Reference data

Retail industry statistics

How large US retail actually is, how it breaks down by subsector, how long new stores last, what returns cost the industry, and how retailers get financed. Every figure traces to a federal dataset or to the organisation that produced it, and every calculation Fundur made is shown in full.

Establishment, employment and payroll figures are Census County Business Patterns, reference year 2023. Firm counts are Statistics of US Businesses 2022. Sales are Census Monthly Retail Trade Survey, calendar year 2025, not seasonally adjusted. Survival figures are BLS Business Employment Dynamics, cohorts tracked to March 2025. Returns figures are the National Retail Federation’s 2025 Retail Returns Landscape. SBA figures are computed by Fundur from SBA’s 7(a) FOIA loan-level file, snapshot 30 June 2026. Last reviewed 28 August 2026.

1,043,373retail establishments in the United States

$7.519Tin US retail sales, calendar 2025

84.4%of new retail establishments survive year one — above the 77.9% all-industry rate

$849.9bnof merchandise returned in 2025, on the NRF’s measure

Industry size

Retail by subsector, in stores and in people

County Business Patterns is the Census Bureau’s annual census of establishments with paid employees. It counts locations rather than companies, and excludes businesses with no employees. Retail trade is NAICS 44–45.

US retail establishments, employment and annual payroll by subsector, 2023
NAICS	Subsector	Establishments	Employees	Annual payroll
44–45	Retail trade, total	1,043,373	16,046,207	$605.7bn
445	Food and beverage stores	160,298	3,333,898	$97.7bn
448	Clothing and accessories stores	122,162	1,505,822	$38.9bn
441	Motor vehicle and parts dealers	119,780	2,015,007	$131.2bn
453	Miscellaneous store retailers	116,482	831,551	$25.9bn
447	Gasoline stations	109,174	1,016,960	$29.7bn
446	Health and personal care stores	89,276	1,066,142	$43.0bn
454	Nonstore retailers	87,725	1,108,365	$63.4bn
444	Building material and garden supply	72,618	1,424,157	$55.4bn
452	General merchandise stores	59,516	2,666,463	$81.8bn
442	Furniture and home furnishings	44,942	409,563	$17.6bn
451	Sporting goods, hobby, book and music	42,352	471,722	$12.6bn
443	Electronics and appliance stores	19,048	196,557	$8.5bn

12.48%

of all US establishments with employees are retail establishments, employing 11.48% of the private-sector workforce. Retail has proportionally more locations than jobs, because the average store is small.Fundur calculation from Census County Business Patterns 2023, US summary file

15.4

employees per retail establishment on average, against 16.7 across all US private employers. Annual payroll per employee is $37,746, against $68,342 across all private employers — a headcount measure, not a full-time wage.Fundur calculation from County Business Patterns 2023

$7.519T

in US retail sales in calendar 2025. Food services and drinking places are a separate $1.181 trillion line in the same survey and are not included in the retail total.Census Bureau, Monthly Retail Trade Survey time series · not adjusted, twelve months summed

Business structure

645,404 companies, 1,045,890 stores

Statistics of US Businesses counts firms rather than locations. The gap between the two numbers is the whole chain-store industry, and it is smaller than the retail landscape’s reputation suggests.

1.62

establishments per retail firm on average. The great majority of retail companies operate a single location.Fundur calculation from Statistics of US Businesses 2022 · NAICS 44–45

59.9%

of retail firms have fewer than five employees, close to the 63.0% rate across all US employer firms. 90.8% have fewer than 20 employees.Census Bureau, Statistics of US Businesses 2022

16.41%

of calendar 2025 retail sales were e-commerce — $1.234 trillion of $7.519 trillion, on the same Census survey. That share is a transaction method, not a set of companies; the distinction is set out on our e-commerce statistics page.Fundur calculation from Census Bureau, Quarterly Retail E-Commerce Sales, four quarters of 2025 summed

Survival

New retail establishments outlast the average new business

BLS has tracked every private-sector establishment opening since 1994. Retail trade sits above the all-industry survival curve at every horizon it publishes, and the gap widens with age.

Share of establishments still operating, by years since opening. Each row is a different opening cohort, because a cohort has to age before it can be read at that horizon.
Years since opening	Retail trade	All private industry	Gap	Cohort measured
1 year	84.4%	77.9%	+6.5 pts	opened year to March 2024
5 years	59.8%	51.4%	+8.4 pts	opened year to March 2020
10 years	44.2%	34.7%	+9.5 pts	opened year to March 2015

Across the 31 cohorts BLS has measured to a full year, retail’s worst first year was the one opening in the year to March 2008, of which 80.4% survived. Its best was the year to March 2017, at 91.4%. First-year retail survival has never been below 80% in the series.

Two limits worth carrying into any citation. The series counts establishments, so a store that closes on relocation, on sale, or as one unit of a surviving chain is recorded as a death. And it is a survival series, not a profitability series — a store still open at year five may be doing badly.

Returns

$849.9 billion of returns, and the denominator problem behind it

The National Retail Federation, with Happy Returns, surveys retailers annually on returns. It is the standard source, and it is routinely quoted against the wrong denominator.

15.8%

of surveyed retailers’ annual sales were expected to be returned in 2025, totalling $849.9 billion — down from 16.9% and $890 billion in 2024.National Retail Federation and Happy Returns, 2025 Retail Returns Landscape

19.3%

of online sales were expected to be returned in 2025, against 15.8% across all channels. The online premium is the single most reliable finding in the series.National Retail Federation, 2025 Retail Returns Landscape

of all returns are fraudulent, on the same survey.National Retail Federation, 2025 Retail Returns Landscape

15.8% and 11.3% are both correct, and they are not the same statistic. The 15.8% is a share of the sales of the retailers NRF surveyed. Set the same $849.9 billion against Census’s measured $7.519 trillion of total US retail sales for calendar 2025 and it is 11.3%. The gap is the denominator: Census retail trade includes motor vehicle dealers, gasoline stations and other lines where returns are rare or structurally impossible, and NRF’s respondent base is not the Census universe. Quoting “15.8% of US retail sales are returned” conflates the two. Quote it as NRF states it — a share of surveyed retailers’ sales — or do the Census arithmetic and say which you have done.

Financing

Retail SBA lending has the widest gap between the average loan and the typical one

SBA publishes every 7(a) approval as a loan-level record. The figures below are computed by Fundur from that file for fiscal years 2020 to 2025, with cancelled approvals excluded.

SBA 7(a) approvals, FY2020–FY2025, cancelled approvals excluded. Computed by Fundur from SBA loan-level data.
	Retail trade (44–45)	All 7(a)
Approvals	37,312	302,593
Share of all 7(a)	12.33%	100%
Total approved	$22.23bn	$159.22bn
Median approval	$250,000	$198,300
Mean approval	$595,659	$526,186
Mean as a multiple of median	2.38×	2.65×
New or not-yet-open borrowers	33.0%	33.5%
Approvals later cancelled	12.70%	12.93%

The median retail 7(a) approval is $250,000 and the mean is $595,659. Reporting the mean alone — which is the more commonly published figure — describes a loan more than twice the size of the one a typical retail borrower actually received.

Retail borrows in line with its footprint: it holds 12.48% of US establishments and took 12.33% of 7(a) approvals over six years. Its share of new-and-not-yet-open borrowers, 33.0%, is also at the all-industry rate of 33.5% — unlike food services, where half of all approvals go to businesses that are not open yet.

Method

How these numbers were produced

Census figures are read directly from the published data files rather than from a secondary summary. Establishment, employment and payroll counts come from the County Business Patterns 2023 US summary file, using the rows where legal form of organisation is set to all; the subsector rows sum exactly to the retail trade total shown. County Business Patterns covers establishments with paid employees only. Firm counts come from Statistics of US Businesses 2022, the most recent year published, which is why the firm and establishment counts in this page come from different reference years and differ slightly.

Sales figures are the twelve not-seasonally-adjusted monthly estimates for calendar 2025 in the Monthly Retail Trade Survey time series, summed. The e-commerce share is the four not-adjusted quarterly e-commerce estimates for 2025 divided by the four matching total-retail estimates in the same release; those four quarterly totals sum to exactly the $7,518,692 million annual figure in the monthly series, which is the arithmetic check that they are the same universe.

Survival figures are quoted directly from BLS Business Employment Dynamics Table 7, which reports survival since birth for every annual opening cohort from 1994 onward. A cohort can only be read at a given age once it has reached it, so the one-, five- and ten-year figures come from three different cohorts, each named in the table.

Returns figures are quoted as the National Retail Federation states them. NRF surveys retailers; it does not measure the Census retail universe. The 11.3% recalculation in the box above is ours, is arithmetic only, and is labelled as ours.

SBA figures are computed by Fundur from SBA’s 7(a) and 504 FOIA loan-level file, 7(a) extract, snapshot dated 30 June 2026 (SHA-256 beginning 6c1e9132, 181,130,871 bytes). From 388,338 parsed rows we keep complete fiscal years 2020 to 2025, gross approval above zero and a valid approval date, then exclude the 12.93% of approvals recorded as cancelled and never disbursed, leaving 302,593 approvals. Industry assignment uses the NaicsCode field on each record. The file contains 368 exact duplicate rows in this window, 0.11%; removing them does not move the retail median at all and moves no percentage here by more than 0.02 points. No rate filter is applied because this page publishes no interest-rate figure; our SBA rate work uses a narrower 302,513-approval population and is documented on our SBA rates and fees analysis.

What these numbers cannot tell you. County Business Patterns and Statistics of US Businesses count only employers, so every establishment and firm figure omits the sole proprietor with no payroll. The BLS series counts establishments, not owners or brands, and cannot separate a failure from a relocation or a sale. NRF’s returns survey is a survey of retailers, with a respondent base that is not the Census universe and is not published as a probability sample. The SBA file records approvals, not applications, so it says nothing about who was declined, and covers only the 7(a) programme. Retail subsector definitions here are NAICS 2017 as used in these Census releases; the 2022 revision reorganised the retail sector and later files are not directly comparable at subsector level.

Reuse and citation

You are welcome to reproduce any figure or table on this page in noncommercial reporting, research or educational material, with attribution and a link back to this page. No permission request is needed. Where a figure is a Fundur calculation rather than a published statistic it is labelled as one above, and we would rather you carried that label through.

If a figure here contradicts one you have published, we would rather you checked our method than took it on trust — it is set out in full above, and every input file is linked.

Suggested citation: Fundur, “Retail Industry Statistics,” last reviewed 28 August 2026. https://fundur.com/resources/retail-industry-statistics/

Sources

Every source on this page

US Census Bureau — County Business Patterns, 2023 US summary data file · establishments, employment, annual payroll
US Census Bureau — Statistics of US Businesses, 2022 US and state detailed employment sizes · firms by employment size
US Census Bureau — Monthly Retail Trade Survey sales time series, 1992 to present · calendar 2025, not adjusted
US Census Bureau — Quarterly Retail E-Commerce Sales · 2026 Q2 release, 18 August 2026
US Bureau of Labor Statistics — Business Employment Dynamics, Table 7: Retail Trade and Total Private · last modified 17 December 2025
National Retail Federation and Happy Returns — 2025 Retail Returns Landscape · press release 15 October 2025
US Small Business Administration — 7(a) & 504 FOIA loan-level data · 7(a) extract, snapshot 30 June 2026

Fundur is a financing marketplace, not a lender. If you run a store and are working out what your own business could borrow, our retail business loans page covers the options and what lenders weigh.