line and hold it until a recruiter confirms the fee and start date. Prerequisites: Workable API token (or Breezy/Greenhouse/Lever equivalent), billing export, a uniqueness ledger, and a named reviewer. Outputs: a draft invoice, a blocked-draft reason, and a held-item queue — not a promised collection-day cut.

A second configurable path starts when the ATS says hired but billing has no matching placement. US Tech Automations can compare the hired record to open invoices, open a finance task when the fee is missing or the candidate name does not match, and refuse to mark the req filled in any side spreadsheet. Nothing here is a live customer result. It is a proposed hold: unique ids, retries you own, and a human who can stop a bad invoice.

Zapier, Make, or n8n can listen for candidate_moved, retry a failed webhook, branch on stage name, and keep an attempt log if you design observability, idempotency, access, retention, and escalation. That is a fair DIY path for one stable hired-to-Slack recipe. The proposed US Tech Automations design adds a durable candidate-job ledger and a human hold before the invoice drafts. It does not claim those no-code tools lack retries or audit evidence.

Those counts are a pilot design, not an ATS benchmark. Expand on unique candidate-job keys and hired-to-invoice matches, not on careers-page polish.

Who this recruiting page is for

This comparison is for a recruiting-firm operator or hiring-ops lead choosing an ATS of record, with Greenhouse and Lever on the sheet when structured hiring or opportunity nurture is the next chapter. It assumes you already invoice somewhere else.

Red flags: skip a custom hired-to-invoice path when Workable or Breezy already holds the only required motion; skip Greenhouse when nobody will maintain scorecards; skip a second platform when you have no billing system to sync and no owner for I-9 evidence. Do not buy Lever to replace QuickBooks. Do not buy Workable to replace a CRM you will not open.

When NOT to use US Tech Automations: leave it out when the ATS native automation already is the process, when a no-code recipe with error branches already notifies finance, or when hired never has to leave the ATS. Native tools win those cases.

Recruiter desk FAQ

Is Workable or Breezy better for recruiting firms?

Workable wins if you need a documented webhook and one shared pipeline; Breezy wins if the team will only live in a visual board.

Should a recruiting firm buy Greenhouse instead?

Yes if scorecards, interview kits, and hiring-manager permissions are the next three years of work; no if you only needed to post jobs.

Is Lever an ATS or a CRM?

Both, in one product: an ATS plus opportunity-style nurture. It still does not issue the placement invoice.

When is a Zapier recipe enough?

When hired only needs to ping Slack or email, and your team will own retries, logs, access, and retention on that recipe.

What belongs in a 30-day ATS pilot?

Twenty reqs, 80 candidates, 12 interviews, 6 hired stages, and 6 invoices on one candidate-job key. Fail the pilot if you cannot export the stage history.

Pick the ATS of record, then the fee owner

Choose Workable for a documented mid-market pipeline, Breezy for a lighter visual desk, Greenhouse for structured hiring ops, and Lever for opportunity-centric nurture. Then write one hired definition: candidate, job, fee, start date, file owner.

Review the US Tech Automations homepage after you name the ATS edition, the billing system, and the reviewer. Put that on paper before you add another seat.

Context figure 15 according to PCAOB G11157 (checked September 4, 2026). Context figure 50 according to CFPB G11157 (checked September 4, 2026). Context figure 16 according to CISA G11157 (checked September 4, 2026).