Pay belongs to the requisition
A company-level wage statement cannot replace the amount, basis, location, and date attached to the exact role.
HireTea Independent hiring research
Original research · 2026 edition
What job applicants can verify before they apply
Across 25 published employer dossiers, HireTea traced the public application route, status return path, posting identity, pay evidence, and schedule detail. This report describes what dated public sources established through 2026-07-12. It does not rate private hiring practices or predict an applicant's outcome.
25 employer dossiers coded
415 linked source records
94 dated field findings
37 structured job records
Executive finding
A visible Apply button is only the first layer of transparency
The official front door was the strongest indicator: 21 of 25 dossiers documented the applicable entry and handoff, while four retained only a partial route. The return path was much less consistent. Only 12 dossiers documented a scoped status route; 6 were partial and 7 were not established in the reviewed public material.
A record may not be an opening
An ID and location improve traceability, but evergreen or catalog records can still disclaim current headcount.
The return lane matters
Hourly, professional, franchise, property, and business-line applications can send candidates to different accounts.
25-employer evidence matrix
Public evidence was coded without turning missing information into a score
Each bar uses one coded row per employer dossier. “Not observed” means the reviewed public sources did not establish the indicator; it does not mean the employer lacks the information or process. The bars describe this research set, not all U.S. employers.
Documented
A dated public source supports every element required by this indicator.
Partial
Some public evidence exists, but at least one required element is missing, role-limited, indirectly accessible, or deferred to a later stage.
Not observed
The reviewed public sources did not establish this indicator. This is not evidence that the employer lacks it.
Application route
An official careers entry, the final Apply destination or system, and the applicable role, banner, or business scope were identified.
Status return path
An official post-submission destination or contact, the action to take there, and the applicable role scope were identified.
Posting identity
A job or requisition identifier, title, location context, URL, and observation date were retained.
Pay specificity
Numeric base pay and its hourly or annual basis were tied to an identifiable requisition, role, or location and a dated observation.
Schedule detail
Employment type and a shift, hours, days, or other actionable schedule detail were tied to the observed role.
Company-level data
Five evidence states for every dossier
This table is an evidence-coverage matrix, not a league table. A documented state can be narrow and role-specific; a partial state preserves a limitation instead of hiding it; and a not-observed state is never converted to zero. Coverage grade A, B, or C describes the retained sample depth, not employer quality.
| Employer | Apply route | Status path | Posting ID | Pay | Schedule | Evidence coverage |
|---|---|---|---|---|---|---|
| Walmart | Documented | Documented | Documented | Documented | Documented | B One or two current records were available, or access, representativeness, or role coverage was limited. |
| Amazon | Documented | Documented | Documented | Documented | Documented | B One or two current records were available, or access, representativeness, or role coverage was limited. |
| McDonald's | Documented | Partial | Partial | Documented | Partial | B One or two current records were available, or access, representativeness, or role coverage was limited. |
| The Home Depot | Documented | Documented | Partial | Partial | Partial | B One or two current records were available, or access, representativeness, or role coverage was limited. |
| FedEx | Partial | Not observed | Documented | Documented | Documented | B One or two current records were available, or access, representativeness, or role coverage was limited. |
| Target | Documented | Documented | Documented | Documented | Documented | B One or two current records were available, or access, representativeness, or role coverage was limited. |
| Kroger | Documented | Documented | Documented | Partial | Documented | A Three or more current official posting records were available for normalization. |
| UPS | Documented | Documented | Not observed | Partial | Partial | C A fresh current posting sample is still needed; retained records were evergreen, catalog-like, expired, or not posting-specific. |
| CVS Health | Documented | Documented | Documented | Documented | Documented | A Three or more current official posting records were available for normalization. |
| Costco Wholesale | Documented | Not observed | Documented | Partial | Not observed | C A fresh current posting sample is still needed; retained records were evergreen, catalog-like, expired, or not posting-specific. |
| TJX Companies | Documented | Not observed | Documented | Documented | Partial | A Three or more current official posting records were available for normalization. |
| Lowe's | Documented | Documented | Documented | Partial | Documented | A Three or more current official posting records were available for normalization. |
| Marriott International | Partial | Not observed | Partial | Partial | Not observed | C A fresh current posting sample is still needed; retained records were evergreen, catalog-like, expired, or not posting-specific. |
| Chipotle Mexican Grill | Documented | Partial | Partial | Documented | Partial | B One or two current records were available, or access, representativeness, or role coverage was limited. |
| Walt Disney Parks | Documented | Documented | Documented | Documented | Documented | B One or two current records were available, or access, representativeness, or role coverage was limited. |
| Publix | Documented | Partial | Partial | Documented | Documented | B One or two current records were available, or access, representativeness, or role coverage was limited. |
| Albertsons Companies | Documented | Partial | Documented | Documented | Documented | A Three or more current official posting records were available for normalization. |
| Starbucks | Documented | Documented | Documented | Documented | Documented | B One or two current records were available, or access, representativeness, or role coverage was limited. |
| Walgreens | Documented | Partial | Documented | Documented | Documented | A Three or more current official posting records were available for normalization. |
| Chick-fil-A | Partial | Not observed | Partial | Documented | Partial | B One or two current records were available, or access, representativeness, or role coverage was limited. |
| Alphabet / Google | Documented | Documented | Documented | Documented | Not observed | B One or two current records were available, or access, representativeness, or role coverage was limited. |
| Dollar General | Documented | Not observed | Documented | Partial | Partial | B One or two current records were available, or access, representativeness, or role coverage was limited. |
| Hilton Worldwide | Documented | Documented | Not observed | Not observed | Not observed | C A fresh current posting sample is still needed; retained records were evergreen, catalog-like, expired, or not posting-specific. |
| Apple | Documented | Partial | Documented | Documented | Partial | A Three or more current official posting records were available for normalization. |
| Taco Bell | Partial | Not observed | Partial | Documented | Partial | B One or two current records were available, or access, representativeness, or role coverage was limited. |
A Three or more current official posting records were available for normalization.
B One or two current records were available, or access, representativeness, or role coverage was limited.
C A fresh current posting sample is still needed; retained records were evergreen, catalog-like, expired, or not posting-specific.
Posting-level sample
Numeric pay appeared on 22 of 37 structured job records
The posting-level denominator is separate from the 25-employer matrix. It contains only structured records with an ID, title, URL, and observation date: 37 job postings or cards across 9 employers. Numeric pay tied to the exact record was documented on 22 (59%). This is a purposive research sample, not a U.S. market disclosure rate.
22/37 structured records with numeric pay tied to the record
All eight sampled TJX postings showed a starting range. Three of eight Lowe's records did. None of the eight Costco role records stated a location-specific starting rate, and every Costco record warned that it did not establish a current opening. All three sets were observed on 2026-07-12.
| Employer | Structured job records | Numeric pay on record | Share in this sample |
|---|---|---|---|
| Walmart | 2 | 2 | 2/2 |
| Amazon | 2 | 2 | 2/2 |
| McDonald's | 2 | 2 | 2/2 |
| The Home Depot | 2 | 2 | 2/2 |
| Target | 2 | 2 | 2/2 |
| Kroger | 3 | 1 | 1/3 |
| Costco Wholesale | 8 | 0 | 0/8 |
| TJX Companies | 8 | 8 | 8/8 |
| Lowe's | 8 | 3 | 3/8 |
The two Target Workday records retained pay in their official structured detail and source-backed fact even though the compact sampled-page summaries omitted the numbers. The downloadable posting CSV preserves the numeric evidence text, coding decision, and underlying URL for every retained record.
What the numbers hide
Five patterns that change an applicant's evidence checklist
01
The application lane controls the status lane
Amazon separated hourly candidate accounts from Amazon Jobs profiles. Home Depot separated hourly BrassRing from salaried Workday. UPS directed hourly candidates back through a unique message link and professional candidates to Workday. Saving only the company name is not enough to reconstruct the right return path.
02
A requisition can be identifiable without proving headcount
Costco's eight sampled records had identifiers, titles, locations, URLs, and observation dates. Each also warned that the role record did not mean a job was currently open. Posting identity and opening evidence are therefore different fields.
03
The logo may not name the employer
McDonald's franchise cards, Marriott and Hilton properties, licensed Starbucks counters, Albertsons banners, and Disney resort entities all require an operator or legal-employer check. The final offer controls payroll, benefits, and many policies—not the logo alone.
04
Matching titles can expose different evidence
Lowe's retained matching merchandising and fulfillment titles across locations, but only three of eight records displayed pay. TJX's eight sampled records all displayed a starting range. The useful comparison is record against record, not title against title in the abstract.
05
Schedule detail changes the meaning of pay
A range attached to seasonal status, part-time hours, an overnight window, a weekend minimum, or a fixed end date describes a different decision from the same number without schedule context. Twelve dossiers documented actionable schedule detail; nine were partial and four were not observed.
Source-linked evidence ledger
Why each company received its coded states
Every row below pairs a representative official route with the field finding that drove the coding and the limitation that remains. The company dossier contains the fuller source register and dated observations.
01
Walmart
BResult cards exposed facility type, location, shift, and pay before a requisition was opened.
Limit: Two current requisitions were retained; another role family would broaden the sample.
02
Amazon
BOfficial instructions separated the hourly candidate account from the profile-based Amazon Jobs route.
Limit: Two current data-center postings were retained, while the hourly search returned an automated-access error.
03
McDonald's
BRestaurant cards identified location and franchise-owned status before the application handoff.
Limit: One restaurant card and one corporate requisition were retained; the restaurant detail endpoint was region-limited.
04
The Home Depot
BApplication status was split between hourly BrassRing and salaried Workday routes.
Limit: Two official indexed posting records were available, but direct requests from the research environment returned access errors.
05
FedEx
BThe Company field and Apply vendor changed across two current warehouse postings.
Limit: The two retained records represented different FedEx businesses and did not establish one universal return path.
06
Target
BA warehouse requisition exposed two classifications, four shift windows, pay ladders, and an assessment before application.
Limit: Two current Workday records were retained; another role family would broaden the sample.
07
Kroger
AThree current requisitions exposed different banners, recruiting metadata, and pay visibility.
Limit: Only one of the three retained detail records exposed populated public pay fields.
08
UPS
CHourly candidates return through a unique message link while professional candidates use Workday.
Limit: No stable current individual requisition was retained for the employer-level snapshot.
09
CVS Health
AStore, pharmacy, and Aetna records exposed different entities, pay bands, schedules, and qualification gates.
Limit: The three retained roles are examples from distinct business lines, not a census of CVS Health openings.
10
Costco Wholesale
CAll eight role records carried a warning that the record did not establish a current opening.
Limit: The catalog-like records linked to a role pay table but did not state location-specific starting rates or current headcount.
11
TJX Companies
AEight OPEN posting records showed materially different local starting ranges and role duties.
Limit: The sample covered selected U.S. store and loss-prevention roles observed on one date.
12
Lowe's
AEight current records showed that matching titles could expose pay in one location and omit it in another.
Limit: Only three of the eight retained records displayed a populated public pay range.
13
Marriott International
CThe search surface separated Marriott, franchise, and third-party operator context from the hotel brand.
Limit: The retained corporate requisition had expired, so a fresh current posting sample is needed.
14
Chipotle Mexican Grill
BA Chicago crew page preserved store, job number, pay, employment type, and a dated refresh signal.
Limit: One current crew record was retained, and its refresh date was not treated as the original posting date.
15
Walt Disney Parks
BTwo current resort records tied legal employer, status, pay, schedule, and job ID together.
Limit: The two examples represented different resorts and legal entities rather than a company-wide pattern.
16
Publix
BA current non-retail requisition exposed pay, status, schedule, age, licensing, and response timing.
Limit: The current sample was non-retail; the retail route uses evergreen applications that do not prove a store opening.
17
Albertsons Companies
ACurrent records separated the parent company, banner, requisition, location, and final employing entity.
Limit: Banner and portal evidence still requires confirmation against the legal employer named in the final offer.
18
Starbucks
BTwo current postings tied different work, pay, expiration dates, and application state to exact requisitions.
Limit: Licensed Starbucks counters can use a different employer and application system from company-operated stores.
19
Walgreens
AStore, pharmacy, shift-lead, and pharmacist records exposed distinct duties, pay, licensing, and benefit conditions.
Limit: The public support route does not establish a single status dashboard or identical status labels for every role family.
20
Chick-fil-A
BOne official location page exposed local pay and employment type before its vendor application opened.
Limit: Four openings came from one Operator location and cannot establish another Operator's process or schedule.
21
Alphabet / Google
BTwo current professional requisitions tied role identifiers, windows, qualifications, and pay to exact pages.
Limit: The retained records did not establish actionable work-hour or shift information.
22
Dollar General
BStore number and requisition distinguished two Sales Associate records that shared a broad title.
Limit: Only some current store records exposed numeric pay, and the conditional-offer route is not an application-status path.
23
Hilton Worldwide
CThe public application path returned candidates to a profile while employer identity still varied by property.
Limit: No current individual posting with a retained identifier, numeric pay, and schedule was available in the snapshot.
24
Apple
AFour current retail, technical, corporate, and hardware records exposed role-specific pay and qualification evidence.
Limit: The nationwide seasonal card showed a weekly-hours field that requires store-level confirmation.
25
Taco Bell
BA current restaurant posting identified its operator, pay, age gate, physical requirements, and local benefits.
Limit: One detailed restaurant sample cannot establish another operator's vendor, status path, or schedule.
Methodology and codebook
Two units of analysis, no composite score
Employer-level indicators use one coded row per dossier. Posting-level findings use only the 37 structured records with an ID, URL, and observation date. These denominators are never pooled. The employer set is HireTea's first 25 published dossiers, not a probability sample or a ranking of the 25 largest U.S. employers.
1. Source scope
Coding starts with official employer career pages, current postings, public applicant FAQs, candidate-account guidance, official vendor handoffs, and employer or operator notices. The 415 linked records resolve to 406 unique URLs. Source registers preserve their individual access dates; the employer field-report snapshots were reviewed from 2026-07-11 through 2026-07-12.
2. Manual normalization
A reviewer mapped the source evidence into five named indicators using the documented, partial, and not-observed codebook. Keyword counts and the site's older benchmark scores were not used. Missing evidence was not imputed, averaged, or converted into a penalty.
3. Posting sample
The structured sample contains 46 page observations across 9 employers. Only 37 records typed as a live posting, live result card, or official indexed posting enter the posting denominator. Search pages, process pages, application forms, and program pages stay outside it.
4. Pay coding
Posting-level pay is documented only when a numeric amount or range belongs to the exact record. Company averages, separate pay tables, broad title ranges, and benefit copy do not enter the numerator. A range's maximum is not treated as the likely offer.
5. No application submission
HireTea did not create candidate accounts, submit applications, complete assessments, or claim firsthand applicant outcomes. Login-only status labels, recruiter response times, interview quality, and actual offers are outside the measurement scope.
6. Snapshot integrity
The report and both CSV files are generated from one versioned cohort module. The build fails if the expected
25 dossiers, 415 source records, 94 findings, 46 page observations, 37 job records, or 22 pay-documented records
drift without an explicit report update. It also verifies the complete row-level snapshot against SHA-256
3faae8ef73434fe1141d9310fb1730ed4cbb32646787b65bdebcdbb7c154bd71, so a changed URL, date, observation, or record cannot
silently retain version 2026.1.
| Indicator | Required for documented |
|---|---|
| Application route | An official careers entry, the final Apply destination or system, and the applicable role, banner, or business scope were identified. |
| Status return path | An official post-submission destination or contact, the action to take there, and the applicable role scope were identified. |
| Posting identity | A job or requisition identifier, title, location context, URL, and observation date were retained. |
| Pay specificity | Numeric base pay and its hourly or annual basis were tied to an identifiable requisition, role, or location and a dated observation. |
| Schedule detail | Employment type and a shift, hours, days, or other actionable schedule detail were tied to the observed role. |
Limitations
What this report cannot establish
Public evidence changes, postings expire, and candidate accounts reveal information that anonymous research cannot inspect. A second location can differ from the sampled one even when the company and title match. Multiple application systems do not by themselves prove separate recruiting teams or screening rules. A not-observed state is a research gap, not proof of absence.
The employer cohort was selected from HireTea's first published dossiers and overrepresents large retail, restaurant, logistics, hospitality, and consumer employers. The structured posting sample is concentrated in nine companies because only those dossiers used the current sampled-page schema. Narrative examples from the other 16 dossiers inform their company-level coding but do not enter the 37-record posting denominator.
Do not infer market prevalence
The 22/37 pay result describes retained records only, not the U.S. labor market.
Do not infer private practice
The report measures public evidence, not internal hiring quality, speed, or fairness.
Recheck before applying
Open the current posting and save its ID, location, date, pay, schedule, employer, and final Apply destination.
Open research files
Audit the coding or reuse the raw observations
Both CSV files are free, require no account, and use explicit state labels. The company file contains one row per dossier with evidence reasons and limitations. The posting file contains the 37 structured records, their URLs, dates, numeric pay evidence where observed, pay-disclosure state, opening-state caveat, and retained observation text.