How to Start a Tutoring Business in 2026

Solo tutoring is a job. A tutoring agency is a business. The transition from one to the other is what most tutors never figure out — and it is the single decision that separates a $60K side income from a $400K agency that runs without you in the room.

The good news: tutoring is one of the cheapest service businesses you can start in 2026. A laptop, a Stripe account, a scheduling tool, and a subject you can teach better than the average parent can explain. Demand is structurally higher than it was in 2019 — pandemic learning loss is still showing up in middle-school math scores, AP exam pressure is rising, and the test-prep market did not die when test-optional admissions came in (it shifted to ACT/SAT-required programs and to schools that quietly reinstated requirements).

This guide walks through the model decision (solo vs agency), the subjects that actually pay, what your real startup costs look like at each tier, how to price without leaving money on the table, where students actually come from, how to hire tutors when you scale, and the four mistakes that kill new tutoring businesses in the first 18 months.

There are three models, and they pay very differently.

Solo tutor. You teach every session. You charge $50-150/hour depending on subject and geography. Your ceiling is roughly your billable hours per week — if you teach 25 hours a week at $100, that is $130K/year gross before taxes and self-employment costs, which is a great solo income but it is a job. You stop working, you stop earning. Vacations cost you twice (the lost revenue plus the trip).

Tutoring agency. You stop teaching (or teach a small caseload to stay sharp) and you book sessions, match students to tutors, handle billing, and own the brand. You pay tutors $30-60/hour and bill clients $80-200/hour. The 30-50% margin is your business. Your ceiling is no longer your hours — it is how many tutors you can recruit, vet, and retain. A 12-tutor agency running each tutor at 15 billable hours a week at a $40/hour spread is $375K/year in margin.

Hybrid. Most successful tutoring businesses pass through this stage. You keep 5-10 of your own highest-paying private students (often the ones who started with you) at $120-150/hour, and you build the agency layer around them. This funds the agency build-out without you having to take on debt or investors. Plan to stay in hybrid for 12-24 months.

The trap: most solo tutors talk themselves into staying solo because the agency model feels like a step away from the work they love. It is. That is the point. If teaching is the work you want to do forever, stay solo and price up. If building something that runs without you is the goal, the agency transition is the only path.

Generalist tutors compete with every college student on Wyzant. Specialists charge 2-3x as much and have parents on waitlists.

Standard subjects (lower margin). K-12 math, K-12 reading and writing, general science, foreign-language conversation. These are the easiest to fill but the easiest to undercut. Expect $40-80/hour as a solo tutor.

High-margin specialties (this is where the business actually lives):

- AP and IB tutoring. AP Calculus BC, AP Chemistry, AP Physics C, IB Higher-Level Math. Parents pay $100-200/hour for tutors who have a track record of 4s and 5s. Specialize in 2-3 AP subjects and you can fill a calendar from January through May every year. - Test prep. SAT, ACT, GRE, GMAT, MCAT, LSAT. MCAT and LSAT tutors charge $150-300/hour. The work is repeatable (same content every cycle) and the contracts are large (most students buy 30-60 hour packages). - Special-needs and learning-difference tutoring. Orton-Gillingham for dyslexia, executive-function coaching for ADHD, slow-paced math for dyscalculia. Requires certification (Wilson, Barton, or OG training) but commands $90-180/hour and parents tend to be long-term clients (12-36 months). - College admissions. Essay coaching, Common App support, school selection. $150-400/hour, often sold as fixed packages ($3K-15K). - Computer science and competition math. AMC/AIME prep, USACO coding, AP Computer Science. Smaller market but very high willingness to pay among tech-industry parents.

Pick a specialty before you build your website. "Math tutor" is a category. "AP Calculus and AMC competition prep tutor" is a brand.

Tutoring is one of the least regulated service businesses in the US, but the structure decisions still matter.

LLC formation. Form an LLC in your home state. $50-500 depending on state ($800/year minimum franchise tax in California, which is the main outlier). The LLC separates your personal assets from any client-related liability — important the moment a parent claims their child's grades dropped because of a tutor you assigned.

Tax election. Single-member LLCs file as sole proprietorships by default (Schedule C). Once your net income clears ~$50K, talk to a CPA about an S-corp election to reduce self-employment tax. Most tutoring agencies make this transition between year 2 and 3.

Professional liability insurance. Required if you teach minors. Hiscox, Next Insurance, and Coverage.com all sell solo-tutor policies. $300-600/year covers most claims a parent might bring.

Contractor classification (the big one). This is where most growing tutoring agencies get into trouble. If you control how, when, and where tutors work — they wear your branded shirts, follow your curriculum, take students you assign — they are W2 employees, not 1099 contractors, under both federal IRS rules and most state tests (California's ABC test is the strictest). Misclassification penalties run into five and six figures. Two safer paths: (1) keep tutors as genuine contractors who set their own hours, bring their own materials, and can subcontract; (2) hire tutors as part-time W2 employees from day one and budget for payroll taxes, workers' comp, and unemployment. Most successful agencies eventually move to W2 because retention and quality are dramatically better — but the cost is real, roughly 15-20% more per billable hour.

Three pricing models, and you should probably offer all three.

Hourly. $50-150 for solo tutors, $80-200 for agency-billed sessions. Easiest to sell to first-time buyers. Hardest to retain — every session is a re-decision for the parent. Use hourly for trial sessions and one-off help.

Session packages. 12-session packs are the sweet spot. Discount 10-15% off the hourly rate ($120/hour x 12 = $1,440 sold as a $1,300 package). Packages lock in revenue, reduce no-shows (parents don't want to lose a paid slot), and let you book recurring weekly times. Sell a 20-session pack at the AP/SAT level for the run-up to exams — many parents prefer one bill to twelve.

Monthly subscription. $400-1,200/month for X sessions per week, with rollover rules and a cancellation policy. This is the model that turns tutoring from a transactional business into a recurring-revenue business. The math is the same as a private-school billing cycle, which is exactly why parents accept it. Subscription works best for K-8 ongoing support and for executive-function coaching where the value is consistency, not exam prep.

Price anchoring: list your highest-tier package first on your pricing page. The 30-session AP package at $4,500 makes the 12-session pack at $1,800 look reasonable, even though the 12-session pack is what you actually want most parents to buy.

Four channels, ranked by what actually works for new tutoring businesses.

Private and charter school partnerships. This is the highest-ROI channel that almost nobody pursues seriously. Private schools — especially K-8 schools without their own learning-support staff — quietly maintain a list of 3-5 tutors they refer parents to. Get on that list and you have a referral pipeline that runs without ad spend. The way in: email the head of school, the learning-support coordinator, or the upper-school dean, offer to do a free parent workshop on study skills or test prep, and follow up. Plan on 6-12 months to convert. One school relationship can be worth $50K-150K/year in steady referrals.

Parent referrals. Once you have 5-10 happy clients, ask. Most tutors never ask. Offer a $50-100 referral credit for any parent who sends a new family that signs up. Track it in your CRM so you actually pay it.

Marketplaces (Wyzant, Tutor.com, Varsity Tutors). Useful for getting your first 3-5 students fast, but the take rate is brutal — Wyzant takes 25% of your first session and a declining percentage after, Varsity Tutors keeps the parent relationship entirely. Use marketplaces as a launchpad, then move clients to direct billing after the first or second session (within the platform's terms of service — read them carefully).

Local and online ads. Google Local Services Ads work surprisingly well for "tutor near me" searches. Facebook and Instagram ads work for the parent-of-elementary-age demo, especially in higher-income zip codes. Skip TikTok unless your specialty is something teenagers buy themselves (rare). Budget $300-1,000/month and expect a 60-90 day learning period before the cost-per-acquisition stabilizes.

Hiring is the bottleneck in every agency that gets stuck.

Sourcing. University graduate students in education, math, or the sciences are the best single source — they need flexible work, they want teaching experience, and they will accept $30-50/hour. Post on the career boards of every university within 30 miles. Retired teachers are a smaller but extremely high-quality pool — they often want to work 8-15 hours a week and they bring decades of classroom craft. Wyzant and Care.com are noisier but workable for filling specific gaps.

Vetting. Three-stage process: (1) resume screen for subject expertise and teaching experience, (2) 30-minute video interview where they teach you a 10-minute lesson on a topic they choose, (3) paid trial session with a real student where you observe (with parent consent). Most failed hires fail in stage 2 — they know the material but can't break it down for a 14-year-old. The 10-minute teach-back filters them out before you waste a session on a paying client.

Background checks. Non-negotiable for anyone working with minors. Use Checkr or Sterling. $30-50 per check. Run it before the trial session.

Ongoing training. A 30-minute monthly call where you walk through one tricky student case, one curriculum update, and one billing-or-process item. Keeps tutors aligned and gives you a feedback channel.

Retention pay structure. Pay base hourly + a small per-student retention bonus (e.g., $25 every time a student renews a 12-session package). Tutors who feel ownership of student outcomes stay longer and quality compounds. Plan to lose 30-40% of tutors in their first six months no matter what — university students graduate, schedules change, and that is the nature of a flexible-work model.

Under-pricing on day one. Charging $40/hour because you feel awkward asking for $90. The problem is not just the lost revenue — it is that you anchor your future clients to a price you cannot scale. The tutor charging $90 from session one will be at $130 within two years. The tutor who started at $40 will still be at $55 because raising prices on existing clients is hard. Price for the tutor you intend to be in 18 months, not the tutor you are today.

No written agreements. A parent disputes a credit-card charge for a session their kid attended. You have no contract, no signed parent agreement, no policy on cancellations or no-shows. Stripe sides with the parent. You lose $1,200 in chargebacks in your first year. Use a real services agreement signed at intake, with cancellation terms (48 hours), no-show policy (full charge), and a clear refund policy. Run it through ESign so it is timestamped and binding.

One-school dependency. You build a relationship with one private school and 70% of your revenue comes from it. The head of school changes. The learning-support coordinator retires. Your pipeline goes to zero in 90 days. Diversify across at least three referral sources by year two — multiple schools, parent referrals, and at least one paid acquisition channel.

No tutor backup pipeline. A parent's 4 p.m. Tuesday session is sacred. The tutor gets sick. You have no one to cover. The parent finds a different tutor who can show up reliably. You lose a $4,800/year client. Build a bench of 2-3 backup tutors per subject — even if they only work 2-4 hours a month — and make coverage a contractual expectation.