Small businesses do not share one staffing pattern, which is why small business administrative workload must be measured inside the company. The U.S. Small Business Administration Office of Advocacy counted 36,207,130 small businesses in its 2026 FAQ, using its research definition of an independent business with fewer than 500 employees. Of those businesses, 82.3 percent had no employees. That scale matters: a delegation decision designed for a 200-person company will not fit an owner-operator, and a single national benchmark cannot reveal how much administrative work is on your calendar.
This article combines official scale data with a workload model you can fill in using your own task log. The result is a capacity estimate, not a promise of revenue or savings.
What official data says about small-business scale
The SBA reports 36.2 million small businesses, including about 29.8 million nonemployer firms and 6.4 million employer firms. Its count combines the Census Bureau's Statistics of U.S. Businesses and Nonemployer Statistics programs. The Census Bureau separately reported 30,427,808 nonemployer establishments for 2023 and 5.58 million firms with at least one employee but fewer than 500 employees.
These totals use different reference years and units, so they should not be added together. A firm may operate more than one establishment. More important for planning, the official figures show that the small-business landscape is lean. Many owners have no employee to absorb scheduling, record updates, inbox triage, document preparation, or recurring follow-up.
The Census Bureau's Business Dynamics Statistics organizes measures by firm and establishment characteristics. It is useful for market context, but it does not measure the administrative hours inside your business. That gap is why the next step must use observed work instead of a borrowed industry percentage.
What the national data cannot tell you
The Bureau of Labor Statistics American Time Use Survey reports time spent working across broad employment and occupation groups. In the 2025 results, self-employed people who worked on an average day worked 6.14 hours on that day. This is not an estimate of administrative workload. It covers total work time, relies on population averages, and does not identify which tasks another person could safely perform.
Do not convert that figure into a claim such as "owners spend X hours on admin." The official sources cited here do not support such a statement. Measure the work directly instead.
A transparent administrative workload model
Track recurring work for two representative weeks. Record each task as it happens, including interruptions. Then calculate monthly demand with this formula:
Monthly task hours = occurrences per month × minutes per occurrence ÷ 60
Add a review allowance for delegated work:
Delegated capacity needed = total transferable task hours × (1 + review rate)
The review rate is your planning assumption. Write it down. For example, if observed transferable work totals 32 hours per month and you expect review to add 15 percent, the modeled need is 36.8 hours. This arithmetic does not assign a dollar value to the time and does not assume that every recovered hour becomes revenue.
| Work category | Monthly frequency | Minutes each | Monthly hours | Transfer test |
|---|---|---|---|---|
| Calendar and meeting coordination | 24 | 12 | 4.8 | Rules can be documented |
| CRM or customer record updates | 80 | 6 | 8.0 | Required fields are defined |
| Routine follow-up preparation | 40 | 10 | 6.7 | Approval boundary is clear |
| Invoice or order status checks | 20 | 15 | 5.0 | Exceptions route to an owner |
| Weekly report assembly | 4 | 45 | 3.0 | Source data is accessible |
The sample totals 27.5 hours before review. Replace every input with your own observations. Keep nontransferable judgment, regulated approvals, sensitive personnel decisions, and final financial authority outside the delegated total.
Decide what to delegate first
Start with tasks that are frequent, rules based, reversible, and easy to inspect. A recurring CRM update with required fields is a stronger first candidate than resolving an unusual customer dispute. Scheduling within documented constraints is easier to quality-check than negotiating a contract.
Use four questions for each line in the log:
- Can someone follow a written trigger, input, and definition of done?
- Can the owner review a sample or exception queue instead of every action?
- Can access be limited to the systems and fields required for the task?
- Can an error be detected and corrected before it causes material harm?
If the answer is no, redesign the process or retain the task. The daily sales VA task guide offers examples of recurring work, while the sales coordination guide shows how coordination responsibilities can be bounded.
Choose capacity only after the scope is stable
Convert the observed monthly hours into a practical weekly range, then add review time and predictable peaks. Avoid selecting a full-time role merely because the task list feels long. A stable 8 to 10 hours per week calls for a different arrangement than a queue that is active throughout every business day.
Compare delivery models only after the scope, access rules, service windows, and quality checks are written. The outsourcing versus hiring comparison explains the operating tradeoffs without treating one model as universally better.
For the first 30 days, measure queue age, completion accuracy, exceptions, owner review time, and missed handoffs. Those are observable operating measures. Revenue may change for many reasons, so do not attribute it to delegation without a credible comparison design.
A practical four-week delegation test
Week 1: Observe
Log task frequency, elapsed time, system access, and the reason each interruption occurred. Do not estimate from memory at the end of the week.
Week 2: Define
Write the trigger, inputs, steps, completion evidence, deadline, and escalation rule for the two most repeatable categories. Remove permissions that are not necessary.
Week 3: Pilot
Delegate a limited queue. Review a sample plus every exception. Record rework time separately so the model does not hide management effort.
Week 4: Decide
Compare actual demand, completion quality, queue age, and review load with the initial assumptions. Expand only if the process is stable. If review consumes most of the transferred time, improve the instructions or keep the work with its current owner.
Limits and source notes
The SBA and Census figures describe the scale and structure of U.S. small businesses. They do not measure an individual company's administrative burden. The BLS time-use figure describes total work time among self-employed people who worked on an average day, not delegable work. The sample workload table is an illustration created for this article and is not survey evidence.
Official datasets are revised and may use different reference years, definitions, and units. Check the linked source before using a figure in a board paper or staffing decision.
Build the workload map before you hire
Bring a two-week task log to a free consultation. We can help separate repeatable support work from owner-only decisions and turn the transferable queue into a role scope.