The mobile mechanic is invisible to the federal government.

No instrument of the US statistical system measures mobile automotive repair. Each one misses it for a different, documentable reason. This is the first original research report from Mechanics Alliance, and we are publishing the method before we publish any number.

How this was built. Every source below was fetched, downloaded and parsed. Every number was read out of a file, not quoted from an article. Where a claim could not be verified against a primary source, it is labeled unverified and left in. That labeling is the point: it is what makes the rest citable.

Published 9 September 2026. Sources are named inline so anyone can reproduce this.

The one finding that organizes everything else

There is no official count of mobile mechanics in the United States. Not a rough one, not a dated one. None. Not because the data is hidden or expensive, but because every existing instrument samples a frame that excludes mobile operators by construction.

Auto Care counts bays. Ratchet+Wrench and ASA survey shop owners. NADA surveys dealership payroll. ASE Connects surveys employers about employees. BLS OEWS excludes the self-employed outright. Nobody is hiding this data. Nobody has a sampling frame that could produce it.

Five instruments, five different blind spots

01  ·  Bureau of Labor Statistics

Two federal counts of the same trade, 121,000 apart

The Occupational Employment and Wage Statistics survey counts 704,640 automotive service technicians and mechanics. Employment Projections, which does include the self-employed, counts 825,800. The roughly 121,000 gap is, essentially, everyone who works for themselves.

OEWS is built on employer payroll records. A person who works for themselves never appears on one. Employment Projections Table 1.2 puts self-employment in this occupation at 14.6%, and that single national percentage is the ceiling of what the federal government will tell you about this population. No wage attached. No state. No metro. No split between someone who owns a shop and someone working out of a van.

Source: BLS OEWS May 2025, SOC 49-3023, national file. BLS Employment Projections, occupation.xlsx, released 27 August 2026.

02  ·  NAICS classification

It names mobile repair, then files it with the garages

The 2022 NAICS Index File contains exactly one mobile repair entry: 811111 — Mobile automotive and truck repair services. So the classification system does name the activity. It then files it into the same code as fixed-location general garages, diesel shops and fleet maintenance.

There is no separate code, no sub-code, and no type-of-operation flag. The Economic Census variable that could split businesses by delivery model is scoped to Wholesale Trade only, not to repair.

This is not an oversight and we will not call it one. NAICS classifies on the nature of the work, not the delivery model, and it folds mobile into 811111 on purpose. Stated that way the finding is unrebuttable, which matters more than stating it dramatically.

Source: US Census Bureau, 2022 NAICS Index File. Cross-checked against the 2022 NAICS Descriptions file, which carries no mobile wording in any repair code.

03  ·  Census business data

One survey sees the self-employed but not deeply enough

Nonemployer Statistics is the only federal series that captures businesses with no paid employees, which is most of this trade. It publishes NAICS to five digits only, so 81111 is as deep as it goes. County Business Patterns reaches six digits but counts employers exclusively. Neither can isolate mobile.

The two also cannot be joined below five digits, because CBP 2023 is built on NAICS 2017 while NES 2023 uses NAICS 2022. The join looks clean and is wrong.

Source: Census Nonemployer Statistics 2023 (nonemp23us.txt) and County Business Patterns 2023 (cbp23us.txt), both read directly from the bulk files.

04  ·  OSHA

A solo operator is outside the agency entirely

Federal injury recordkeeping does not reach a one-person business. The regulation is explicit: self-employed individuals are not covered by the OSH Act.

The practical consequence is that no injury data exists for this trade at all. Not high numbers, not low numbers. None. Anyone claiming a mobile mechanic injury rate is not reading a federal source.

Source: 29 CFR 1904.31(b)(1).

05  ·  The industry's own instruments

The trade counts itself in bays

The Auto Care Association Factbook, the industry's flagship reference, reports 269,548 service outlets. Its capacity unit is bays and outlets. A metric built on premises cannot, by construction, count an operator who has none.

This is worth saying carefully. The industry is not excluding mobile operators out of malice. It built its measuring stick before this way of working existed, and the stick has no room for a van.

Source: Auto Care Association Factbook, 36th edition.

What that costs the people doing the work

Invisibility is not an abstraction. These are questions with real money attached that no existing survey has any reason to ask, because every one of them is unique to working without premises.

0
Surveys that ask what a callout costs
A trip fee exists only in mobile work, so no shop-based survey has ever had a reason to ask about it. There is no instrument in existence that could produce this number.
0
Studies of the mobile billable hour
Shop benchmarks are abundant. The mobile equivalent has never been measured, and windshield time between jobs is the dominant cost with no shop analogue.
0
Measures of what a comeback costs
A shop comeback is a bay. A mobile comeback is an entire second trip. The metric means something materially different and nobody tracks it.
0
Auto-repair-specific collection times
No premises means no vehicle held as collateral, which changes the collection dynamic fundamentally. No source publishes this for auto repair at all.

What the data does show

Where the federal record can see this trade, even partially, the picture is not the one the industry describes when it talks about shops.

54%
gross under $25,000 a year
Of 373,301 nonemployer auto repair businesses. Mean receipts are $60,577, and the typical business is nowhere near it.
60.2%
have one to four people
102,158 of 169,572 auto repair establishments that do have employees. This industry is not made of big shops.
23.5%
first-year exit rate
For auto repair establishments in year one. By year 26 it falls to 7.1%. This is an annual exit rate, not a cohort survival rate.

Sources: Census Nonemployer Statistics 2023 and County Business Patterns 2023, NAICS 8111. Census Business Dynamics Statistics 2023, the only federal source reaching four-digit auto repair for failure data.

There is a way to count this trade, and nobody has used it

A first credible national estimate does not require waiting for members to join and fill in a survey. It requires two files that already exist.

1. California is the countable state

California requires an Automotive Repair Dealer registration to charge for repair, and regulation requires a separate registration for each vehicle used for mobile repair. The Business and Professions Code separately requires the application to state a plate number if the applicant is engaged in mobile automotive repairs. A mobile-only operator is therefore directly distinguishable in California's licensee file. No other state produces this.

2. Three more states corroborate the definition

Florida names mobile motor vehicle repair shops and self-employed individuals in statute. New York makes a mobile unit a regulated category with its own signage rule. Hawaii references each mobile repair facility as registered with its board. Four independent registries, four different definitions, one cross-check.

3. Scale it with Census

Nonemployer Statistics gives state-level establishment counts for 81111. Apply the California-derived mobile share, corroborate against the other three registries, and carry the uncertainty band honestly rather than hiding it.

That is a real methodology, and this page is us publishing it before we publish the number. We will also publish the state files we used, so the estimate is reproducible by anyone who doubts it. Reproducibility is what makes a number get cited, and being cited is the entire point.

One caution we will hold ourselves to: a survey of self-selected founding members produces a sample that is not representative of the mobile population, and we will never present one as if it were. The registry route is what makes a population estimate defensible, and it can be done before membership reaches any threshold at all.

What we could not verify

Publishing the gaps is not a weakness in a research report. It is the reason to trust the rest of it.

Method

Primary sources only. Every federal file named here was downloaded and parsed rather than quoted second hand. Where a widely repeated industry claim failed against its own primary source, we recorded the failure rather than repeating it. Figures from job boards and salary-aggregation sites are deliberately excluded from this report: they are self-reported, their methodology is undisclosed, and they do not survive the standard we are applying to everyone else.

Principal sources: BLS OEWS May 2025; BLS Employment Projections Table 1.2; BLS QCEW 2025; Census Nonemployer Statistics 2023; Census County Business Patterns 2023; Census Business Dynamics Statistics 2023; 2022 NAICS Index File and Descriptions file; IRS SOI Sole Proprietorship Table 2, TY2023; 29 CFR 1904.31; California Business and Professions Code and CCR Title 16; Auto Care Association Factbook, 36th edition.

We are going to count this trade.

Founding membership is free for the first 100 mobile mechanics, and founding members see the first national estimate before anyone else does.

Become a Founding Member Challenge a number