Hello — I’m the Harborfield Assistant. I can explain Harborfield, compare reports, or help you find a relevant starting point. Final product access is determined separately through eligibility.
Loading
Preparing the requested page…
Loading
Preparing the requested page…
Hello — I’m the Harborfield Assistant. I can explain Harborfield, compare reports, or help you find a relevant starting point. Final product access is determined separately through eligibility.
Sources & context
Harborfield uses relevant public evidence to support defined research questions. This page explains source selection and interpretation; it is not a public dataset or benchmark database.
Research architecture
TraceableSource-to-report method
Provenance and assumptions stay connected through validation, analysis, QA, and delivery.
Interpretation
ControlledA source must be relevant to the question being examined. We consider who produced it, how the information was collected, the geography and population it describes, its reference period, and the limitations on comparison. Authority alone does not make a broad statistic suitable for a narrow service question.
Recency is assessed against the decision and the source's publication cycle. A recently retrieved page can still describe an older period.
Census sources can provide business, demographic, and economic context for a defined U.S. geography. The relevant dataset, reference year, and industry or population definition must accompany any observation.
Nonemployer Statistics describe businesses with no paid employees. Nonemployer, employer, and establishment measures are not interchangeable: the specific dataset determines what is counted. A broad industry category may include activities outside the customer's defined service.
A count or other contextual measure does not establish purchases, demand for a particular offer, or the likely outcome of a solo practice.
BLS sources provide employment, wage, and labor-market context. The series or program, geography, occupation or industry classification, and reference period determine what a measure can support.
Employee wage statistics are not solo-practice revenue, owner income, profit, or a recommended service price. Employment counts and self-employment measures also describe different populations. Comparisons must preserve those distinctions.
Where relevant to a customer research question, public service descriptions, pricing pages, and directories can support observations about visible offerings. Service scope, pricing basis, included work, and retrieval date matter when deciding whether observations are comparable.
These observations are a non-exhaustive sample unless the source supports completeness. A listing or advertised price does not establish sales, customer volume, utilization, or business performance. This description does not promise that every engagement has usable observations from every source family.
Missing evidence is not silently replaced with zero. An older observation retains its reference period and is not presented as a current measurement merely because it was retrieved recently.
Important gaps, conflicting observations, and limits on comparison remain visible. If the evidence does not support a conclusion or a calculation, Harborfield narrows the statement or identifies the result as unavailable rather than inventing a substitute.
General public context shown on Harborfield helps explain the research approach. It is not a personalized finding about a visitor's business and is not a substitute for a defined customer research engagement.
Sources used for a customer deliverable are selected for that engagement's scope and recorded with the relevant provenance and limitations. Private customer information and engagement-specific source records are not published by this page.
Citing or using an external source does not mean that its publisher endorses Harborfield or its interpretations. This product uses the Census Bureau Data API but is not endorsed or certified by the Census Bureau. BLS cannot vouch for data or analyses derived from its data after retrieval.
Continue