Online Math Tutoring
The shortcut: Pick one narrow grade band or one curriculum (Singapore Math, Saxon, IXL-aligned) and own it. Generalists on Wyzant fight over $35/hour scraps. Specialists charge $80-$120/hour and keep parents for years.
Industry: Childcare & Education | Investment level: Micro — $300-$1,000 | Time to launch: 2-4 weeks (laptop, Zoom, an LLC, and your first 3 students gate the launch)
Best for: A current or former teacher, a strong-in-math undergrad, an engineer who likes explaining things, or anyone who actually moved a kid's math grade and got told "you should do this for money." You're a fit if you can hold a 50-minute session with a fidgety 11-year-old, write a 6-week plan around a specific gap (fractions, integers, word problems), and answer a parent text within a day. What you'll likely make: $600-$1,500 month 3, $1,800-$3,200 month 6, $3,200-$6,000 month 12. Math is in Section 4.
Market Opportunity
The parent calling you usually isn't shopping for a tutor — she's reacting to a report card. The kid pulled a D on a fractions unit, the next test is in three weeks, and Mathnasium quoted $399/month for sessions in a strip mall 25 minutes away. You're the alternative: someone who shows up on a screen at 7pm Tuesday, works the actual gap, and doesn't cost a car payment.
That panic buy is the entry point. The bigger market: most parents who hire a tutor for a one-month "fix" keep them for the rest of the school year once a grade moves. The global online tutoring market sits in the $10-$12 billion range and is growing about 15% per year through 2030. What that means for you on the ground: a kid who books two sessions a week at $70 is worth roughly $5,000 over a single school year.
The trap is the marketplace tax. Wyzant takes 25% of every dollar you earn until lifetime billings cross $2,000, then drops to 20%. Outschool keeps about 30% of group-class revenue. Coaches who never move clients off the marketplace stay permanently capped at marketplace margins. Your first Wyzant client should also get an email and a nudge to book directly next time.
Launch With AI
Pro section. AI doesn't replace the work — it cuts the parts that drained you (writing a different lesson plan for every student, generating practice problems at the right difficulty, drafting weekly parent reports, recording session recaps). Spend the saved time on what AI can't do: sitting with a fidgety 11-year-old at 7pm, reading the moment they're about to give up on fractions, and the warm parent text after the breakthrough session.
The trap most new tutors fall into: they think AI tutoring tools (Khanmigo, Magic School AI) will threaten their business. Wrong direction — those tools are noise compared to a human who shows up every Tuesday. The actual leverage is using AI as your prep assistant: the kid still needs you on Zoom, but you stop spending 90 minutes of unbilled prep on every 50-minute lesson. The tutors clearing $4-6K/month aren't grinding harder, they're spending 15 minutes on AI-assisted prep instead of 90 minutes per session.
Important up-front: AI cannot teach a kid math. It also cannot write parent reports about a session it didn't see, or replace the verbal "I see what you did there" moment that builds trust. Most importantly, AI tools that interact with under-13 students directly trigger COPPA exposure (Section 5) — your AI use should stay tutor-side, not student-side, unless you have written verifiable parental consent and a documented data-handling plan. You set the structure; AI prepares the materials and writes the parent updates.
AI Tools You'll Use
| Tool |
Price |
What it does |
| ChatGPT Plus |
$20/mo |
Diagnostics, lesson plans, practice problems, weekly parent reports |
| Wolfram Alpha Pro |
$7.25/mo |
Math problem solving, step-by-step solutions, equation verification before you teach it |
| Loom Free |
$0 |
3-min weekly recap videos parents can watch — answers "what did you cover this week?" |
| Canva Pro (Magic Studio) |
$15/mo |
Custom worksheets, parent-facing progress reports, mom-group / Nextdoor carousels |
| Otter.ai (free tier) |
$0 |
Auto-records sessions for parent review (with verifiable parental consent only) |
The Workflow
Run a 10-question diagnostic on a new student (ChatGPT, ~20 min per intake). Most new tutors waste session 1 on small talk and a vague "where are you struggling?" Instead, send a diagnostic before session 1. Paste:
"I'm tutoring a [grade] student whose parent says they're struggling with [topic — fractions, algebra 1, integers, word problems]. The kid is on the [Singapore Math / Saxon / Common Core / IXL] curriculum. Build me a 10-question diagnostic that escalates in difficulty — 2 below grade level, 4 at grade level, 4 one grade above. For each question include: the answer, the most common wrong answer (so I know what their thinking is when they get it wrong), and the underlying skill being tested. Format: clean PDF-ready, no extras."
Drop into Canva to format, send via Acuity intake. The kid does it in 15 minutes before session 1. You arrive at session 1 already knowing exactly which 3 skills are the gap. Parent already knows you take this seriously. That's the difference between a $35/hr Wyzant tutor and a $90/hr specialist.
Build the 6-week lesson plan from the diagnostic (ChatGPT, ~30 min once per student). Per the plan body, the parent buys "get fractions to a B by midterm." That's a 6-week sprint with a measurable goal. Paste the diagnostic results:
"This [grade] student scored [paste results] on the diagnostic. The parent's stated goal: [paste — e.g., 'get fraction unit grade to a B by midterm in 6 weeks']. Build me a 6-week lesson plan, one 50-min session per week. For each session: (1) the specific skill we're targeting, (2) the warm-up problem (review of last week), (3) the 'I do, we do, you do' instruction segment, (4) 5 practice problems with answers, (5) the homework problem with answer, (6) the parent-facing 1-line goal for the week. Tone: structured but warm — this is for a real kid who's been frustrated."
Save as the student's plan in Notion or Google Docs. Reuse the structure for every new student in that grade × curriculum × goal combo. By your 5th student in the same lane, you have a battle-tested template that takes 5 minutes to customize.
Generate practice problems at the right difficulty (Wolfram Alpha + ChatGPT, ~10 min/week). The hardest part of writing your own practice problems is making sure they're solvable, not too easy, and not too hard. Wolfram Alpha is the math-specific verifier:
"Generate 10 practice problems for a [grade] student working on [specific skill — e.g., 'multiplying fractions with mixed numbers']. Difficulty progression: 3 entry-level, 4 mid-difficulty, 3 challenge problems. Include answers. Then for each problem, paste the equation into Wolfram Alpha to verify the answer is correct and the solution path makes sense for a [grade]er — flag any problem where the path requires above-grade techniques."
You verify the Wolfram step-through before teaching — catches the AI hallucinations that would otherwise embarrass you in front of a parent's kid. Wolfram's step-by-step solutions also become your teaching script for the harder ones.
Weekly parent reports — the retention engine (ChatGPT + Loom, ~15 min/student/week). Parents pay for tutoring but don't see the session. The retention question is always "is this working?" — and "their grade went up" lags by 4-6 weeks. Bridge the gap with a weekly report. After each session, paste your 5-line note:
"This is my session note for [student name, grade] this week: [paste — e.g., 'Reviewed last week's homework, 7/10 correct. Today: introduced multiplying fractions with mixed numbers. Got it on her own by problem 4. Next week: division of fractions. Homework: 5 practice problems p.42.']. Convert into a 150-word parent email: warm but specific, mentions the win this week, the one thing we worked through together, the goal for next week, and a 1-line homework reminder. Tone: confident tutor, not corporate. Subject line under 45 characters."
For high-touch parents, also record a 3-minute Loom: open Loom, share screen with the worksheet you used, narrate "here's what we covered today, here's where she struggled, here's the practice for next week." Loom Free covers 25 videos/month — plenty for a full book of weekly clients. The 3-minute Loom is the difference between a parent who tells 2 friends and a parent who silently doesn't rebook.
Mom-group + Nextdoor marketing batch (Canva + ChatGPT, ~90 min/month). Per Section 6, your highest-yield channels are Facebook mom groups and Nextdoor — you should be posting once a month into 3-5 groups. Paste:
"I tutor [grade band] math online. My specialty: [Singapore Math / Algebra 1 + 2 Common Core / SAT Math]. I'm IXL-certified and have [N] testimonials. Open slots: [N]. Build me: (1) one 200-word Facebook mom-group introduction post (conversational, mentions a specific result like 'last 4 students moved from C to B in 6 weeks'), (2) a 5-slide Nextdoor carousel for parents — slide 1 hook, slides 2-3 the diagnostic-first approach, slide 4 a sample weekly report (anonymized), slide 5 a CTA to book a free 15-min consultation, (3) a 3-line text I can send to 30 contacts who have school-age kids around back-to-school + report-card season."
Build the carousel in Canva (Magic Studio's brand kit). Schedule the FB posts via Meta Business Suite to fire on the high-spike weeks (last week of August, mid-November after first-quarter report cards, week after winter break). One mom-group post in a 3,000-member group routinely produces 5-15 inquiries per the plan body.
Time Saved Per Week
Roughly 6-8 hours/week once your diagnostic + lesson-plan template + parent-report flow are dialed in (assuming 10 weekly students):
- Lesson planning per student: 90 min/week → 15 min (template + ChatGPT customizes)
- Practice problem writing + verification: 60 min/week → 15 min (Wolfram verifies)
- Weekly parent reports for 10 students: 2-3 hours/week → 30 min (ChatGPT drafts from your notes)
- Marketing posts for mom groups + Nextdoor: 90 min/month → 30 min (templates ready)
- Diagnostic for new student intake: 60 min per intake → 15 min (template reused)
Trade that time for: a 5th and 6th student per week (each adds $560-$800/month at $70-$100/hour, 2 sessions/week), filming one Outschool group class per month for lead-gen, and personally calling any parent whose kid skipped a session. Retention is everything in tutoring — a kid who stays the full school year is worth $4,000-$5,000 of LTV.
Total AI Stack Cost
- Budget tier ($20/mo): ChatGPT Plus only. Skip Wolfram Alpha Pro (free Wolfram works for verification at lower volume), use Loom Free instead of paid, Canva Free instead of Pro, Otter Free under 600 minutes/month. Right for the first 3-5 students while you're stabilizing.
- Full tier ($42/mo): ChatGPT Plus + Wolfram Alpha Pro + Canva Pro + Loom Free + Otter Free. Worth it once you cross 8 weekly students and prep + parent-report volume becomes the bottleneck.
- Compare: A part-time virtual assistant doing your parent reports and lesson planning runs $400-$1,000/month. A freelance worksheet designer is $30-$80/worksheet. The full AI stack is $42/month — and you keep instructional control, which matters because the parent is paying for your judgment, not an outsourced template.
Cancel anything you don't open in a 7-day window. Don't stack 3 different AI student-facing tools — your AI use stays tutor-side until you've worked through the COPPA paperwork (Section 5) for any tool that interacts with the under-13 student directly.
Your First Win
30 minutes from now you'll have a diagnostic ready for your first prospective student. Open ChatGPT (free tier works for this one). Paste:
"I tutor [grade band, e.g., 5th-8th grade] math online. My specialty: [Singapore Math / Algebra 1 + 2 / Common Core / SAT Math prep]. Build me a 10-question math diagnostic for a [grade] student whose parent says the kid is struggling with [topic — fractions, integers, word problems, algebra, etc.]. Question difficulty: 2 below grade, 4 at grade, 4 one grade above. For each question include: (1) the question itself, (2) the correct answer, (3) the single most common wrong answer and what it tells me about how the kid is thinking, (4) the specific skill being tested. Format as a clean PDF-ready worksheet I can drop into Canva. Add a 1-line note at the top: 'Please complete in 15 minutes without help, no calculator.'"
Drop the result into Canva, format it, save it as your "Intake Diagnostic v1" template. Send it to your next inquiring parent before booking session 1. You've just changed the conversation from "how much do you charge?" to "your tutor took our kid seriously enough to assess before charging us anything." That's the move that justifies the $80-$120/hour price band.
Product / Service Offering
You're selling one thing in three formats:
- 1-on-1 weekly tutoring (the main product). A 50-minute session, once or twice a week, with a written 4-6 week plan attached to a specific outcome ("get fractions to a B by midterm"). Delivered through Zoom + a shared whiteboard like Bitpaper. $60-$120/hour depending on grade band and your credential.
- Test-prep cohort (the seasonal upsell). SAT Math, ACT Math, ISEE/SSAT, or end-of-year state tests — a defined 8-12 week sprint priced as a package ($800-$2,400) instead of per session. Higher commitment, no flake-out.
- Group class on Outschool (the lead source). A weekly $20-$40/seat group class on a narrow topic ("5th-grade fractions weekly review"). Outschool finds you the students; some convert to private 1-on-1. Treat it as paid lead-gen, not the main income.
Pick one grade band and one curriculum to anchor your listings. "K-12 math tutor" gets ignored. "5th-7th grade Singapore Math specialist" or "Algebra 1 + Algebra 2 Common Core" gets booked.
Revenue Model
Unit economics for a solo tutor working from a laptop, no employees:
| Service |
Price |
Variable cost (platform + payment fees) |
Your time |
Take-home per session |
| 1-on-1 direct (own site, Acuity + Stripe) |
$70/hr |
$2.30 (Stripe 2.9% + $0.30) |
50 min + 10 min prep |
~$67 |
| 1-on-1 specialist (SAT/algebra) |
$100/hr |
$3.20 |
50 min + 15 min prep |
~$96 |
| 1-on-1 via Wyzant (first $2K lifetime) |
$70/hr |
$17.50 (25% Wyzant cut) |
50 min + 10 min prep |
~$52 |
| Group class on Outschool (8 seats × $25) |
$200/class |
~$60 (30% Outschool cut) |
45 min + 30 min prep |
~$140 |
| Test-prep package (10 sessions × $90) |
$900 flat |
$26 |
8.5 hrs of work |
~$870 |
Your first $1K month = 3 regular students at $70/hour doing 5 sessions/week total. Wyzant fills the early slots; your own site takes over by month 3.
Your first $3K month = 10 weekly students × 2 sessions × $70 = ~$1,400/week, plus one Outschool group class running ($300-$500/month) and one specialist client at $100/hour. Total: $3,200-$3,600/month. Or run the specialist play alone: 5 SAT-prep students × 4 sessions/month × $100 = $2,000, plus 5 regulars = ~$3,400/month with fewer clients.
The two seasonal spikes that fund the year: back-to-school (August-September) and post-report-card panic (November and February). Summer "bridge" sessions for kids moving up a grade fill June-July.
Startup Costs
- Laptop, headset, webcam. $400-$700 used if you don't have them.
- Zoom Pro: $15.99/month for unlimited session length. The free tier's 40-minute cap kills 50-minute lessons.
- Bitpaper or equivalent shared whiteboard: free tier works for 1-2 students, paid plans from $10/month at bitpaper.io.
- Acuity Scheduling (booking + payment + Zoom integration): $20/month at acuityscheduling.com. Calendly free tier works as a starter.
- Stripe: 2.9% + $0.30 per transaction, no monthly fee.
- Curriculum backbone: Khan Academy is free at khanacademy.org. IXL ($9.95-$19.95/month at ixl.com) is usually paid by the parent, not you.
- IXL Tutor Certification badge via Wyzant — free, earns a marketplace trust signal. Do it before launch.
- LLC + EIN + insurance: $35-$500 LLC filing depending on state — see the LLC University 50-state table. EIN is free at IRS EIN Online — never pay a third party. General + professional liability for a solo tutor runs $300-$600/year via Hiscox or Insureon.
- A simple one-page website: Carrd ($19/year) or a free Squarespace trial. Takes an evening.
Realistic all-in: $300 if you already have a working laptop and defer the LLC for 60 days; $1,000 if you file the LLC, bind insurance, buy a year of Acuity + Zoom + Bitpaper, and grab a domain.
Legal & Formation
Business entity. Single-member LLC the moment you take money from a parent who isn't your sister. It separates your personal accounts from any complaint a parent files later. Sole prop is fine for the first month while you test, but file the LLC before client number five. Get your EIN free directly at the IRS — services that charge $50-$300 for "EIN filing" are reselling a five-minute free form. S-corp election almost certainly doesn't make sense in year one for this slug — wait until net profit is sustainably above the $80K-$100K range.
Licenses & compliance. Private tutoring is not state-licensed in any U.S. state — that's a real entry advantage. You don't need a teaching credential, a state registration, or a tutoring-specific permit. What you do need to take seriously: mandatory reporter status (every adult in a regular tutoring relationship with a minor is a mandatory reporter in essentially every state — the standard is "reasonable suspicion," confirm your state's specific training requirement at the Child Welfare Information Gateway). And a basic background check ($25-$75 via Checkr or Sterling) is table stakes — a parent will ask, and the answer should be "yes, here's the report." Tutoring services aren't subject to sales tax in most states. FERPA only applies if a school district hires you directly and hands you student records — parent-paid tutoring is exempt.
Industry-specific risk. COPPA is the real exposure. The Children's Online Privacy Protection Rule covers any online service that collects personal info from children under 13. The moment you ask a parent for a 10-year-old's name, school, and grade level to set up a Bitpaper account, you're collecting child data. The fix: get verifiable written parental consent in your enrollment form, do not allow under-13 students to create their own accounts, and document what data you store. A checkbox on a signup page is not "verifiable parental consent" under the FTC standard. The penalty cap is $51,744 per violation as of 2024. Second trap: never promise a specific grade or test score. "I guarantee an A" is an FTC unfair-and-deceptive-practices risk. Phrase it as "targeted practice on the gap, measured weekly."
Marketing & First Customers
Your first 5 students come from people who already trust you. Not Google ads — texts.
- Direct text to 30 parents you know. Coworkers with school-age kids, neighbors, your kid's classmates' parents. Offer a $40 first session (one-time, then $65-$70). You will sign 2-4 in a week.
- Wyzant for early lead flow. List with the IXL Tutor badge and one narrow specialty in your headline. Accept the 25% cut for the first 5-10 clients to build reviews. Move every Wyzant client to direct booking by session 4-6 with a 10% loyalty discount as the bridge.
- Outschool group class as paid lead-gen. Run one weekly $25/seat group class on a narrow topic. Two or three students convert to private 1-on-1 over a semester.
- Local Facebook mom groups and Nextdoor. A single post — "I tutor 5th-8th grade math online, three open slots, $65/hour, IXL-certified" — in a 3,000-member neighborhood group routinely produces 5-15 inquiries. Repeat monthly.
- Push outreach hard around the school-year spikes: the last week of August, the week after first-quarter report cards (mid-November), the week after winter break, and right before spring testing.
Ask every parent for a Google review by session 6. Twenty reviews and you stop having to discount the first session.
First 90 Days
- Week 1. File the LLC. Get the EIN free at IRS. Set up a business checking account. Run a background check on yourself ($25-$75) and save the PDF.
- Week 1-2. Pick your specialty (one grade band + one curriculum). Buy domain, build a one-page Carrd site. Set up Zoom Pro, Bitpaper, Acuity + Stripe.
- Week 2. Complete the IXL Tutor Certification. List on Wyzant with the badge. Bind a $300-$600/year general + professional liability policy via Hiscox or Insureon.
- Week 2-3. Write your enrollment form with COPPA-compliant parental consent language. Write your intake questionnaire (current grade, last test scores, specific gap, parent goal in 6 weeks).
- Week 3. Text 30 warm contacts. Aim for 3 paying students signed by end of week 4.
- Week 4-6. Run sessions. Capture parent testimonials and Google reviews after session 4 of each student. Confirm any state-required mandatory reporter training is on file.
- Week 6-8. Launch one Outschool group class on your specialty. Post in two local Facebook mom groups and Nextdoor.
- Week 8-12. Raise rate for new clients to $80-$100/hour. Existing clients keep their original rate. Move all Wyzant clients to direct booking. Target: 8-12 weekly students by day 90, ~$1,800-$2,500/month.
Common Pitfalls
- Never moving clients off Wyzant. The 25% cut adds up fast. Every client you keep on the marketplace past session 6 costs you $15-$25/session in fees. Build a one-page direct-booking site in week one and move them.
- Promising a specific grade or score. Phrase outcomes as "targeted practice on the gap, measured weekly." A guarantee in writing is an FTC risk and a chargeback magnet.
- Skipping the COPPA paperwork. A checkbox isn't verifiable parental consent. Use a signed enrollment form (Acuity supports digital signatures), do not let under-13 students create their own accounts, and document what data you store. The penalty per violation is in the five figures.
- Generalist headline. "K-12 math tutor" gets buried. "5th-8th grade Singapore Math + IXL specialist" gets booked. Pick one grade band and one curriculum and own it for the first 12 months.
Get your full launch plan — take the free 60-second quiz.