Salary Transparency in 100 Fortune 100 Management Vacancies: August 2026

Abstract

MTF Institute analysed 100 distinct U.S. management vacancies captured from eight employers in the 2026 Fortune 100 on 10 August 2026. The study asks how often the sampled advertisements disclosed a numerical base-pay or salary range, how disclosure patterns differed by employer and functional family, and how wide the published ranges were.

Eighty-four vacancies contained at least one numerical range: 77 were coded on an annual basis and seven on an hourly basis. Fourteen advertisements displayed more than one range, usually because an employer distinguished national and higher-cost-location ranges or listed multiple role levels. Among disclosed primary ranges, the median width was 50.0% of the range midpoint. These findings describe this purposive 100-vacancy snapshot; they are not a population estimate, legal-compliance assessment or market-pay benchmark.

Report number: MTF-RR-2026-08-21-01
Publication date: 21 August 2026
Author: MTF Institute Editorial Team
DOI: 10.5281/zenodo.22043595
Archival files: Searchable PDF and supporting workbook are available in the Zenodo record.

Research question

In a defined sample of 100 current U.S. management vacancies from Fortune 100 employers:

  1. How often was a numerical base-pay, salary or hourly-pay range visible?
  2. How did disclosure frequency vary across the sampled employers and functional families?
  3. How wide were the published ranges relative to their midpoints?
  4. What should candidates and managers infer—and not infer—from a posted range?

Scope and method

Sampling frame

The unit of analysis is one distinct U.S. management vacancy. The frozen corpus contains 100 unique URLs captured on 10 August 2026 from Amazon, Apple, Alphabet (Google), UnitedHealth Group, CVS Health, McKesson, Microsoft and JPMorgan Chase. All eight employers were members of the 2026 Fortune 100 when the corpus was constructed.

The sample spans 10 title-based functional families: engineering/data/technology; operations/facilities/construction; commercial/marketing/partnerships; programme/project/delivery; legal/risk/compliance; product management; health/care delivery; finance/banking; design/content/experience; and general/organizational leadership.

This is a purposive, diversified vacancy sample, not a random sample of all Fortune 100 vacancies or all U.S. management work. Employer counts range from five to 18, and only eight Fortune 100 employers are represented.

Inclusion rules

A vacancy was included when the captured employer page:

  • was a distinct U.S. management-role URL;
  • contained the job title and substantive vacancy description;
  • belonged to one of the eight sampled Fortune 100 employers; and
  • was available in the frozen 10 August research corpus.

Duplicate URLs were excluded. Related-job recommendations were not treated as observations.

Pay-range coding

A vacancy was coded range disclosed when its captured body contained a numerical lower and upper amount explicitly presented in salary, base-pay, total-target-cash or hourly-pay context.

The method recognized:

  • annual ranges labelled annually, per year, base pay or salary;
  • hourly ranges labelled hourly or containing values consistent with the employer's stated hourly-pay context;
  • “between X and Y” wording;
  • ranges displayed with USD in parentheses; and
  • additional location- or level-specific ranges.

The first listed qualifying range was treated as the primary range. Where a page displayed multiple ranges, this was recorded separately. Annual and hourly amounts were never combined into one midpoint analysis.

For every primary range:

midpoint = (lower bound + upper bound) / 2

relative width = (upper bound - lower bound) / midpoint

The relative-width measure allows ranges at different salary levels to be compared without treating a USD 50,000 span as equally broad in every context.

Validation

The corpus contains 100 records and 100 unique source URLs. Employer and functional-family totals reconcile to 100. Extracted low and high outliers, all hourly ranges, non-disclosures and multi-range records were reviewed against the captured vacancy text. The coding workbook preserves one row per vacancy and the source URL.

Results

1. Numerical ranges appeared in 84 of 100 vacancies

Measure Result
Valid distinct vacancies 100
Employers 8
Functional families 10
Numerical ranges disclosed 84
Disclosure rate in this sample 84.0%
Primary annual ranges 77
Primary hourly ranges 7
Vacancies with multiple ranges 14
Median relative width 50.0% of midpoint
Median annual midpoint USD 181,600

The median annual midpoint is descriptive of the 77 annual primary ranges in this particular sample. It should not be used as a universal management salary benchmark: the sampled employers, seniority mix, functions and locations are uneven.

2. Employer patterns dominated the sample result

Employer Vacancies Ranges disclosed Sample disclosure rate Median relative width
Apple 18 16 88.9% 48.0%
Alphabet (Google) 15 15 100.0% 32.6%
Amazon 13 12 92.3% 30.0%
UnitedHealth Group 13 13 100.0% 56.4%
JPMorgan Chase 12 0 0.0% Not applicable
McKesson 12 11 91.7% 50.0%
Microsoft 12 12 100.0% 63.4%
CVS Health 5 5 100.0% 41.7%

The result is polarized. Four sampled employers displayed a numerical range on every captured page, while none of the 12 sampled JPMorgan Chase pages did. This is not a claim about every vacancy or a permanent employer policy. It is evidence about the pages in the frozen corpus on the capture date.

3. Disclosure varied less by function than by employer

Functional family Vacancies Ranges disclosed Sample disclosure rate
Engineering, data and technology 16 15 93.8%
Operations, facilities and construction 15 14 93.3%
Commercial, marketing and partnerships 12 11 91.7%
Program, project and delivery 12 11 91.7%
Legal, risk and compliance 10 6 60.0%
Product management 9 7 77.8%
Health and care delivery 8 8 100.0%
Finance and banking 7 4 57.1%
Design, content and experience 6 4 66.7%
General and organizational leadership 5 4 80.0%

These rates should be interpreted cautiously. Each functional group mixes employers, levels and locations, and several cells are small. The table is useful for locating differences inside the sample, not for ranking entire professions.

4. A posted range was often broad

The median primary range width was 50.0% of its midpoint. In practical terms, the typical disclosed interval in this sample covered a large span rather than a narrow expected offer.

Fourteen postings contained multiple ranges. Examples included separate U.S. and higher-cost-location ranges or more than one job level on the same page. A candidate who copies the widest visible number without identifying the applicable level and location may misread the posting.

The range is therefore a boundary, not an offer prediction.

What the data does not show

The analysis does not identify the offered salary, final negotiated amount, bonus, equity, benefits, working hours, job quality or total compensation. It does not adjust annual midpoints for location, seniority, employer, function or company size. It does not evaluate whether an employer satisfied a particular state or local law.

Some pages described base pay; others used salary, hourly pay or total target cash language. The coding records the visible label and basis but does not force unlike compensation concepts into one comparison.

Practical application: the RANGE-5 interview method

Candidates can convert a published range into five decision questions.

R — Role level

Which internal level, grade or title does the range describe? Multi-level postings can contain several ranges and different expectations.

A — Applicable location

Which range applies to the work location or remote arrangement? A national range and a named high-cost-market range are not interchangeable.

N — Non-base compensation

What bonus, commission, equity or other compensation is separate from base pay? Compare like with like before evaluating total value.

G — Grounds for placement

What evidence influences placement inside the range—scope, rare skills, leadership experience, location, internal equity or another factor? Prepare examples tied to those criteria rather than arguing from the maximum alone.

E — Evolution

How does compensation change after entry? Ask about review timing, progression through the range, promotion boundaries and whether the role can be re-levelled.

What managers and HR teams should do

For hiring managers, a range is more useful when the job description also clarifies role scope, level, location logic and the relationship between base and variable compensation.

For HR and compensation teams:

  1. align the published range with the approved job level;
  2. distinguish base pay from total target cash and incentives;
  3. explain geography-specific ranges consistently;
  4. prevent multi-level postings from obscuring which range matches which responsibilities;
  5. review whether unusually broad ranges are informative to candidates; and
  6. retain the source and date behind market decisions.

MTF's Job Grading and Salary Bands: A Governance Audit provides a complementary internal-governance framework. The Degree vs Experience study examines a different selection signal in the same broader vacancy-research programme.

Readers who want to deepen their ability to connect job architecture, workforce decisions, change and people analytics can review MTF Institute's Executive Certificate in Strategic Human Resources, Leadership & Change.

Limitations

  • The sample is purposive and covers eight Fortune 100 employers, not all 100 companies.
  • Employer counts are unequal, ranging from five to 18 vacancies.
  • Vacancy pages can change or close after the 10 August capture date.
  • Functional families are based on title taxonomy and simplify hybrid roles.
  • Numerical extraction was validated against captured text but does not replace employer clarification.
  • Posted ranges are not accepted offers or market-price observations.
  • The study is descriptive and does not establish causal explanations for employer differences.

Data and archival record

The Zenodo record contains the searchable report PDF, a 100-row coding CSV and a supporting workbook with methodology, employer results and functional-family results: 10.5281/zenodo.22043595.

References