Exam Preparation and Tutoring in the United States

NAICS 2022 code 611691 — an investor's primer

1. Overview

This is the business of helping people pass a test or catch up in a subject: SAT and ACT college-admissions coaching, kindergarten-through-grade-12 (K-12) academic tutoring, and review courses for professional licensing exams (bar, CPA, medical boards, nursing, securities). It is a large, highly fragmented, labor-intensive service industry, not a conventional stock-market sector. Its customers are families, students, schools and districts, colleges, libraries, employers, and government programs.

Federal statistics count roughly 9,820 establishments, about 108,755 employees, and around $3.0 billion in annual payroll [1], with employer-firm receipts near $7.3 billion in 2022 [2]. As explained in Section 3, the true economic footprint is meaningfully larger because most tutoring is delivered by tiny or individual operators the official counts miss.

Why an investor cares: demand is driven by durable forces — competitive college admissions, pandemic-era learning loss, and a steady stream of people who must pass a credentialing exam to work. But the field is being reshaped in real time by free artificial intelligence (AI) tools, and it is overwhelmingly a private-market industry. Public-market ways in are few (essentially one U.S.-listed pure-play), so most capital enters through private routes: buying or operating a franchise center, private-equity ownership of the scaled brands, or venture bets on AI-tutoring startups. The best opportunities usually surface through company-level diligence rather than industry-wide market forecasts.

2. What it is and how it's structured

Scope. The North American Industry Classification System (NAICS) code 611691 covers establishments primarily engaged in offering (a) preparation for standardized exams and (b) academic tutoring [1]. In plain terms: test-prep courses, one-to-one and small-group tutoring, remedial help, and professional-exam review — delivered in person, online, as self-study products, or under school contracts. Classification is based on an establishment's primary activity, so a single brand can generate revenue across several NAICS categories.

What it excludes (named adjacent codes) [1][4]:

• Automobile driving schools — 611692 — and All Other Miscellaneous Schools and Instruction — 611699. Note that widely cited market-research figures often bundle 611691 with driving schools, inflating the number (see Section 3).
• Fine-arts schools (611610), sports and recreation instruction (611620), and language schools (611630).
• Elementary and secondary schools (611110), junior colleges (611210), and colleges and universities (611310). A tutoring center is not a school.
• Computer training (611420), professional and management training (611430), technical and trade schools (611519), and educational-support services (611710).
• Educational software publishing and pure content (parts of NAICS 5132/5182). A homework-answer app can straddle this boundary.

Ownership mix. Four structures coexist: (1) national franchise networks of bricks-and-mortar learning centers (Kumon, Mathnasium, Sylvan, Huntington); (2) online marketplaces and membership platforms (Varsity Tutors/Nerdy, Wyzant, Tutor.com); (3) exam-prep publishers that sell courses, question banks, and adaptive software (Kaplan, Becker, UWorld, Barbri, The Princeton Review); and (4) a very long tail of independent local tutors and small centers. That independent tail is the numerical majority of the industry.

3. How big it is

Federal ground-truth figures for NAICS 611691:

Average pay works out to roughly $27,700 per employee [1] — low, because a large share of the workforce is part-time hourly tutors. Payroll absorbs roughly 40% of employer receipts (comparing 2023 payroll to 2022 receipts) [1][2], confirming this is a people business, not a capital-intensive one. The Small Business Administration (SBA) size standard is a government-contracting classification, not an estimate of average firm revenue.

The undercount caveat matters here. County Business Patterns (CBP) measures establishments with paid employees, and the Economic Census likewise generally covers only employer establishments — so the ~$7.29 billion receipts figure counts employer firms only [1][2][4]. Academic tutoring is dominated by sole proprietors, gig tutors, and cash-paid private lessons that never show up in employer statistics; these nonemployer independents are a structurally large part of this specific industry. Census Nonemployer Statistics exist to capture them, but our ground-truth file does not include a 611691 nonemployer total, so we do not state a precise adjusted market size [5]. As a directional (not official) proxy, private market research that bundles tutoring with driving schools put the broader category near $18.9 billion in 2025-26, growing at a low-single-digit rate [6], while the online-tutoring slice alone is about $1.8 billion and has shrunk since its lockdown-era peak [7]. The honest read: the federal employer core is ~$7.3 billion, and the full academic-tutoring economy including independents is materially larger — most likely in the low-double-digit billions. No industry-wide profit figure, average hourly rate, student count, churn rate, or utilization rate is asserted here, because our federal file does not supply one.

Fragmentation. The industry is unconcentrated. The largest four firms hold about 22.1% of revenue, the top eight 29.9%, the top twenty 40.7%, and the top fifty 50.5% [2]. The Herfindahl-Hirschman Index (HHI, a standard concentration gauge that sums the squared market shares of all firms) is just 186.6 [2] — far below the ~1,500 level economists treat as the low end of "moderately concentrated." No single company dominates.

4. The investable universe

Public pure-plays are scarce. The table separates the one U.S.-listed pure-play from diversified and adjacent names; share prices, market caps, and multiples belong here, not in the prose above.

Major private and other owners (where most of the industry actually sits):

• Kumon — Japanese, privately held; the largest by footprint with roughly 25,000 centers across 50+ countries and a big U.S. presence. Kumon North America is a subsidiary of the Japan-based Kumon Institute of Education; local centers are generally independently operated franchises [17].
• Mathnasium — 1,000+ math-tutoring centers; owned by private-equity firm Roark Capital since 2021 [14].
• Sylvan Learning — franchise network acquired by Unleashed Brands in 2024; Unleashed Brands is backed by Seidler Equity Partners [15][16].
• Huntington Learning Center — privately/family-owned franchise.
• The Princeton Review + Tutor.com — private, and long private-equity-owned (historically by Primavera Capital, a Hong Kong-based firm); a September 2025 company statement described the businesses as then "100% U.S.-owned and operated" but did not name the current controlling owner [18]. Ownership is best treated as unresolved on the public record.
• Professional-exam publishers — Kaplan (test/professional prep), Becker (CPA), UWorld (medical/nursing/professional), Barbri (bar exam) — mostly private or PE-backed, high-margin content businesses.
• Wyzant — online tutor marketplace, owned by IXL Learning (private). Other notable private/franchise operators include C2 Education, Magoosh, Revolution Prep, ArborBridge, and Club Z; ownership and financials are generally not public.

5. How the money works

Owners make money on a few economic engines, all governed by tutor labor as the dominant cost. Revenue commonly comes from hourly or packaged tutoring, memberships, group classes, exam-prep courses, school/district contracts, franchise royalties, enrollment fees, and licensed content. The main costs are tutor compensation, center rent, local managers, customer acquisition, platform and content development, payment processing, refunds, and overhead.

1. Consumer-pay (retail). Parents and students pay out of pocket. Revenue = active students × sessions (or hours) × price, minus tutor compensation. The levers that matter are enrollments/active members and revenue per member; tutor utilization (billable hours per tutor) and retention/churn; and customer-acquisition cost, a real drag for online players competing for search traffic. Because tutor pay scales with delivery, gross margin is the spread between price charged and tutor wage — technology and group formats widen it.

2. Institutional / contract (B2B and B2G). School districts pay for "high-dose" tutoring (multiple small-group sessions weekly). The metrics are contracts signed, bookings, per-student pricing, and renewals — and this segment lives or dies on public funding cycles (Section 6).

Franchise economics are different again: the franchisor collects royalties (a percentage of each center's revenue) plus fees and bears little capital cost, while the franchisee funds real estate, labor, and local marketing. For the franchisor it is a low-capital, recurring royalty stream where center count and same-center enrollment are the growth levers; for the franchisee it is a local small business whose returns depend on utilization and neighborhood demand.

Exam-prep publishing behaves almost like software: content and question banks are expensive to build once, then licensed at high incremental margin (Kaplan, Becker, UWorld). This is the most defensible, highest-margin corner of the industry.

Useful operating metrics an analyst should demand include active learners and conversion; revenue per learner; paid instructional hours and tutor utilization; tutor pay as a percentage of revenue; same-center revenue and franchisee churn; school-contract bookings, renewals, and payment timing; and customer-acquisition cost, lifetime value, refunds, and cash conversion. Nerdy's 2025 filing illustrates the shape without providing an industry benchmark: ~33.2 thousand active members, ~15.8 thousand active experts, ~$364 average monthly revenue per member, and a ~58% gross margin, with higher traffic before the school year and standardized tests, softer summers, and sensitivity to district funding cycles [8].

Seasonality and cyclicality. Demand pulses with the academic calendar and exam windows (spring SAT/ACT, July/February bar exams, back-to-school). First-quarter payroll is about a quarter of the annual total [1], i.e., fairly even but exam-season weighted. The consumer segment is discretionary and recession-sensitive; professional-exam and institutional segments are steadier.

6. What drives demand

• Standardized-test policy. The test-optional era is reversing. Dartmouth, then the rest of the Ivy League, MIT, and major public flagships (the Florida and Georgia systems, LSU, Auburn, Alabama) are reinstating SAT/ACT requirements for 2027 admissions [19] — a direct tailwind for admissions-test prep.
• Digital-assessment change. The digital SAT launched nationally in March 2024 as a shorter, adaptive, software-based test [26]; the ACT began phasing in a redesign from April 2025 (a composite of English, math, and reading, with science and writing optional in applicable formats) [27]. Format changes reset practice materials and lift demand for updated prep.
• Learning loss and high-dose tutoring. The 2024 National Assessment of Educational Progress (NAEP) reported lower national average reading scores in grades 4 and 8 than in 2022 [25], sustaining demand for academic recovery. The pandemic learning-loss boom of 2023-24 was largely funded by federal Elementary and Secondary School Emergency Relief (ESSER) dollars, which expired September 30, 2024 [20] — a demand cliff for the institutional segment. States are partly backfilling: Florida committed $20 million, Louisiana ~$30 million (with a bill to reach $45 million), and New York $10 million, with roughly 23 states now offering some tutoring funding [21].
• School-choice vouchers. As of 2025, 18 states run active education savings account (ESA) programs (21 programs total), and in nearly all of them approved tutoring is an eligible expense [22] — a new, growing consumer-subsidy stream, gated by vendor-approval rules.
• Credentialing exams. Bar, CPA, USMLE (medical), NCLEX (nursing), securities licenses, and the CFA generate steady, relatively cycle-resistant demand tied to the number of people entering regulated professions.
• Demographics, income, and AI. Competitive admissions and enrichment-oriented family spending sustain the consumer segment, rising and falling with household income; and AI increasingly improves personalization and tutor productivity — a demand shaper as much as a threat. The main demand risk is that test-optional admissions, free official practice, and AI reduce the need for paid prep in low-differentiation segments; high-stakes exams and human tutoring for stubborn learning problems should be more resilient.

7. Regulation

NAICS is a statistical classification, not a license, and tutoring is lightly regulated at the federal level — there is no national license to be a tutor. Requirements are a state-by-state patchwork [23]:

• Individual freelance tutors face minimal burden — usually just local business registration.
• Some states (California, Florida, New York) apply private-school statutes to tutoring centers that keep attendance and run structured curricula, adding recordkeeping and disclosure duties [23].
• Background checks for tutors working with minors are required in varying forms by state; there is no central federal database [23].

Several federal rules still bite:

• The Federal Trade Commission (FTC) requires advertising claims to be truthful and substantiated; guaranteed-score, grade, or admissions claims are a material liability [28].
• The Children's Online Privacy Protection Act (COPPA) governs online services directed to children under 13, requiring privacy notices and verifiable parental consent [29].
• The Family Educational Rights and Privacy Act (FERPA) governs student records held by federally funded schools; vendors relying on the "school official" exception must stay under the school's control and use data only for the contracted purpose [30].
• The FTC Franchise Rule requires a franchisor to give prospects a Franchise Disclosure Document (FDD) of 23 specified items at least 14 days before signing or payment [31] — directly relevant to anyone buying a tutoring franchise.

Two indirect layers matter as much as direct licensing. First, professional-exam prep is shaped by the credentialing bodies — state bar examiners, the AICPA/NASBA for the CPA, the USMLE program, FINRA for securities — whose exam formats and pass standards dictate the product. Second, public-money tutoring carries compliance strings: ESA voucher dollars require state vendor approval (with tightening fraud scrutiny) [22], and district contracts bring procurement rules, evidence-of-effectiveness standards, and FERPA data-privacy obligations.

The cautionary regulatory tale is China 2021: Beijing effectively banned for-profit academic tutoring for compulsory-education grades overnight, forcing New Oriental, TAL, and Gaotu to convert or exit and wiping out most of their share value [24]. It is a reminder of how policy-exposed this industry can be, even if U.S. politics point the other way today.

8. Competitive dynamics and consolidation

The industry is structurally fragmented (HHI 186.6; top-four share 22.1%) [2] with low barriers to entry — anyone can hang out a shingle, and free alternatives (Khan Academy, and the College Board's free Khan-partnered SAT prep) cap pricing at the low end. Competition runs on brand, measured outcomes, tutor supply and matching, district relationships, retention, and technology rather than scale monopoly. A national brand does not automatically create local pricing power; local execution stays decisive.

Consolidation is happening at the top even as the tail stays fragmented. Private equity has rolled up the scaled brands (Roark/Mathnasium, Unleashed Brands/Sylvan) [14][15][16], Kaplan is the scale leader in professional and test prep, and Nerdy is the venture-backed online consolidator (public via a 2021 SPAC). The defining competitive force now is AI: free tools like ChatGPT and AI search results gutted the low-end homework-help model — Chegg's revenue fell ~39% in a single year [11] — while the same technology lets tutoring providers deliver cheaper, always-on help at scale (AI companions, adaptive practice). AI is simultaneously the biggest threat and the biggest efficiency lever. Expect consolidation to keep coming through franchise aggregation, PE roll-ups, and platform acquisitions rather than one firm capturing the national market.

9. Risks

• AI substitution. Free or near-free AI can commoditize homework help and compress pricing; Chegg is the cautionary case [11]. Premium human tutoring, measured outcomes, and institutional contracts are more defensible than answer-lookup products.
• Public-funding cliffs. The ESSER expiry shows how quickly business-to-government (B2G) demand can drop when episodic federal money ends [20]; state backfill is partial and uneven [21].
• Policy reversibility. Today's tailwind (the test-required reversal) could swing back; any renewed de-emphasis of standardized testing would cut admissions-prep demand — and the China precedent shows policy can be existential [19][24].
• Low barriers / free competition. Easy entry and free official practice structurally cap pricing power for undifferentiated players.
• Tutor labor. Recruiting and retaining qualified tutors, and worker-classification (employee vs. contractor) exposure, pressure the core cost line.
• Consumer protection. Unsupported score, grade, or admissions claims can trigger refunds, FTC enforcement, and litigation [28].
• Child safety and data privacy. Online services handling minors' data face heightened COPPA/FERPA, cybersecurity, and reputational exposure [29][30].
• Discretionary/cyclical demand. Consumer tutoring spending falls with household budgets in a downturn.
• Private-company opacity. Private operators rarely disclose audited segment data, customer concentration, retention, or unit-level profitability.

10. How to invest, and the outlook

Public routes (limited). Nerdy (NRDY) is the only U.S.-listed pure-play — a micro-cap, high-volatility bet on the online-tutoring and school-contract thesis [8][9]. Graham Holdings (GHC) offers indirect, diluted Kaplan exposure inside a conglomerate [10]. Chegg (CHGG) is a distressed AI-disruption turnaround, not classic tutoring [11]. Pearson (PSON) is a broad global assessment/education play [12], and Duolingo (DUOL) is the adjacent language-tech name [13]. The China ADRs (EDU/TAL/GOTU) are the largest listed operators but carry China-specific regulatory and geopolitical risk [24]. There is no clean, large-cap, U.S. pure-play index-style exposure here — treat NAICS 611691 as a screening starting point, then read segment disclosures and separate live human tutoring from assessment delivery, language learning, publishing, and software. Focus on learner retention, tutor utilization, gross margin after tutor cost, school-contract renewals, seasonality, refunds, deferred revenue, and balance-sheet risk. Tickers, prices, dividends, and multiples belong in that company-specific work.

Private routes (where most capital goes). This is fundamentally a private-market industry. The classic small-business entry is owning and operating a franchise center (Mathnasium, Sylvan, Huntington, Kumon) — you pay a franchise fee and ongoing royalties (review the FDD [31]) and earn on local operating economics. Institutional investors access scale through private equity (Roark, Seidler/Unleashed Brands, and other sponsors) that owns the branded platforms, and venture capital funds the wave of AI-tutoring startups. For any center or roll-up, demand center-level revenue, paid instructional hours, tutor utilization, lead sources, renewal rates, prepaid balances, refunds, rent, royalties, teacher turnover, and student outcomes — and test whether central technology and marketing produce measurable gains before paying for projected synergies.

Outlook (forward-looking judgment). The underlying demand — competitive admissions, persistent learning gaps, and a steady flow of licensure candidates — is durable, and the reinstatement of college-admissions testing plus expanding ESA voucher eligibility are genuine near-term tailwinds for consumer and test-prep providers [19][22]. The offsetting headwinds are the post-ESSER funding reset for school-contract tutoring [20][21] and the AI wildcard, which will keep hollowing out undifferentiated homework-help while rewarding providers that pair human tutors with technology and prove outcomes [11]. Expect continued consolidation at the top of a still-fragmented field, with the most defensible economics in high-margin professional-exam publishing and outcomes-backed institutional tutoring. The broad category should keep growing at a low-single-digit pace off a roughly low-double-digit-billions base [6], but returns will concentrate in operators that turn AI from a threat into a cost advantage and win on execution, trusted outcomes, and disciplined customer acquisition rather than on broad industry growth alone.

---

Sources

1. U.S. Census Bureau, County Business Patterns 2023 (NAICS 611691: establishments, employment, annual and Q1 payroll), 2025. https://www.census.gov/data/datasets/2023/econ/cbp/2023-cbp.html
2. U.S. Census Bureau, 2022 Economic Census — Concentration of Largest Firms (NAICS 611691: firms, receipts, CR4/8/20/50, HHI), 2025. https://data.census.gov/table/ECNSIZE2022.EC2200SIZECONCEN
3. U.S. Small Business Administration, Table of Size Standards (NAICS 611691), 2023. https://www.sba.gov/document/support-table-size-standards
4. U.S. Census Bureau, Economic Census — Understanding NAICS / Industry Classification (employer-only coverage; adjacent codes), 2022. https://www.census.gov/programs-surveys/economic-census/year/2022/guidance/understanding-naics.html
5. U.S. Census Bureau, Nonemployer Statistics (captures no-employee sole proprietors; no 611691 total in our file), 2025. https://www.census.gov/newsroom/press-releases/2025/2023-nonemployer-statistics.html
6. IBISWorld, Tutoring & Driving Schools in the US — Industry Report (bundled category, directional only), 2025-2026. https://www.ibisworld.com/united-states/industry/tutoring-driving-schools/1544/
7. IBISWorld, Online Tutoring Services in the US — Industry Report, 2025. https://www.ibisworld.com/united-states/industry/online-tutoring-services/6038/
8. Nerdy, Inc., Form 10-K for FY2025 (active members ~33.2k, active experts ~15.8k, ARPAM ~$364, ~58% gross margin, seasonality), 2026. https://www.sec.gov/Archives/edgar/data/1819404/000181940426000015/nrdy-20251231.htm
9. StockAnalysis, Nerdy Inc. (NRDY) — market cap and price, 2026. https://stockanalysis.com/stocks/nrdy/
10. Graham Holdings Company, 2025 and Fourth-Quarter Earnings (Kaplan and Supplemental Education revenue), Feb. 2026. https://www.businesswire.com/news/home/20260224509125/en/Graham-Holdings-Company-Reports-2025-and-Fourth-Quarter-Earnings
11. Chegg, Inc., Form 10-K FY2025 (Academic Services revenue decline; generative-AI risk); Forbes, "Chegg Stock Down 99%," Oct. 2025. https://www.sec.gov/Archives/edgar/data/1364954/000136495426000021/chgg-20251231.htm
12. Pearson plc, Annual Report and Accounts 2025 (Assessment & Qualifications), 2026. https://plc.pearson.com/sites/pearson-corp/files/annual-reports/2025/pearson-annual-report-2025.pdf
13. Duolingo, Inc., Form 10-K for FY2025 (language platform; Duolingo English Test), 2026. https://www.sec.gov/Archives/edgar/data/1562088/000162828026012494/duol-20251231.htm
14. Mathnasium / BusinessWire, "Roark Capital Acquires Mathnasium" (1,000+ centers), Nov. 2021. https://www.businesswire.com/news/home/20211116006328/en/Roark-Capital-Acquires-Mathnasium
15. Unleashed Brands, "Unleashed Brands Announces Acquisition of Sylvan Learning," 2024. https://www.prnewswire.com/news-releases/unleashed-brands-announces-acquisition-of-sylvan-learning-302064146.html
16. Seidler Equity Partners, "Unleashed Brands" (portfolio backing), 2026. https://sepfunds.com/company/unleashed-brands/
17. Street Fight, "BOOM: Tutoring & Learning Centers" (Kumon ~25,000 centers), Jan. 2024; Kumon North America corporate materials. https://streetfightmag.com/2024/01/04/boom-tutoring-learning-centers/
18. Wikipedia, "The Princeton Review" (Primavera Capital history); Tutor.com, "New Ownership. Same Commitment" (Sept. 2025 U.S.-ownership statement). https://www.tutor.com/cmspublicfiles/WWW/tutorcom-CEO-message-on-US-ownership.pdf
19. College Transitions, "Top Colleges Rolling Back Test-Optional Policies," 2025. https://www.collegetransitions.com/blog/top-colleges-rolling-back-test-optional-policies/
20. K-12 Dive, "Need for high-dosage tutoring remains after ESSER funds end," 2024. https://www.k12dive.com/news/need-for-high-dosage-tutoring-remains-esser-funds-end/746751/
21. EdWeek Market Brief, "From ESSER Boom to Market Reset: States Bolster High-Dosage Tutoring's Next Phase," May 2026. https://marketbrief.edweek.org/education-market/from-esser-boom-to-market-reset-states-bolster-high-dosage-tutorings-next-phase/2026/05
22. National Conference of State Legislatures, "Education Choice State Policy Scan: Education Savings Accounts," 2025; Savvy Learning, "ESA Documentation Requirements for Tutoring," 2025. https://www.ncsl.org/education/education-choice-state-policy-scan-education-savings-accounts
23. National Tutoring Authority, "State-by-State Tutoring Service Regulations and Requirements," 2025. https://nationaltutoringauthority.com/state-by-state-tutoring-regulations/
24. Yicai Global, "New Oriental, TAL Education, Other E-Tutors Crash as China Bans Extra-Curricular Classes," July 2021. https://www.yicaiglobal.com/news/new-oriental-tal-education-other-e-tutors-crash-as-china-bans-extra-curricular-classes
25. National Center for Education Statistics, NAEP Reading 2024 (grades 4 and 8 below 2022), 2025. https://nces.ed.gov/nationsreportcard/reading/default.aspx
26. College Board, "Digital SAT Launches Across the Country," 2024. https://newsroom.collegeboard.org/digital-sat-launches-across-country-completing-transition-digital-and-providing-simpler-testing
27. ACT, "ACT Test Enhancements for Higher Education," 2025. https://www.act.org/content/act/en/products-and-services/the-act-postsecondary-professionals/resources/enhancements-higher-ed.html
28. Federal Trade Commission, "Advertising and Marketing Basics," 2026. https://www.ftc.gov/business-guidance/advertising-marketing/advertising-marketing-basics
29. Federal Trade Commission, "Children's Online Privacy Protection Rule (COPPA)," 2026. https://www.ftc.gov/business-guidance/resources/childrens-online-privacy-protection-rule-not-just-kids-sites
30. U.S. Department of Education, "Who Is a School Official Under FERPA?," 2026. https://studentprivacy.ed.gov/faq/who-school-official-under-ferpa
31. Federal Trade Commission, "A Consumer's Guide to Buying a Franchise" (Franchise Rule; FDD), 2026. https://www.ftc.gov/business-guidance/resources/consumers-guide-buying-franchise