SBA context
36.2 million
U.S. small businesses reported by the SBA Office of Advocacy. This is a broad population context only.
Research proposal · evidence as of 2026-10-09
A market question, not a fabricated forecast.
This page separates national context, a limited survey sample, a proposed customer-research pilot and a future sizing method. It does not claim validated demand, customers, revenue or a current addressable-market estimate.
Evidence at a glance
NFIB sample
521 respondents
Small employers drawn from NFIB membership; this is not a representative sample of all U.S. businesses.
Hard boundary
No extrapolation
Do not multiply the SBA population by an NFIB percentage to invent a market or customer count.
Primary sources
What the two sources actually tell us
These links are the evidence boundary for the figures on this page. The sources provide context and a reason to investigate a problem; they do not establish AI Clouds Company OS demand or willingness to pay.
June 30, 2025 · SBA Office of Advocacy
A broad small-business population
The SBA report says the United States has 36.2 million small businesses. Use this as national population context only. SBA's total includes nonemployer businesses, so it is not interchangeable with an employer survey sample and is not an AI Clouds Company OS TAM or customer forecast.
Read the SBA sourceJune 25, 2025 · NFIB survey
A bounded website-adoption signal
NFIB reports that 82% of participating small employers had a business website and 18% did not. For employers with 1–9 employees, the reported split was 75% with a website and 25% without. A random sample of 20,000 owners from NFIB's membership database was invited, and 521 respondents completed the survey; it is not representative of all U.S. businesses.
Read the NFIB sourceProposed customer focus
Start with independent local service SMBs in the U.S.
The initial target is a research segment, not a claimed market total: owner-led or independently operated local service businesses that need a credible web presence and may have limited time, budget, technical skill or maintenance capacity.
Possible examples include home services, personal services and other locally delivered services. The final sector and geography should be chosen before sampling, with exclusions recorded rather than quietly broadened.
Owner-research pilot
- Name one U.S. city or metro and a narrow set of service sectors.
- Define a permission-respecting sample and record selection criteria, date and sample size.
- Interview owners about discovery channels, website pain, budget, upkeep and adoption blockers.
- Separate interest, a qualified problem, a pilot invitation, a signed commitment and a paid customer.
- Publish what the pilot could disprove, including weak pain or low willingness to pay.
Category framing
Alternatives to compare, without invented competitor claims
The relevant comparison is how an owner gets a trustworthy, useful and maintainable site—not a claim that one category always wins. A pilot should compare actual options available to the chosen segment.
Do-it-yourself builders
Compare setup effort, content responsibility, accessibility checks, maintenance and owner confidence.
Freelancers and agencies
Compare discovery, evidence collection, review cadence, handoff, cost structure and ongoing support.
Industry or vertical platforms
Compare bundled listings, templates, lead flow, portability and the degree of control the owner retains.
AI website builders
Compare speed with fact checking, human approval, editing control, source quality and maintenance expectations.
Sizing method
TAM, SAM and SOM are future calculations
No dollar value or customer count is asserted here. Once the pilot has a defined geography, segment, price hypothesis and evidence, the calculations can be shown with their assumptions and sensitivity rather than presented as facts.
TAM
Defined population
A documented count of businesses that fit the broad problem definition, using a named source and clear inclusions/exclusions.
SAM
Reachable segment
The subset in the chosen U.S. geography, service sectors and operating model that the planned offer can actually serve.
SOM
Testable near-term share
A conservative scenario based on capacity, acquisition channels, conversion evidence and retention—not a percentage selected to make a forecast look attractive.
Method note: The SBA and NFIB figures above are not inputs that can be multiplied together. Any future model must add its own dated, segment-specific source, assumptions, validation evidence and uncertainty range.
Value hypothesis
Why this might matter—and what remains unknown
Some smaller local businesses may need a first website. Others may already have one but need better content quality, accessibility, conversion support or upkeep. An evidence-checked production workflow could be valuable if it reduces owner effort while preserving factual accuracy and human control.
That is a hypothesis, not a measured outcome. It remains uncertain whether the problem is urgent enough, whether owners prefer this category of solution, what they can pay, how much ongoing support costs, and whether the proposed workflow creates better outcomes than the alternatives above.
Current internal engineering artifacts do not answer those questions. They are not customers, revenue, conversion rates or proof of market demand.
Continue exploring the candidate
Read the technical evidence pages or return to the customer-facing overview.