Home / Blog / 8 ProfessionLens Ranking Lists for Workforce Research
8 ProfessionLens Ranking Lists for Workforce Research
ProfessionLens, 2026-11-20
21,462 institutions formed the comparison population for the 2025 Center for World University Rankings, which ranked the top 2,000, according to CWUR's publication page. That's not a workforce count. It illustrates a rule that applies to workforce ranking lists too: the population behind the published order matters as much as the order itself.
Our professional counts use public licensing rolls and federal files pulled on a stated count date. The published reference period is September 2026, with individual source releases documented separately. A file can contain duplicates, outdated affiliations or lapsed licenses. A count describes what survives the published counting rules, not everything a reader might infer about employment, availability or performance.
For research and citation, require a reference date, base count and named source file. Without them, you can't reproduce the result or cite it as a verified count. The Synopsix people analytics roadmap provides context for connecting workforce analysis to decisions, but the evidence still needs an inspectable foundation.
The eight list types below follow a decision sequence. Start with employer counts for brokerages, health systems, carriers and engineering firms. Then adjust for geography through state and county rates. Use executive titles to investigate documented leadership roles, and licensing retention to examine continuity. For each, the useful question is the same: what was counted, from which file, on what date, and what would change the answer?
Table of Contents
- 1. Brokerage Size Ranking by Licensed Agent Headcount
- 2. Health System Ranking by Physician Count
- 3. Insurance Carrier Ranking by Appointed Producers
- 4. Engineering Employer Ranking by Professional Engineer Headcount
- 5. States Ranked by Licensed Healthcare Professionals per Capita
- 6. Counties Ranked by Nurse Practitioners per 100,000 Residents
- 7. Employer Ranking by Executive Title Holders
- 8. Licensing Retention Ranking by State
- How the seven ranked lists differ in what they count and where the numbers come from
- Turn a Ranking into a Decision You Can Defend
1. Brokerage Size Ranking by Licensed Agent Headcount
Licensed agent headcount measures recorded affiliation, not sales performance. Start with the state real estate licensing file named on the count page. Record its release date, the extraction date and the license statuses included before comparing brokerage totals.
A brokerage can have a substantial licensed footprint without having the highest transaction volume. Don't translate affiliated agents into employees, full-time agents or productive agents unless another source establishes those categories. The narrower claim is more defensible: the file associates these distinct licensees with this brokerage under the stated rules.

The check that changes the answer
Normalize brokerage identities before ranking them. Preserve the original legal name, then add a documented parent or brand mapping. Otherwise, you risk splitting a business across subsidiaries or combining independent franchises that shouldn't share a total.
For a recruiting comparison involving eXp Realty and Compass, first define the territory. Then compare distinct eligible licensees affiliated with each organization on the same reference date. Keep unresolved affiliations separate rather than assigning them through a loose name match.
- Status filter: State whether inactive, expired or suspended records are excluded.
- Counting unit: Distinguish unique people from licenses and branch affiliations.
- Entity rule: Explain whether the total represents a legal brokerage, brand or consolidated group.
The firmographic attributes for account scoring framework is useful context for treating organizational scale as an attribute rather than a quality judgment. For workforce research, attach the licensing-file definition to that attribute. A larger recorded footprint supports a recruiting-market comparison, but it doesn't establish higher productivity or better agent retention.
2. Health System Ranking by Physician Count
Say a workforce planner wants to compare Kaiser Permanente and Cleveland Clinic. The first task isn't sorting physician totals. It's deciding what affiliated physician means and whether the available records support that definition.
Name the exact NPPES release, CMS clinician file or state licensing extract used, along with its reference date. These files serve different purposes. Don't assume that a physician's appearance at an address establishes employment by the health system associated with that location.
A defensible count needs an explicit relationship rule. That might be a documented organizational affiliation or a defined match between clinician and organization records. Label the result according to that rule rather than presenting every association as an employee relationship.
Practical rule: A physician can be counted accurately as an individual and still be assigned incorrectly to an organization.
Separate identity from affiliation
Deduplicate individual clinicians before aggregating them by health system. An individual National Provider Identifier can help resolve repeated clinician records, but it doesn't establish the employer relationship. Keep the identity check and the affiliation check separate.

For a regional recruitment decision, rerun the comparison using physicians with documented practice locations inside the target region. A national affiliation total answers a different question from the local pool available to recruit.
The count cannot establish appointment availability, clinical workload or care quality. It also doesn't show whether a physician divides time across organizations. Report those limits beside the total, and preserve ambiguous matches in an exceptions file. The decisive check is whether changing the affiliation rule changes which system leads.
3. Insurance Carrier Ranking by Appointed Producers
Count appointed producers, not presumed employees. The relevant evidence is the state insurance department's appointment file, identified by release or extraction date. The claim should describe the carrier, jurisdiction, appointment status and eligible producer population.
An appointment-based ranking concerns a recorded distribution relationship. It doesn't establish that every producer actively sells the carrier's products, works exclusively with that carrier or generates comparable premium volume. Those are separate questions requiring separate records.
For a market-entry assessment, an analyst could compare carriers serving the same state and line of authority. The analyst should retain carrier identifiers and producer identifiers rather than relying on company names alone. Corporate naming differences can otherwise create false splits.
Keep relationships and people separate
A producer can appear in multiple appointment records. Before sorting carriers, build distinct outputs for the questions you're asking:
- Unique producers: Deduplicate eligible people within each carrier.
- Appointment records: Retain qualifying producer-carrier relationships under the file's rules.
- Coverage notes: Identify excluded statuses, missing identifiers and unresolved carrier mappings.
Say a carrier appears to lead when every appointment row is counted. Another carrier leads after producer deduplication. Neither ordering is automatically wrong, but they describe different units and must carry different labels.
A distribution-network count is not a market-share estimate.
The check that most often changes the interpretation is the unit of count. Test whether the ranking holds when you switch from appointment rows to distinct producers. Then verify that all carriers were assessed using compatible status rules and the same reference period. If state files differ in coverage, present the state comparisons separately rather than manufacturing a national league table from incompatible extracts.
4. Engineering Employer Ranking by Professional Engineer Headcount
An engineering employer ranking should begin with eligible licensed professional engineers, not every employee whose title contains engineer. Name the state board file, its count date and the status filter. Then explain how each licensee was connected to an employer.
The caveat belongs before the conclusion: a licensing board record may establish a credential without establishing current employment. If the employer field is absent or stale, the credential alone can't support an employer ranking. Keep those people in an unassigned category unless a documented source resolves the relationship.
For a procurement shortlist, an analyst might compare firms' documented licensed engineering footprints within the project's jurisdiction. That can help frame a capacity question. It doesn't establish available project hours, relevant experience or the staff assigned to a proposed contract.
Test the geography and identity rules
A person licensed in several states shouldn't become several people in a consolidated headcount. Preserve jurisdiction-specific license identifiers, then document how records were resolved to individuals. Show both the geographic coverage and any unresolved cross-state matches.
- Active status: Apply the published eligibility rule consistently.
- Discipline coverage: Report the available discipline classifications without inventing missing ones.
- Employer evidence: Distinguish board-record affiliation from separately documented employment.
The importance of defining the eligible population isn't unique to workforce lists. Investopedia's history of the Fortune 500 explains that the original ranking focused on industrial companies rather than every major business sector. Absence from a ranking can therefore reflect inclusion rules rather than small scale.
The decisive check is whether the employer order survives person-level deduplication across jurisdictions. If it doesn't, describe the original result as a license-record ranking, not a ranking of distinct engineers.
5. States Ranked by Licensed Healthcare Professionals per Capita
A per-capita ranking compares recorded workforce density, not total workforce size. Its numerator needs a named licensing file and count date. Its denominator needs a named population file and population reference year.
The caveat is substantial: licensed, active and practicing are different populations. A licensing roll can support a count of eligible license holders under its status rules. It doesn't automatically establish where those people deliver care, how much they work or whether they accept patients.
Use the definitions in our licensed healthcare professionals guide to keep the population label attached to the rate. Don't change licensed professionals to available providers when summarizing a result.

Read the numerator before the rate
For a workforce-planning comparison between Texas and California, display the underlying professional counts beside the population-adjusted results. Keep profession definitions consistent. Don't compare a combined healthcare population in one state with a single profession in another.
- Population alignment: Record the denominator's geography and year.
- License alignment: Apply compatible status and profession filters.
- Location alignment: Explain whether geography follows residence, mailing address or practice location.
Our counts use complete public files rather than samples, but completeness of extraction doesn't guarantee completeness of the underlying workforce picture. The rolls themselves can contain errors, duplicates or lapsed licenses.
The check that changes the answer is denominator and location alignment. Rerun the comparison using a consistent population vintage and the same geographic assignment rule. A state's rank can support a statement about recorded density. It can't, by itself, support a statement about access to care.
6. Counties Ranked by Nurse Practitioners per 100,000 Residents
The nurse practitioner count and county population year must travel with the rate. Name the state board or other public professional file used for the numerator. Name the population release used for the denominator, and state how addresses were assigned to counties.
The heading's standard rate unit doesn't solve the underlying location problem. A residential address describes where a licensee lives. A practice address describes a documented work location. Neither automatically identifies the population that receives that practitioner's care.
For regional planning, use the nurse practitioners by state comparison as context, then inspect county records under a consistent geographic rule. Don't interpret a state total as evidence that practitioners are evenly distributed within it.

Examine the county boundary
Say a recruitment team is choosing between neighboring counties. A practice cluster near the boundary could affect residents on both sides, while the ranking assigns the recorded practitioners to only one county. Examine adjacent counties before treating a low local rate as evidence of an isolated service gap.
Interpretation check: A low recorded county rate is a reason to investigate coverage, not proof that residents receive no care nearby.
Preserve unmatched addresses and failed county assignments in the coverage notes. Don't discard them and describe the retained records as the entire licensed population. In our tables, a share whose base is under 20 prints as n/a, and counts under 5 print as <5.
The decisive check is practice location versus licensed residence. If that change reverses the order, publish the distinction. The county ranking supports a geographically defined count, not an unsupported shortage designation.
7. Employer Ranking by Executive Title Holders
A documented title holder is not the same thing as a verified organizational position. Start with the report's reference date and its underlying file. Our executive-title dataset, list_titles_by_company.csv, carries company-level flags and people counts.
The source caveat is essential. Our title research uses press coverage, company pages and professional-network job titles, as specified in each report's methods. Profile titles are self-reported. They aren't filings, proxy statements or organizational charts.
A leadership researcher could compare organizations with identified chief financial officers or chief operating officers. The defensible conclusion concerns documented title matches within covered companies. It doesn't establish reporting lines, decision authority, compensation or how top-heavy an organization is.
Audit coverage before comparing shares
The denominator is companies with any person rows, not every company nominally belonging to a comparison list. We calculate title shares on that covered population. Missing people records aren't evidence that an organization lacks the role.
- Title definition: Preserve the published matching rule and accepted variants.
- Holder evidence: Record the source, source date and any current-status uncertainty.
- Validation status: Distinguish manually reviewed matches from unreviewed pattern matches.
Our methods distinguish review depth by role. For chief revenue officer matches, we read five matches per list by hand; the remaining output wasn't individually verified. A reader shouldn't describe that entire category as manually confirmed.
For a worked comparison, inspect whether co-presidents, divisional executives and former office holders qualify under the same rule. Then rerun the count after excluding ambiguous matches.
The check that changes the answer is often eligibility, not arithmetic. A broad title search can identify wording that a narrower organizational question would exclude. Publish the matching rule so another analyst can test that boundary.
8. Licensing Retention Ranking by State
Retention requires a starting cohort and a later observation date. Identify both public licensing extracts, their reference dates and the rule for matching records. Without those elements, a change in total license count isn't a retention measure.
The caveat comes before any workforce conclusion: an active license indicates status in the observed file. It doesn't establish continued work in the profession. A missing or inactive record doesn't, by itself, prove retirement, migration or occupational exit.
Our Florida real estate first-renewal research provides a specific licensing question to inspect through its cohort definition and underlying rows. Don't generalize a profession-specific renewal result to every licensed workforce in a state.
Follow the same people through time
Use stable license identifiers where possible. Document how replacements, identifier changes and duplicate records were handled. Names alone are a weak matching key because spelling changes and shared names can create false exits or false continuations.
For a state comparison, define equivalent observation points relative to each jurisdiction's renewal schedule. Otherwise, you could compare a cohort already observed after renewal with another still awaiting its renewal deadline.
- Starting cohort: State who qualified at the initial date.
- Later status: Separate active, lapsed, surrendered and unresolved records where available.
- Comparable exposure: Align the time each cohort had to reach the event being measured.
Say a recruitment planner wants to identify a stable professional population. Retention can help frame that investigation, but it must be paired with evidence about actual work and geography before supporting a staffing decision.
The decisive check is renewal-cycle alignment. Recalculate using comparable cohort timing before interpreting differences between states. Our research reports include their underlying CSV rows so readers can inspect the matching and exclusions.
How the seven ranked lists differ in what they count and where the numbers come from
ProfessionLens publishes seven ranked lists. The table covers those list types and the executive-title research view, which is built from separate report data. Each row is a checkable claim: what is counted, which file supplies it, where it applies, and what the resulting position can support.
| List or view | Profession counted | Primary source | Geography | Best for |
|---|---|---|---|---|
| Brokerage Size Ranking by Licensed Agent Headcount | Licensed real estate agents affiliated with each brokerage | State real estate licensing records | State, county, metro, or national | Comparing brokerage scale and agent pools |
| Health System Ranking by Physician Count | Physicians affiliated with or practicing at each health system | NPPES, CMS clinician records, and licensing records | Regional or national | Comparing health system workforce scale |
| Insurance Carrier Ranking by Appointed Producers | Licensed insurance producers with active carrier appointments | State insurance department appointment and licensing filings | State, by line of authority | Evaluating carrier distribution strength |
| Engineering Employer Ranking by Professional Engineer Headcount | Licensed professional engineers linked to employers | State engineering board licensing records | State or multistate | Comparing credentialed engineering capacity |
| States Ranked by Licensed Healthcare Professionals per Capita | Licensed healthcare professionals, measured per resident population | State licensing records and Census population estimates | U.S. states | Investigating workforce density and regional gaps |
| Counties Ranked by Nurse Practitioners per 100,000 Residents | Licensed nurse practitioners, measured per 100,000 residents | State licensing or federal clinician records and population data | U.S. counties | Investigating local provider concentrations and possible shortage signals |
| Executive Title Holders (research view) | Identified holders of defined executive titles | Press coverage, company pages, and self-reported professional-network job titles | Organizations or sectors | Studying documented leadership roles in a covered company population |
| Licensing Retention Ranking by State | Professionals who keep licenses active across reference periods | State licensing board records | States | Assessing continuity of recorded license status |
Read the source column before the rank column. Licensing rolls can carry errors, duplicates or lapsed licenses, so a high count records credentials on the count date rather than a live workforce census. Job titles in the executive view are self-reported on professional network profiles; they are not filings, proxy statements or org charts. A list position can show documented footprint or distribution. It cannot show productivity, available capacity or an employment relationship beyond the stated affiliation rule. The date and file behind each count belong beside the number whenever the list is cited.
Turn a Ranking into a Decision You Can Defend
Choose the measure before choosing the leading organization or location. Employer counts suit questions about documented professional footprint, competitive scale and recruiting pools. They don't establish productivity, available capacity or employment relationships beyond the stated affiliation rule.
Use state per-capita and county rates to investigate distribution. Keep the professional count, population denominator and location rule visible together. A rate without those components is difficult to interpret, even if the arithmetic is correct.
Match the evidence to the question
Executive-title counts suit questions about documented leadership roles within a covered company population. Licensing retention suits questions about continuity of recorded license status. Neither supports a broader organizational or labor-market conclusion without additional evidence.
The distinction matters when moving from reporting to action. The discussion of turning dashboards into business insight provides decision-making context, but a dashboard can't repair a mismatched counting unit. Analysts still need to establish that the displayed measure answers the operational question.
Before citing a ranking, record:
- List identity: Name the list and the population eligible for inclusion.
- Reference date: Separate the count date from the publication date and source release.
- Base count: Give the denominator behind every share and suppress shares with an insufficient base.
- Source file: Identify the licensing roll, federal extract or documented company-source file behind the rows.
- Reproduction rule: Preserve the filters, deduplication key, entity mappings and exclusions.
Then test the definition most likely to alter the result. For brokerages, that may be parent-company consolidation. For health systems, it may be affiliation. For geographic rates, it may be address type. For retention, it may be the renewal observation window.
Cite the narrowest supported conclusion
Write that an organization had the largest recorded eligible count within the defined comparison, rather than calling it the largest employer without qualification. Write that a county had a lower recorded density, rather than declaring a shortage from the rate alone. These aren't cosmetic distinctions. They keep the conclusion within what the rows can establish.
A ranking without a date and a source is a claim, not evidence. Downloadable rows exist so another reader can rerun the deduplication, inspect exclusions and challenge the grouping rule. You don't need to accept the published order on trust, and your readers shouldn't have to accept yours that way either.
ProfessionLens counts licensed professionals and executive titles by profession, geography and employer, with dated sources and documented methods. Visit ProfessionLens to find a comparison relevant to your workforce question, inspect its counting rules and use the underlying research rows to verify the result before citing it.