The question
How many companies rest on one person, what does that person hand over first, and what breaks when the team grows?
Method Twenty-two published sources from 2011 to 2026, most of them official statistics and peer-reviewed studies, each read on its primary release and marked with who ran it and whom it counted or asked.
Most companies are one person. In the United States, 29.8 million of 36.2 million small businesses have no employees1; in the United Kingdom, 4.1 million of 5.6 million2. What such a company needs in order to run, and what changes when it grows to ten people or to three hundred, is asked far more often than it is measured.
We read twenty-two published sources for an answer: thirteen releases of official statistics and official surveys from the United States, the United Kingdom and the European Union, seven peer-reviewed studies, one working paper and one company's study of its own records. No vendor survey is among them: the ones we found on how owners spend their time could not be traced to a primary release, so we left them out. Beside every reference we say what kind of evidence it is and whom it counted or asked. Every number below is taken from its primary release; the selection, and what we make of it, are ours.
Key findings
Most companies are one person. 82.3% of US small businesses have no employees1, and 74% of UK private sector businesses employ nobody but their owners2.
Few of them hire, and most never set out to. Of 4.4 million US businesses started without employees in 1997, 2.5% had hired seven years later3. Asked at the start, 24.3% of new owners wanted the business to be as large as possible4.
An owner with a team works a day more each week than an employee. In the European Union, self-employed people with employees work 47.5 hours a week full time, against 39.5 for employees5.
The first help is not an employee. 38% of US firms without employees used contract workers in a year, and 40% used AI, most of them for writing6.
With employees, the hard part becomes people. 16% of US firms without employees name hiring or keeping staff as a challenge; among firms with employees, 46% do6.
Software arrives with size. In the European Union, 41.1% of enterprises with 10 to 49 people use ERP software, and 88.7% of those with 250 or more7.
How many companies rest on one person?
Most of them. In the United States, 82.3% of small businesses, 29.8 million of 36.2 million, had no employees in the count for 20221. A year later the number of businesses without employees had reached 30.4 million, out of 36.4 million businesses of every size8. In the United Kingdom, 4.1 million of 5.6 million private sector businesses, 74%, employed nobody but their owners at the start of 20262.
Figure 1
Each step up in size holds a fraction of the companies below it
UK private sector businesses by number of employees, start of 2026
| Measure | Value |
|---|---|
| No employees | 73.7% |
| 1 to 9 | 21.5% |
| 10 to 49 | 4% |
| 50 to 249 | 0.7% |
| 250 or more | 0.15% |
The European Union counts differently and arrives at the same shape. Of 34.0 million enterprises in its business economy in 2024, 94.4% had fewer than ten people working in them9, and of 26.1 million self-employed people aged 15 to 64 in 2025, 17.7 million had no employee10. The share is not shrinking. In the United States, businesses with employees fell from 24.6% of all businesses in 2012 to 21.6% in 202311.
A business of one is often the whole of a working life, not a sideline. In a survey by the US central bank, 62% of firms without employees were their primary owner's only job6.
How many of them grow a team?
Few, and most never set out to. Asked while they were starting up, 24.3% of 1,214 new business owners in the United States said they wanted the business to be as large as possible. The rest wanted a size they could manage themselves or with a few key employees, and the median owner expected to employ four people after five years4.
Tax records show what follows. Of the 4.4 million US businesses that started without employees in 1997, 2.5% had hired a first employee seven years later3.
Figure 2
Of a hundred businesses started alone, two took on an employee within seven years
US businesses started without employees in 1997, seven years on
| Measure | Value |
|---|---|
| No longer reporting revenue | 85% |
| Still in business, no employee | 13% |
| Hired a first employee | 2% |
The picture changes for the minority that starts out to build a company. In a panel of new US businesses that the same authors describe as more oriented to growth, 38.0% of those that began without employees had hired one year after starting and 58.8% by seven years. When a business hires at all, it hires early: after the first few years, few do3.
Plans run ahead of the record. 30% of US firms without employees told the central bank's 2025 survey that they would add employees within twelve months6. In the United Kingdom, 58% of businesses with no employees aimed to grow sales over three years, 17% planned to recruit, and 47% planned none of the growth steps the survey lists12.
Hiring is also not a door that closes behind a company. Among US firms without employees that were more than two years old, planned to hire and had answered the survey in several years, 40% had employed someone in at least one year since 202013. Staying in business is its own test: across 1994 to 2022, 67.7% of new US establishments with employees survived two years and 49.2% survived five1.
Where does a founder's time go?
Into more hours than an employee works, and into many small pieces. In the European Union in 2025, self-employed people with employees who worked full time put in 47.5 hours a week in their main job. Those without employees put in 43.6, and employees 39.55.
Figure 3
An owner with a team works a day more each week than an employee
Usual weekly hours in the main job, full-time workers, EU, 2025
| Measure | Value |
|---|---|
| Employees | 39.5 h |
| Self-employed, no employees | 43.6 h |
| Self-employed with employees | 47.5 h |
What fills those hours has been watched, though rarely. Researchers followed twelve founders in Switzerland, Austria and Germany through four working days each, 542 hours in all. The six who ran young companies worked 11.8 hours a day on average, and 63% of the things they did took five minutes or less14. Their time was spread across every function of the company.
Figure 4
Administration and finance took more than a day of a start-up founder's week
Share of working time by function, six founders in the start-up stage
| Measure | Value |
|---|---|
| Product development | 21% |
| People and hiring | 16% |
| Marketing, sales and PR | 14% |
| Administration and finance | 27% |
No official statistic says how many hours an owner spends on administration. The figures in circulation come from surveys we could not trace to a primary release, and we do not repeat them. What official sources do count is the cost: federal paperwork cost US small businesses more than USD 81 billion in 2025, over 80% of it from the tax authority alone1.
What does a founder hand over first?
Work goes to people outside the company before it goes to an employee. 38% of US firms without employees used contract workers in the twelve months before the 2025 survey, more than the 30% that planned to hire6. In the United Kingdom, the adviser a business of one names most often is its accountant: 34% of those that sought outside advice, ahead of consultants and business networks at 21% each12.
- The books, half on paperAmong UK sole proprietors and partnerships without employees that file income tax returns, 55% keep paper records, 54% spreadsheets and 48% record keeping software. Software stood at 32% a year earlier12.
- Software to run the business47% of UK businesses with no employees use technology or web-based software to manage the business, against 70% of employers with fewer than 250 staff. For selling online the two are close: 26% and 29%12.
- AI, nearly as often as employers40% of US firms without employees used AI in late 2025, against 46% of firms with employees. One in ten of those users called it fully integrated6.
Figure 5
Solo owners give AI the writing first, and the back office far less often
What US firms without employees use AI for, among those that use it
| Measure | Value |
|---|---|
| Writing or marketing | 87% |
| Individual productivity | 63% |
| Planning or analysis | 61% |
| Administrative business functions | 36% |
| Customer service | 29% |
| Coding | 20% |
The uses say what is handed over: words first. Writing or marketing leads at 87%. Administrative business functions, the back office a solo owner carries alone, stand at 36%, and 51% of the users named accuracy as a challenge6.
Whether the help helps has been measured in one place we found, and the answer is uneven. In a field experiment with 640 small business owners in Kenya, an AI business assistant had no measurable effect on revenue or profit on average. Owners who were doing well before may have gained more than 15%, and owners who were struggling did nearly 10% worse. The difference lay in which advice they chose to act on, not in the advice they were given15.
Founders who start alone do not wait longer to hire. Among startups that keep their ownership records on one equity management platform, companies with a single founder rose from 23.7% of new startups in 2019 to 36.3% in the first half of 2025, and they took on a first employee a median of 399 days after incorporation, against 480 days for teams of founders16.
What changes with the first employees?
The list of problems changes, and people move to the top of it. In the 2025 survey, 16% of US firms without employees named hiring or retaining qualified staff as a challenge. Among firms with employees it was 46%. Rules move the same way, from 16% to 25%. Finding customers is the challenge that barely moves: 62% without employees, 57% with6.
Figure 6
With employees, staff becomes the problem a solo owner rarely names
Operational challenges of the past 12 months, US small firms
| Measure | No employees | With employees |
|---|---|---|
| Staff | 16% | 46% |
| Supply chain | 19% | 30% |
| Regulation | 16% | 25% |
An employer also meets obligations that a business of one does not have. In the United Kingdom, 37% of businesses with no employees named taxation as a major obstacle and 30% named regulations. Among employers with fewer than 250 staff, taxation was named by 67% and regulations by 47%12.
The founder's own day changes too. In the observation study, founders of companies in the growth stage spent 82% of their time communicating, against 64% for founders in the start-up stage, and 56% of their time with people inside the company, against 29%. Time spent working alone on analysis and concepts fell from 28% to 12%14. The sample is twelve people, so this is a description, not a rate.
What breaks between ten people and a few hundred?
The supply of the founder's own attention. Diaries of 1,114 chief executives of manufacturing firms in six countries, firms with 300 employees at the median, show a job that is almost all contact: the average executive spent 70% of working time with other people, in a week of about 50 hours17. A company grows past what one person can attend to only when decisions are handed over, and the studies of that step agree on where it fails.
- Delegation needs records the owner can trustIn a field experiment on 28 textile plants of 17 Indian firms, owners kept decisions such as purchasing to themselves, because they did not trust managers from outside the family and could not keep track of materials and money. The size of a firm followed the number of the owner's brothers and sons (a correlation of 0.689 with employment) far more closely than it followed how well the firm was managed (0.223). Once better records made the plants easier to see into, owners delegated more18.
- Founders run less structured companiesAcross 13,436 observations of manufacturing firms with 50 to 5,000 employees in 32 countries, firms led by their founder scored lower on basic management practices than firms under other forms of ownership, and the gap remained after country, industry, size, age and skills were accounted for. Managers in founder-led firms also tended to overestimate how well managed their firm was19.
- Flat does not scale by itselfIn a large sample of game development start-ups, a flatter structure went with more creative success, and with haphazard execution and commercial failure. A few levels of managers helped a start-up succeed commercially and survive20.
Figure 7
The larger the team, the more its obstacle is people
Share of UK employers, by number of employees, 2025
| Measure | 1 to 9 employees | 10 to 49 | 50 to 249 |
|---|---|---|---|
| Staff a major obstacle | 33% | 58% | 63% |
| Plan to recruit | 38% | 70% | 78% |
| Plan to develop managers | 31% | 56% | 73% |
Official surveys describe the same passage from the employer's side. Among UK employers, staff recruitment and skills were a major obstacle for 33% of those with 1 to 9 employees, 58% of those with 10 to 49 and 63% of those with 50 to 249. Plans to raise the leadership capability of managers rose from 31% to 56% to 73%21.
Which tools does a company take on at each size?
The books first, then systems for what the founder can no longer see. Among UK businesses that file income tax returns, record keeping software is used by 48% of those with no employees12, by 68% of employers with 1 to 9 employees, 79% of those with 10 to 49 and 94% of those with 50 to 249. Among employers that use any business technology, accountancy software is the most common kind, at 82%, followed by payroll software at 64% and electronic invoicing at 62%21.
Figure 8
Each size adds software, and the smallest firms counted have the least
EU enterprises using each kind of software, by size, 2025
| Measure | 10 to 49 people | 50 to 249 | 250 or more |
|---|---|---|---|
| ERP | 41.1% | 69.9% | 88.7% |
| CRM | 24.7% | 43.8% | 65.4% |
| Business intelligence | 11.5% | 34.3% | 69.2% |
| AI | 17% | 30.4% | 55% |
Statistics for the European Union start at ten persons employed and show the climb from there. In 2025, 41.1% of enterprises with 10 to 49 people used ERP software, and 88.7% of those with 250 or more. The widest spread is in software for analysing the company's own data, 11.5% against 69.2%7. We read that as the tool a company takes on when no one person can see the whole of it any more. AI, the newest of the four, already follows the same slope, from 17.0% to 55.0%22.
Among the smallest firms the order may be different. US firms without employees used AI nearly as often as firms with employees6, while 55% of the UK's sole proprietors and partnerships without employees still keep paper records12. A company of one may reach for an assistant before it has a system of record. None of the sources we read measures what that does to the company later.
What this does not show
- A business of one is counted in several ways. The US count takes every tax filer with receipts of USD 1,000 or more in a year11, the UK count estimates unregistered businesses from a survey2, and the EU counts persons employed, owners included9. The shares agree in shape and cannot be compared to the point.
- The record of first hires is old. The cohort followed through tax records started in 1997, and the panel of new businesses in 20043. Both come before the software and AI a founder has today.
- Nobody has measured a solo owner's week at scale. The observation covers twelve founders14, the diaries are of chief executives of manufacturing firms17, and the weekly hours are self-reported5.
- The central bank's survey is not a random sample. It is a convenience sample weighted to the population of firms, and its questions on AI were optional6.
- The studies of management describe, they do not prove. The comparison of founder-led firms is a correlation in manufacturing firms of 50 employees or more19, and the experiment on delegation covers 17 firms in one industry in one country18.
- One source is a company's own records. The figures on solo founders cover startups that use one company's product, not startups at large16.
- No source shows what software or AI does to growth. Whether an assistant lets a company stay at one person longer, or hire sooner, is not measured in anything we read. The one experiment on AI advice found no average effect15.
What we build from this
Each need below is one the evidence names. Beside it is what SalexHub does about it today, with the page that shows it. Where we have nothing to show, the row says so.
One place for the work one person does
Customers, deals, quotes, invoices, purchasing, stock, bookings and leave are in one app, with the Agent working across all of them.
The back office off the owner's morning
Say what the morning needs, and the Agent prepares the quote, the order for what is short, the reminders for late invoices and the total overdue. By default each piece waits for your yes.
A day in fewer pieces
A ready Mini reads your deals and contacts every weekday at 08:00 and reports the deals due or overdue and the customers waiting on your reply.
An accountant who stays
Ask in the chat for the quarter's issued invoices, and the Agent builds an XLSX or CSV file for your accountant. An export always waits for a person.
A way out of paper and spreadsheets
Import a CSV or Excel file into any list, match the columns, and see how many rows are ready, with look-alikes flagged, before anything is written. Undo import takes it back.
Handing over without losing sight
Set an approval rule, and anything over your amount waits for you or the person the rule names. By default the Agent shows each change as a card first, and only the Owner can let it act on its own.
A team that changes
When someone leaves, deactivate the member: sign-in stops at once, their records move to the colleague you choose, and nothing is deleted.
Proof in hours
We have not measured what SalexHub saves in an owner's week. Until we do, we make no claim about it.
Nothing to show yet
References
Grouped by the weight of the evidence, the heaviest first. Open a group to see its sources.
Peer-reviewed studyPublished in a journal after review by other researchers.07
- 03Taking the Leap: The Determinants of Entrepreneurs Hiring Their First Employee
Fairlie and Miranda. Journal of Economics and Management Strategy, 26(1), pages 3 to 34; read in its open version, IZA Discussion Paper 9848, 2017.
Who was askedAll 4,368,700 US businesses that started without employees in 1997, followed for seven years in tax records, and a panel of 4,928 businesses started in 2004.
Back to the text - 04What Do Small Businesses Do?
Hurst and Pugsley. Brookings Papers on Economic Activity, Fall 2011, pages 73 to 142; a journal of invited papers, each discussed by other economists before it is published, 2011.
Who was asked1,214 people in the United States who were starting a business, first asked in 2006, and 2,617 new businesses that survived four years.
Back to the text - 14What Do Entrepreneurs Actually Do? An Observational Study of Entrepreneurs' Everyday Behavior in the Start-Up and Growth Stages
Mueller, Volery and von Siemens. Entrepreneurship Theory and Practice, 36(5), pages 995 to 1017, 2012.
Who was askedTwelve founders in Switzerland, Austria and Germany, six in the start-up stage and six in the growth stage, each observed for four working days.
Back to the text - 17CEO Behavior and Firm Performance
Bandiera, Prat, Hansen and Sadun. Journal of Political Economy, 128(4), pages 1325 to 1369; read in its open version, NBER Working Paper 23248, 2020.
Who was asked1,114 chief executives of manufacturing firms in Brazil, France, Germany, India, the United Kingdom and the United States, each with a diary of one week.
Back to the text - 18Does Management Matter? Evidence from India
Bloom, Eifert, Mahajan, McKenzie and Roberts. The Quarterly Journal of Economics, 128(1), pages 1 to 51; read in its open version, NBER Working Paper 16658, 2013.
Who was asked28 plants of 17 textile firms in India, in a randomized field experiment.
Back to the text - 19Are Founder CEOs Good Managers?
Bennett, Lawrence and Sadun. Measuring Entrepreneurial Businesses: Current Knowledge and Challenges, National Bureau of Economic Research, University of Chicago Press, pages 153 to 185, 2017.
Who was asked13,436 observations of manufacturing firms with 50 to 5,000 employees in 32 countries, from the World Management Survey.
Back to the text - 20The myth of the flat start-up: Reconsidering the organizational structure of start-ups
Lee. Strategic Management Journal, 43(1), pages 58 to 92, 2022.
Who was askedA large sample of game development start-ups. Read in its abstract; the journal is closed.
Back to the text
Official sourceStatistics or rules published by a public authority.13
- 01Frequently Asked Questions About Small Business, 2026
US Small Business Administration, Office of Advocacy. Office of Advocacy, February 2026.
Who was askedAll US firms with fewer than 500 employees, from the Census Bureau's counts of firms with and without employees for 2022.
Back to the text - 02Business population estimates for the UK and regions 2026: statistical release
UK Department for Business, Innovation, Science and Trade. GOV.UK, official statistics in development, 5 November 2026.
Who was askedAll UK private sector businesses at the start of 2026; unregistered businesses are estimated from the Labour Force Survey and tax returns.
Back to the text - 05Average usual weekly hours worked in the main job by professional status (lfsa_ewhun2)
Eurostat. Eurostat database, EU Labour Force Survey, 2025 data, updated 30 October 2026.
Who was askedPeople aged 15 to 64 in employment in the EU; the figures used are for full-time workers.
Back to the text - 062026 Chartbook on Nonemployer Firms: Findings from the 2025 Small Business Credit Survey with Trends over Time
Federal Reserve Banks. Small Business Credit Survey, July 2026.
Who was asked11,552 US firms with fewer than 500 employees, 5,027 of them without employees, surveyed from September to November 2025. A convenience sample, weighted.
Back to the text - 07Integration of internal processes by size class of enterprise (isoc_eb_iip)
Eurostat. Eurostat database, ICT usage in enterprises, 2025 data, updated 27 February 2026.
Who was askedEU enterprises with 10 or more persons employed, financial sector excluded.
Back to the text - 08Census Bureau Releases New Data on Business Owner Characteristics
US Census Bureau. Census Bureau press release, Annual Business Survey and Nonemployer Statistics by Demographics, 20 November 2025.
Who was askedAll US businesses with and without employees, reference year 2023.
Back to the text - 09Enterprise statistics by size class and NACE Rev. 2 activity (sbs_sc_ovw)
Eurostat. Eurostat database, structural business statistics, 2024 data, updated 30 September 2026.
Who was askedEnterprises in industry, construction and market services in the EU, by number of persons employed.
Back to the text - 10Self-employed persons by educational attainment level (lfsa_esgaed)
Eurostat. Eurostat database, EU Labour Force Survey, 2025 data, updated 30 October 2026.
Who was askedSelf-employed people aged 15 to 64 in the EU, with and without employees.
Back to the text - 11Number of U.S. Nonemployers Grew Faster Than Employer Businesses Nearly Every Year From 2012 to 2023
US Census Bureau. Census Bureau, America Counts, 30 July 2025.
Who was askedUS businesses without paid employees, from federal tax records, 1997 to 2023.
Back to the text - 12Longitudinal Small Business Survey 2025: businesses with no employees
UK Department for Business, Innovation, Science and Trade. GOV.UK, 15 October 2026.
Who was asked2,439 UK businesses with no employees, interviewed by telephone and online. Some questions, those on technology among them, were put to about a third of the sample.
Back to the text - 132025 Report on Nonemployer Firms: Findings on Hiring Plans from the 2024 Small Business Credit Survey
Federal Reserve Banks. Small Business Credit Survey, October 2025.
Who was asked5,955 US firms without employees, surveyed from 4 September to 4 November 2024. A convenience sample, weighted.
Back to the text - 21Longitudinal Small Business Survey 2025: SME employers (businesses with 1 to 249 employees)
UK Department for Business, Innovation, Science and Trade. GOV.UK, 15 October 2026.
Who was asked8,022 UK employers with 1 to 249 employees, interviewed by telephone and online.
Back to the text - 22Artificial intelligence by size class of enterprise (isoc_eb_ai)
Eurostat. Eurostat database, ICT usage in enterprises, 2025 data, updated 15 June 2026.
Who was askedEU enterprises with 10 or more persons employed, financial sector excluded.
Back to the text
Working paperResearch released before peer review.01
- 15The Uneven Impact of Generative AI on Entrepreneurial Performance: Evidence from a Field Experiment in Kenya
Otis, Clarke, Delecourt, Holtz and Koning. Harvard Business School Working Paper 24-042, 2023, revised 2026.
Who was asked640 owners of small businesses in Kenya, half of them given an AI business assistant at random.
Back to the text
Company researchA company's study of its own data.01
- 16Solo Founders Report 2025
Carta, with Solo Founders. Carta Data, 9 December 2025.
Who was askedTens of thousands of US startups that keep their ownership records on the company's platform.
Back to the text
- Contains public sector information licensed under the Open Government Licence v3.0. It comes from the UK government's Business population estimates 2026 and the two reports of its Longitudinal Small Business Survey 2025, © Crown copyright 2026. The licence: https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/
- Suggested citation: "2026 Chartbook on Nonemployer Firms: Findings from the 2025 Small Business Credit Survey with Trends over Time." 2026. Small Business Credit Survey. Federal Reserve Banks. https://doi.org/10.55350/sbcs-20260707
- Suggested citation: "2025 Report on Nonemployer Firms: Findings on Hiring Plans from the 2024 Small Business Credit Survey." 2025. Small Business Credit Survey. Federal Reserve Banks. https://doi.org/10.55350/sbcs-20251014
- Both Small Business Credit Survey reports are licensed under a Creative Commons Attribution-NonCommercial 4.0 International licence: https://creativecommons.org/licenses/by-nc/4.0/. This piece takes facts from them and draws its own figures; it reproduces none of their text or charts.
Changes
- 9 November 2026First published.
