Job search
LCA Filings vs. H-1B Petition Data: Build an Employer List Without Treating History as a Job Offer
A real Google LLC LCA and a current engineering posting show the distance between historical administrative records and role-specific sponsorship confirmation.
An LCA record and an H-1B decision describe different administrative events. Use them to discover employers with relevant history, then inspect current vacancies and ask about the specific role. Neither dataset is a job offer, a list of unfilled positions, or confirmation that an employer will sponsor you.
LandOffer publishes this guide and provides a place to browse recent roles. This review uses official sources and downloaded disclosure data checked on October 12, 2026. A real Google example connects historical records to a current posting while preserving the questions the public evidence cannot answer.

Start with the event each source records
The DOL LCA program covers H-1B, H-1B1 and E-3 applications. Employers make labor-condition attestations, including wage-related obligations. Certification is a step in the applicable process; it does not establish that USCIS subsequently approved an H-1B petition or that a worker started employment.
Filter VISA_CLASS before calling an LCA extract H-1B data. Otherwise, the employer list can include activity under different classifications. A third-party database labeled H-1B may use a file that also contains H-1B1 and E-3 records.
The USCIS Employer Data Hub explanation concerns H-1B petition decisions collected from Form I-129 and adjudication information. It reports first decisions, excludes petitions still pending when data are generated, and excludes subsequent decisions such as appeals or revocations.
Use the sources for different parts of employer research. LCA records can provide occupation, requested employment period and worksite details. USCIS data provide a view of recorded H-1B decisions. They do not form a complete, publicly linked timeline for each applicant.
A completed comparison for building your employer list
This matrix applies the official definitions to practical search decisions. The interpretation column is editorial guidance, not a government ranking of employers.
| Question | DOL LCA disclosure | USCIS H-1B Employer Data Hub | Useful interpretation |
|---|---|---|---|
| What event is represented? | Application information and an OFLC determination | USCIS first approval or denial | Keep certification and petition approval separate |
| Which classifications appear? | H-1B, H-1B1 and E-3 | H-1B | Filter visa class before comparing |
| Which date places activity in the release? | Most recent determination date within the reporting period | Fiscal year of first recorded adjudication | Matching fiscal-year labels do not link individual cases |
| What does a worker-related number mean? | Requested workers and employment categories | Published decision measures; current glossary describes approved or denied workers | Neither measures current vacancies or necessarily unique hires |
| Which geography appears? | Employer address and worksite fields | Petitioner mailing city, state and ZIP | Mailing location is not automatically the job location |
| How is an employer named? | Legal business name and separate DBA field | Petitioner name, with common spelling used per tax ID | Preserve original names before normalizing |
| What is missing? | Pending determinations and some excluded information | Pending petitions and later decisions | Absence is not proof that an employer never sponsors |
Save the source, release period, exact employer name, relevant occupation and a short interpretation in your shortlist. “Historical H-1B activity found” is a defensible label. “Will sponsor this opening” needs different evidence.
Read dates before sorting by recency
DOL's performance-data page describes cumulative fiscal-year releases. The FY2026 Q3 files cover determinations from October 1, 2025 through June 30, 2026. A received date can fall before that window because receipt and determination are different events.
The USCIS fiscal-year definition likewise uses adjudication rather than receipt. Federal fiscal years run from October 1 through September 30. A case received in September and decided in October can therefore appear in the next fiscal year's decision data.
This matters when a dashboard says an employer was active “this year.” Determine whether it means a calendar year, federal fiscal year, filing date, decision date or requested start date. Those are not interchangeable definitions of recent recruiting.
Do not add Q1, Q2 and Q3 cumulative totals to create an annual total. The same earlier activity can appear in later releases. Choose one relevant cumulative release, preserve its version, and use CASE_NUMBER rather than treating every quarterly appearance as a new filing.
DOL also warns that later appeals or redeterminations can change a small share of determinations. A saved row and a later release may differ for a legitimate reason. Record the release date before interpreting the difference as an employer error.
Keep cases, requested workers and decisions apart
The FY2026 Q3 LCA record layout defines CASE_NUMBER as an application identifier and TOTAL_WORKER_POSITIONS as requested foreign workers. One case is therefore not a universal unit of one worker or one available job.
The same layout separates new employment, continued employment, changes, concurrent employment and amended petitions. A record for a software occupation can concern someone continuing with an employer. Its technical title does not make it a new opening for external applicants.
Worksite detail introduces another counting trap. The main file describes a first worksite, while additional worksite files provide further locations. Joining those tables can repeat the same case across locations. Count cases or requested workers according to the question you are answering, rather than summing every joined row.
USCIS's current glossary distinguishes six employment categories, including continuation and change of employer. It describes the approval and denial measures as counts of workers. Its introductory organization paragraph still refers to initial and continuing aggregates. Preserve the actual column labels in the export you use and consult the matching definitions; do not silently rename every measure “new sponsored hires.”
An approval percentage also needs its denominator and category. Even a correctly calculated historical ratio would describe the included decisions, not your probability of receiving an offer, obtaining sponsorship or getting an individual petition approved.
Requested employment periods are another separate date pair. A future BEGIN_DATE in a historical LCA does not establish that recruiting remains open until that date. The record describes what the employer requested in that application. Likewise, an END_DATE is not an advertised application deadline. When comparing possible employers, use those fields to understand the historical record, then take the current vacancy's timing from its own source. If the posting gives no deadline, keep it unspecified rather than borrowing a date from an unrelated administrative case.
Check the actual schema instead of repeating old summaries
The DOL page's general exclusions mention FEIN among information not disclosed. However, the current FY2026 Q3 record layout includes EMPLOYER_FEIN, and the downloaded file inspected for this review contains that field with populated employer identifiers. The current layout's exclusion note refers to attorney identifiers instead.
This is a source conflict, not a reason to pretend one version does not exist. If you need an identifier for entity research, inspect your specific release and record its availability. Do not assume every older file has the same columns. An identifier can help distinguish employers without establishing a current role's sponsorship policy.
USCIS publishes the last four digits of a petitioner's tax ID. Those digits alone do not uniquely identify every employer. Its mailing-address fields and possible data-entry errors also limit automatic matching. The datasets do not provide a demonstrated one-to-one LCA-to-petition join for the job you are considering.
The current USCIS hub advertised coverage through FY2026 Q3 during this review. Its instructions route complete downloads through Crosstab View, filters and an Excel or CSV export. Our attempted employer query returned session and resource-limit errors. No employer decision total was successfully obtained from that query; an error is not a zero-result finding.
Worked example: Google history to a current role
The following is an actual, bounded research chain. It is not a claim that a historical LCA corresponds to the current vacancy.
In DOL's FY2026 Q3 disclosure file, case I-200-25267-331421 names Google LLC and the occupation Software Engineer. It was received September 24, 2025, decided October 1, 2025, and marked Certified. It requests one worker under H-1B, with continued employment equal to one and new employment equal to zero.
The employer city is Mountain View, California; the first worksite is Sunnyvale, California. This single row demonstrates both the fiscal-year boundary and the difference between employer address and worksite. It does not describe an unfilled Sunnyvale engineering vacancy.
We stopped after the first five qualifying Google LLC H-1B records containing software in the title. Three indicated continued employment and two indicated change of employer. This deliberately limited inspection is neither an employer total nor a representative sample.
Next, we opened Google's official career search and the Software Engineer detail, ID 137359440260014790. It displayed a Mountain View position and an Apply link. The requirements included a relevant bachelor's degree and two years of experience, with specified development, algorithms and project-lifecycle experience. The responsibilities included engineering work on AI-related systems.
During this October 12 review, no role-specific sponsorship commitment was found in the inspected body. The Google brand and historical Google LLC record make the company a useful research lead; they do not establish the current employing entity, a link to that LCA, or a sponsorship decision for an applicant.

Turn the chain into a usable shortlist entry
Here is the completed entry resulting from that research. The decision labels are editorial judgments about the evidence, not an employer response.
| Shortlist field | Completed value |
|---|---|
| Historical source | DOL FY2026 Q3; case I-200-25267-331421 |
| Filing employer | Google LLC; original spelling retained |
| Historical activity | Certified H-1B LCA for continued software-engineering employment |
| Current opportunity | Google Software Engineer; ID 137359440260014790; Mountain View |
| Evidence date | October 12,2026 |
| Entity connection | Brand-to-filer connection is a lead; current employing entity unconfirmed |
| Role sponsorship | Unconfirmed in the inspected posting |
| Next action | Check technical fit, then ask the verified recruiting contact about this requisition and required support |
The completed entry identifies a role worth checking and the two questions that remain. A technically suitable candidate can keep the role under consideration while seeking a specific answer. Someone who does not meet the listed experience requirements should not let a historical filing count override that separate fit assessment.
For example, a fictional candidate, Mira, needs employer support for her next U.S. role. Her completed note is: “Relevant employer history found; current role fits my background; employing entity and sponsorship still need confirmation.” She does not change it to “sponsor confirmed” because she found more historical records.
Her fictional, unsent question asks whether the identified requisition can consider candidates needing her accurately described type of support and which legal entity would employ the successful candidate. It requests a current policy answer, not a prediction of an immigration outcome. An ambiguous reply leaves the corresponding field unresolved.
Choose the source for the question you have
Use LCA disclosures when you want an initial list connected to occupations and worksites. Use USCIS decisions when you want the agency's recorded H-1B decision history under its published categories. Use the current employer posting and verified recruiting channel when you need to know whether a particular opening can support your circumstances.
Keep all three in separate columns. A missing historical record should trigger proportionate checking, not automatic rejection. A large historical count should not erase a current posting's restriction. Current policy can vary by role, entity, location and timing.
LandOffer has an interest in job-search tools; an ordinary spreadsheet is sufficient for this evidence chain. For your next search, choose one recent role, save one relevant historical record, and write the precise employer question that remains. Keep role sponsorship marked unconfirmed until the employer answers that separate question.


