How to Start an Online Tutoring or Test-Prep Business

Most tutoring businesses do not start as businesses. They start as a favor: a neighbor's daughter who is stuck on fractions, a colleague's son with the SAT in March. Then a second family asks, then a fourth, and one evening you notice you are running something with customers, a calendar and unpaid invoices, but no prices you chose on purpose and no record of who attended what.

This guide is about making those choices deliberately. To start an online tutoring business you need eight decisions, roughly in this order: a niche, a format, a price and a package, a way to find the first ten students, a schedule you can keep, a way to show progress, a way to get paid, and a rule for when to bring in help. None of them needs special software on day one. All of them are easier to change in month one than in month twelve.

A note on numbers before we begin. Every dollar figure below is an illustration with the arithmetic shown, so you can replace my inputs with yours. I have not quoted an average for what tutors earn, because an average across every subject and every country says nothing about your subject, your students or your hours.

A parent does not search for 'a tutor'. They search for help with the specific thing that is going wrong: Algebra 2, the SAT math section, AP Chemistry, conversational Spanish before a move abroad, piano for a nine-year-old. A niche is simply the sentence a happy parent uses when they recommend you. If that sentence is 'she tutors', the recommendation goes nowhere. If it is 'she got my son through AP Calculus', it travels.

Three questions settle the choice. Can you name the result in one line? Is there a date or a milestone that makes a family act now rather than someday? And have you already produced that result for someone, even informally? Two answers of yes is enough to start. A niche is not a cage: Lucas Martin can begin with SAT math, add ACT math after his first cycle, and add AP Calculus when the same families ask for it.

Narrow feels risky because it seems to shrink the pool. Online it does the opposite. You are no longer limited to families within a drive of your house, so the pool for 'AP Chemistry tutor' is every student taking AP Chemistry in a time zone you are willing to teach in. The narrower the promise, the easier it is for the right family to recognize that you are talking to them.

Format decides your income ceiling more than your hourly rate does, because it decides how many students one hour of your attention serves. There are three to choose from, and most tutoring businesses end up running two of them side by side.

• One-to-one. One student, fully adaptive. It is the easiest format to sell and the easiest to start, and it is capped by the hours in your week. Every cancellation is an hour of income gone.
• Small group (three to six students). Same level, same goal, one fixed weekly slot. Each family pays less than a private rate, you earn more per hour, and students learn from each other's mistakes. It only works when the students really are at the same level, which is the hard part.
• Cohort (a group that starts and finishes together). A program with a first day, a last day and a syllabus: eight weeks to the SAT, a semester of AP Biology. It suits test prep especially well, because the date is the same for everyone in the room.

Take three illustrative prices: $75 for a private hour, $30 a seat in a group of four, and $20 a seat in a cohort of twelve. The private hour earns $75. The group hour earns 4 × $30 = $120. The cohort hour earns 12 × $20 = $240. The family's price fell each time and your hour became more valuable each time. That is the whole economic argument for teaching groups.

The arithmetic hides the costs. A group hour needs more preparation than a private one, a cohort needs material that works without you adjusting it on the fly, and neither fills itself. Start one-to-one, because that is where you learn what students actually get stuck on, and open a group only when three or four of your private students are working on the same thing at the same level.

Start from what you need, not from what the tutor down the road charges. You do not know her costs, her hours, or whether she is making a living at that rate.

Work out the rate you need

Sarah Mitchell wants tutoring to bring in $4,000 a month before tax. She can teach 15 hours a week. Over a year that is 15 × 52 = 780 hours, or 780 ÷ 12 = 65 hours a month. Not every slot gets filled: holidays, illness and cancellations happen, so she plans on 80% of them, which is 65 × 0.8 = 52 billable hours. $4,000 ÷ 52 = $76.92. Her private rate needs to be about $77 an hour. If that number looks high for her subject, the answer is not to hope for more hours. It is to change the format, which is exactly what the group and cohort options are for.

Three ways to package the same teaching

Notice that the monthly price is lower per session than the single-session price: $70 against $75. That gap is what you pay the family for commitment, and it is worth paying. A student on a monthly slot is a student you do not have to sell to again every week.

The program is where the hourly math changes. Six students at $640 is 6 × $640 = $3,840. Sixteen 90-minute sessions is 24 hours of teaching. Add 8 hours for grading practice tests and writing reports and you have 32 hours of work. $3,840 ÷ 32 = $120 an hour for you, while each family pays $640 ÷ 16 = $40 a session, well under the private rate.

A $640 program is a real sum to find at once. Splitting it into four payments of $160 keeps the price the same and removes the objection. What matters is that the later payments are collected automatically. A payment plan you have to chase is just late invoices with a nicer name.

Whatever you choose, write down three policies before the first paid lesson: how much notice a cancellation needs, whether a missed session is made up or forfeited, and when a refund is given. Put them on the page where people pay. A policy invented in the middle of a disagreement satisfies nobody.

Ten is the number that matters at the start. It is enough to prove the offer, fill a small group, and produce the two or three results you will talk about for the next year. Those ten do not need advertising. They come from people who already trust you, and from the people those people trust.

1. Write the one-line offer. 'I help juniors raise their SAT math score in eight weeks, online, in groups of six.' If you cannot say it in a line, a parent cannot repeat it to another parent.
2. Tell everyone who already trusts you. Former students, their parents, colleagues, neighbors, families from a school you taught at. One at a time, by name, with the one line and a link. Then ask a second question: 'Do you know a family who needs this?'
3. Offer a free diagnostic, not a free lesson. Twenty minutes, a short quiz, and a spoken summary of the two or three things holding the student back. A free lesson gives away the product. A diagnostic shows the parent a problem they can now see, and you are the person who found it.
4. Go where the parents already are. School parent groups, neighborhood forums, the library notice board, homeschool co-ops. Read each group's rules, and answer questions usefully for a while before you mention what you do.
5. Ask one professional for referrals. A school counselor, a music-shop owner, a classroom teacher who does not tutor. They are asked for recommendations all the time and are glad to have a name to give.
6. Publish one useful thing a week. A worked problem, a common mistake explained, a study plan for the month before a test. Put it on your own website, so it keeps working after the week is over.
7. Ask every happy family for two things at the first good result: a sentence you may quote with their permission, and one introduction.

Keep a simple list of who you told and what they said. The first ten are a counting exercise more than a marketing one: if you have personally told sixty people and run fifteen diagnostics, you will usually know whether the offer is working, and what parents say when it is not. For the longer version, when ten needs to become a hundred, see how to get your first 100 students.

Scheduling is where a tutoring business quietly loses money, and the fix is unglamorous: fixed slots. A student who has Tuesdays at 5 p.m. for the semester has a habit to keep. A student who books week by week has a decision to make every week, and every reschedule is ten minutes of messages you are not paid for.

• Offer a small number of fixed weekly slots, and state each one in the student's time zone as well as yours. Daylight-saving changes do not happen on the same date everywhere, so confirm the slot in writing twice a year.
• Send a reminder before every session. Automatic is better than remembered.
• Use one join link that never changes. 'Which link is it today?' is the most avoidable message in tutoring.
• Leave a gap between sessions. Back-to-back hours with no margin mean every overrun makes the next family wait.
• Record attendance as it happens, not from memory at the end of the month.

Attendance is not bureaucracy. It is the record that settles three arguments before they start: whether a missed session was the student's or yours, whether a make-up is owed, and whether the slow progress a parent is worried about has something to do with the four sessions their child skipped. In a group it also shows you who is drifting while there is still time to do something about it.

On Prolaud a live class can repeat weekly on one join link, enrolled students get a reminder before it starts, and attendance is recorded from who was actually in the room and for how long, with a report you can export. There is more on what an honest attendance record looks like in live class attendance, measured honestly, and the room itself is described on the live classes page.

The person paying is usually not the person in the lesson. A parent cannot see what happens in the hour. They see the charge on their card and, months later, a grade. Everything between those two points is your job to make visible, and reporting it well is how you keep a family for years rather than for a term.

Start with a baseline. Before the first real lesson, give a short diagnostic test and keep the score. Without a starting number, 'she is doing much better' is an opinion. With one, it is a fact you can point to.

For test prep, schedule full practice tests under timed conditions at regular points in the program, and treat the review as the real lesson. A practice test tells you three things a lesson cannot: which topics cost the most points, whether the student runs out of time, and whether the errors come from missing knowledge or from carelessness. Each of those needs a different fix, and a student who knows which one is theirs works on the right thing at home.

Then report, briefly and on a schedule. A useful progress report fits on a phone screen:

• Sessions attended, out of sessions held.
• What was covered since the last report, in plain words.
• One number that moved: a practice score, a quiz result, a piece now played at full tempo.
• The one thing to work on at home this week.
• What comes next, and the date of the next report.

Send it every two to four weeks, even when the news is flat. A flat month reported honestly ('the practice score held at 610, timing is the problem, here is the plan') builds more trust than three months of silence followed by a good result. Silence is what makes a parent start wondering what they are paying for.

This is one place where software saves real hours. On Prolaud you can build timed, auto-graded quizzes for practice, and if a guardian's email is on the student's record, the result of an auto-graded quiz is emailed to that guardian as well. At the end of a program you can issue a certificate of completion, which gives the student something to show for the work. For how to write questions worth asking, see quiz and assessment best practices.

There are two ways money can reach you. In the first, a marketplace or an agency collects from the family, keeps a share, and pays you later. In the second, the family pays you directly through a payment processor, and the money lands in an account with your name on it. The first is a fair trade when the marketplace is the one finding the students. Once the students are yours, the second is the one to build on.

Being paid into your own account means three practical things. The payout schedule is between you and your processor rather than a platform's policy. The customer relationship and the payment history stay with you if you ever change tools. And the cost is visible: you pay your payment processor's standard fee on each charge, and nothing is taken on top of it.

A checkout worth using does four jobs without you. It takes a card. It gives the family a receipt. It grants access to the class or the course the moment the payment succeeds. And it collects later installments on schedule. If any of those needs you to send a message, it will eventually fail in the week you are busiest.

Prolaud works this way. You connect your own Stripe or Razorpay account, students pay on your site, and your processor settles the money to you. Prolaud takes 0% of your sales. A program price can be split into 2 to 12 installments, and with auto-debit switched on the later ones are collected for you; the details are on the payment plans page.

The signal is not ambition. It is a waiting list: families you would have said yes to, turned away for a month or more because your week is full. Before that point, a second tutor is a cost looking for work.

Raise your price before you hire. A full calendar with a waiting list is what a price that is too low looks like. If a 10% increase for new students empties the waiting list, you did not need a second tutor. If the list is still there, you do.

Then run the numbers per session. Suppose you charge $70 for a private hour and pay a second tutor $40 to teach it. That leaves $70 − $40 = $30, before your payment processor's fee and before the time you spend finding the student, matching them, talking to the parent and checking the quality. At ten sessions a week that is 10 × $30 = $300 a week, or $300 × 52 ÷ 12 = $1,300 a month. Whether it is worth it depends on how many of your own hours the managing takes. If it takes five hours a week, you are earning $300 ÷ 5 = $60 an hour to manage, which is less than the $70 you earn teaching.

That arithmetic is why a second tutor makes most sense for hours you could not teach anyway: a subject outside your own, a time zone you are asleep in, or group sessions where one tutor's hour serves six families. Settle four things in writing before their first lesson:

• How they are paid: a fixed amount per session, a percentage of what the student pays, or a fixed amount per enrolled student.
• Who holds the student relationship, and what happens to those students if the tutor leaves.
• What every lesson must include: your diagnostic, your report format, your attendance record. Families bought your standard, not a stranger's.
• Employee or independent contractor. The tests for telling the two apart vary by country and state and carry real penalties when you get them wrong, so take advice. This is not legal advice.

On Prolaud you can add a teacher, give them access to the courses they teach, and record the pay terms you agreed: a percentage split, a fixed rate per class, or a fixed amount per student. It keeps track of what each teacher is owed. It does not split the payment for you; you transfer their pay yourself.

1. Week 1. Write the one-line offer. Work out the rate you need from your own income target and hours. Write the three policies: cancellation, make-ups, refunds.
2. Week 2. Build the diagnostic test. Choose three fixed weekly slots. Set up payment into your own account and run one test payment from start to finish, including the receipt.
3. Weeks 3 and 4. Tell everyone who already trusts you, one at a time. Run free diagnostics. Enroll your first three students on a monthly package.
4. Weeks 5 to 8. Teach. Record attendance every session. Send the first progress reports. Ask the first happy family for a sentence you can quote and one introduction.
5. Weeks 9 to 12. When three students share a goal and a level, open your first small group. Replace the 80% planning assumption with your real number, and look at your price again.

• Tutoring every subject for every age, so nobody can describe what you do.
• Copying a rival's rate instead of working out your own from your income target and hours.
• Selling single sessions forever, and living with the cancellations that come with them.
• Skipping the baseline test, which leaves you with no proof of progress later.
• Reporting to parents only when they ask, which is usually just before they leave.
• Letting a platform hold the customer relationship and a share of every payment long after it stopped finding you students.
• Hiring a second tutor before raising your price.

• How much to charge for an online course: the pricing method in full, with tiers, payment plans and price rises.
• How to get your first 100 students: what to do once the first ten are enrolled.
• Live class attendance, measured honestly: what a record a parent can rely on has to contain.
• Quiz and assessment best practices: writing practice questions that show what a student actually knows.
• What it costs to run your classroom and your checkout as two systems: the month-end work that grows with every student you add.