Skip to main content

Create Online Courses That Sell: The Complete Blueprint

Create online courses that sell: pre-sell before you record, design six-minute lessons, price after platform fees, and launch with a real deadline. 2026.

Founder of IImagined.ai

Published
Feb 22, 2026
Updated
Oct 11, 2026
Reading time
14 min read
Quick answer

To create online courses that sell, reverse the usual order: pre-sell a dated first run before you record, design the course around one result in lessons of about six minutes, set the price after subtracting platform fees, and launch to an email list with a real deadline. On an illustrative $197 course, Gumroad leaves about $170.79 a sale with no monthly fee, while Thinkific Basic leaves $190.99 a sale after a $54 plan and passes Gumroad from three sales a month.

To create online courses that sell, sell the course before you record it: pre-sell a clear result to a small audience, build only what those buyers need in lessons of about six minutes, price it with platform fees already subtracted, and launch to an email list with a real deadline. Most courses that do not sell were made in the opposite order: recorded first, priced by guesswork, and shown to an audience last.

Platform prices and fees checked on October 11, 2026 against the Gumroad, Payhip, Whop, Teachable, Thinkific, Kajabi, Skool and Podia pricing pages. Lesson-length figures are from Guo, Kim and Rubin's 2014 edX study. Legal points link to the FTC and EU pages they come from. This rewrite replaces the February 2026 version, which quoted price bands, list sizes and pre-order targets we could not source.

This is the business blueprint: what to decide, in what order, and what the numbers look like. The production side (outlining with AI, slides, recording, captions and hosting) is in our step-by-step tutorial on how to create and sell an online course. Both sit under the courses and memberships hub, which covers where to host and when a membership fits better.

Who this blueprint is for

It is for someone who knows a subject well enough to get another person a result, and who can reach at least a few people with that problem. It is not a way to find a topic, and it will not work without an audience of some kind. If you have neither yet, start smaller with one of our mini course ideas.

The system: create online courses that sell, end to end

Six stages, in this order. The order is the blueprint. Each stage produces the evidence the next one needs.

From idea to repeat sales
  1. 01
    Validate

    Find proof that people already pay for this result.

  2. 02
    Pre-sell

    Take payment for a dated first run before you record.

  3. 03
    Design

    One result, modules as milestones, lessons of about six minutes.

  4. 04
    Build

    Record for the buyers you already have.

  5. 05
    Price and host

    Subtract platform fees first, then choose the plan.

  6. 06
    Launch and repeat

    An email list, a real deadline, then an evergreen sequence.

Phase 1: validate, then pre-sell before recording

A course sells when a specific person wants a specific result enough to pay for a structured path to it. Start with one sentence: who it is for, what they can do at the end, and why free content has not got them there.

Then look for evidence that people already pay. Two kinds are easy to find: questions you are asked again and again in comments, messages or client work, and paid courses, books or coaching on the same result. Competition shows buyers exist, and the critical reviews of those products show what they leave out. Our validation guide has the full method.

Run a pre-sale, not a survey

A waitlist shows interest. Payment shows demand. Podia's waitlist mode collects email addresses while you build, which is a fair first step, but the stronger test is asking for money. Thinkific's pre-order status lets buyers pay while the content stays locked, and Podia's start date setting charges at purchase and opens access on a date you choose (both checked October 2026).

Record first vs sell first
Record first
  • Months of recording before anyone has paid
  • A price guessed at the end
  • The audience sees it for the first time on launch day
  • Lessons built on what you think buyers need
  • A quiet launch tells you nothing about why
Sell first
  • An outline and a sales page before any video
  • A green-light goal set in advance
  • First buyers shape the lessons
  • A published delivery date and refund promise
  • A missed goal costs a few weeks, not a quarter

Thinkific's guide calls the target a green-light goal: the number of sales, or the revenue, you need before you give yourself permission to build. Set it before you open the pre-sale. The same guide gives three rules worth copying whatever platform you use:

  1. Publish a hard delivery date on the sales page, and tell buyers if it moves.
  2. Send progress updates, weekly if you can, so buyers can see the course being made.
  3. Refund everyone if you miss the goal or cannot deliver.

Offer the pre-sale to your warmest people first: past buyers, clients and subscribers who reply. If it does not reach the goal, you have lost a few weeks. Change the promise or the topic and run it again.

Phase 2: design the course around one result

Work backwards from the end state: "by the end of this course you can ___". Each module is a milestone on the way there, and each lesson teaches one thing and ends with an action. Anything that does not move the student towards the result comes out.

PartWhat it doesExample: a course on starting a newsletter
First moduleAn early, visible win that proves the method worksSet up the newsletter tool and send the first issue
Core modulesOne milestone each, in the order a beginner needs themA writing system, growth channels, sponsorships
PracticeA worksheet, template or assignment that applies each moduleA content calendar and a sponsor pitch worksheet
Wrap-upNext steps, and what the course does not coverA launch checklist and a 90-day plan

Lesson length: what the data says

Keep each video to about six minutes. That figure comes from the largest public study of course video we know of: Guo, Kim and Rubin analysed 6.9 million video watching sessions across four edX courses and found that median engagement time was at most six minutes, however long the video was. Students often got less than halfway through videos longer than nine minutes.

Students who attempted the follow-up problem, by video length
0 to 3 minutes
56%
3 to 6 minutes
48%
6 to 9 minutes
43%
9 to 12 minutes
41%
12 to 40 minutes
31%

Share of video watching sessions followed by an attempt at the assessment problem that came after the video. Source: Guo, Kim and Rubin, How Video Production Affects Student Engagement, L@S 2014 (four edX courses, Fall 2012)

Read the limits before you rely on it. The courses were free university maths and science courses from 2012, the measure is engagement and not learning, and a student who has paid may sit through more. It is still the best evidence available, and it points one way. The same paper has three more findings that are cheap to act on:

  • Plan the split before you record. The authors recommend segmenting lessons into chunks shorter than six minutes at the planning stage, not cutting a long recording afterwards.
  • Show your face some of the time. Videos that mix a talking head with slides were more engaging than slides alone.
  • Skip the studio. Videos with a more personal feel could be more engaging than high-fidelity studio recordings, and instructors who spoke fairly fast and with enthusiasm held attention better.

Designing so recorded students finish

A recorded course has no teacher in the room and no deadline, so the structure has to do that work. Put a real win in the first module. End every lesson with something to do, and give each module a worksheet where the student applies it; our guide on how to make a workbook covers the build. Show the finish line: a short list of milestones the student ticks off. If your topic needs feedback or accountability to work, a recorded course may be the wrong format, and cohort based courses vs self-paced covers that choice.

Phase 3: price with the platform fees already subtracted

No formula turns hours of video into a price. Price from the problem you solve, what buyers already pay for alternatives, and how much of your time the offer includes. Then check what the platform leaves you, because the fee model changes the answer.

  • Compare alternatives. Other courses, books, coaching, or paying someone to do the job.
  • Price support separately. Self-paced access costs you little per buyer. Calls and feedback cost hours, so sell them as a higher tier.
  • Use real deadlines. A launch price is fine if it actually ends.
  • Keep income promises out. "This course pays for itself" is an earnings claim that needs evidence.

Three fee models

Platforms charge a fee on every sale, a monthly plan, or both. Prices below are for monthly billing unless stated, checked October 11, 2026.

PlatformMonthly pricePlatform fee per saleWorth knowing
Gumroad$010% + $0.50, plus card processing of 2.9% + $0.3030% on sales through its Discover marketplace, processing included
Payhip$0, $29 or $995%, 2% or 0%, plus processor feesIts free plan has no feature restrictions
Whop$0None; card processing is 2.7% + $0.30Payout fees from $2.50; optional add-ons cost extra
Teachable$39, $89 or $1897.5% on Starter; 0% on Builder and Growth with Teachable PaymentsStarter allows 5 products, Builder 10
Thinkific$54, $109 or $219None with Thinkific PaymentsUsing Stripe instead adds a 5% gateway fee on Basic
Kajabi$179, $249 or $499None with Kajabi PaymentsThird-party payment providers add 2% on Basic (PayPal excluded)
Skool$9 or $9910% on Hobby; 2.9% on ProCourses and community on both plans
Podia$42 or $84, billed yearly5% on Mover; none on ShakerSupports waitlists and dated pre-launches

Revenue scenarios on a $197 course

Here is what is left after fees and the monthly plan at three sales volumes. The price and the volumes are illustrative inputs, not forecasts. Card processing is 2.9% plus $0.30 on every row except Whop, which publishes 2.7% plus $0.30.

Platform and planMonthly plan3 sales a month10 sales a month30 sales a month
Gumroad$0$512.36$1,707.87$5,123.61
Payhip Free$0$543.41$1,811.37$5,434.11
Whop$0$574.14$1,913.81$5,741.43
Teachable Starter$39$489.64$1,723.12$5,247.36
Thinkific Basic$54$518.96$1,855.87$5,675.61
Teachable Builder$89$483.96$1,820.87$5,640.61
Kajabi Basic$179$393.96$1,730.87$5,550.61

Three things fall out of the table. Gumroad's 10% plus $0.50 is $20.20 on each $197 sale, so Thinkific Basic passes it at three sales a month, Teachable Builder at five and Kajabi Basic at nine. Whop stays ahead of every flat plan on fees alone, because its card rate is lower and it has no plan to pay for; the flat plans earn their price through course tools, not through fees. And at thirty sales the gap between the best and worst row is about $618 a month, which is real money but small next to the difference between a course that sells thirty copies and one that sells three.

To run your own price and volume, the free platform break-even calculator compares Gumroad, Payhip, Whop and a plain Stripe checkout and shows where one overtakes another. For plan limits, tax handling and payouts, see our course platform comparison, and Skool vs Whop vs Teachable if a community is part of the offer.

Phase 4: launch to a list, then keep selling

A launch is a short period where you ask an audience that already knows you to decide. It works through email, because email is the one channel no platform can take away from you.

A launch week with a real deadline
  1. Weeks before
    Build the list

    A free resource tied to the course topic, and free lessons that show how you teach.

  2. Day 1
    Open the cart

    One email with the result, the curriculum, the price and the closing date.

  3. Day 3
    Answer objections

    Who it is for and not for, the refund policy, and real feedback from the first run.

  4. Day 4
    Hold a live session

    People ask the question that is stopping them from buying.

  5. Last day
    Close on time

    One reminder. The deadline is real, so the cart closes or the launch price ends.

  6. After
    Go evergreen

    An automated sequence for new subscribers, or reopen the cart for each new group.

The sales page has to carry the same information every time: who the course is for and not for, the result, the full curriculum, the support included, the price, the refund policy and an FAQ. Our guide to the sales funnel for digital products covers the pages and emails around it.

Launch emails, the evergreen sequence and the funnel behind them are where a blueprint runs out of room. The Digital Products program has full lessons on the pre-sell method, course recording, pricing a course and the five-day email launch.

This is general information for course sellers in the US and EU, not legal advice.

  • Refunds. Choose a policy, state it on the sales page and at checkout, and honor it. EU consumers generally have 14 days to withdraw from an online purchase; for digital content the right ends once they start downloading or streaming after expressly agreeing to that.
  • Earnings claims. If the course teaches people to make money, income screenshots and lifestyle shots are claims. In June 2026 the FTC told businesses to back up their earnings or lifestyle claims, and said unusually successful results from a handful of people do not by themselves support a claim that others will earn the same.
  • Testimonials. The FTC's Consumer Reviews and Testimonials Rule, in effect since October 21, 2024, covers fake or false testimonials, incentives tied to a positive review, and undisclosed testimonials from staff or relatives. Courts can impose civil penalties for knowing violations. Generating a testimonial with AI does not make it any less fake.
  • Marketing email. Under the US CAN-SPAM Act, commercial email needs accurate headers and subject lines, a valid physical postal address and an opt-out honored within 10 business days. The FTC lists penalties of up to $53,088 per email.
  • Sales tax and VAT. Course sales can create tax obligations where your buyers live. Some platforms collect and remit it for you and some leave it with you, so check before you choose one.

The 90-day plan

Thirty days to find out whether anyone will pay, thirty to teach the first group, thirty to turn it into a finished product.

Ninety days from idea to public launch
  1. 1
    Days 1 to 7: write the promise

    One sentence: who it is for, what they can do at the end, why free content has not got them there.

  2. 2
    Days 8 to 14: gather evidence

    List the questions you are asked repeatedly. Read the critical reviews of three paid alternatives.

  3. 3
    Days 15 to 30: pre-sell

    Outline, sales page, price, delivery date and a green-light goal. Offer it to your warmest audience.

  4. 4
    Day 30: decide

    Goal met: build. Goal missed: refund everyone, say so plainly, and change the topic or the promise.

  5. 5
    Days 31 to 60: run the first version

    Teach the first group live or in weekly releases. Note every place they get stuck.

  6. 6
    Days 61 to 80: rebuild as finished lessons

    Short recorded lessons, one action each, with captions and a worksheet per module.

  7. 7
    Days 81 to 85: price and platform

    Subtract fees at your expected volume and pick the plan that leaves the most.

  8. 8
    Days 86 to 90: public launch

    Open to the full list with a real closing date, then switch on the evergreen sequence.

Do this week
  • Write the one-sentence promise and show it to three people in your audience
  • List five questions you are asked again and again
  • Read the one-star and three-star reviews of three paid alternatives
  • Draft a ten-lesson outline where every lesson ends in an action
  • Pick a price and subtract the platform fees at ten sales a month
  • Set a green-light goal and a delivery date you can keep
  • Write the refund line you will put on the sales page

Mistakes that stop courses selling

  • Recording before selling. The most expensive way to find out nobody wanted it.
  • Teaching a topic, not a result. "Everything about email marketing" has no finish line. "Send your first paid newsletter" does.
  • Hour-long lessons. The engagement data points to about six minutes. Split at the planning stage.
  • Pricing by hours of video. Buyers pay for the result, and more hours can make a course harder to finish.
  • Choosing a platform by name. Choose by fee model at your volume, then by the features you will use.
  • Fake deadlines. A timer that resets teaches your list to ignore the next one.
  • Launching once. A course that is offered once is sold once. Build the evergreen sequence before you move on.

Creating online courses that sell: FAQ

How do I create an online course that sells?

Sell it before you build it. Write a one-sentence promise for a specific buyer, check that people already pay for that result, then pre-sell a dated first run to your warmest audience with a sales target set in advance. Build only after the target is met, in short lessons that each end in an action, and set the price after subtracting platform fees.

Should I pre-sell an online course before creating it?

Yes, if you can deliver on a date. A pre-sale shows demand in the only way that counts, payment, and the first buyers tell you what the lessons need to cover. Thinkific has a pre-order status and Podia lets you set a start date, and both charge at purchase. Publish the delivery date, send progress updates, and refund everyone if you cancel. Checked October 2026.

How long should online course lessons be?

About six minutes or less per video is the best-supported figure. A study of 6.9 million viewing sessions on edX found median engagement was at most six minutes however long the video ran, and recommended splitting lessons into chunks under six minutes. The data comes from free university courses in 2012, so treat it as a strong default, not a rule. The course itself should be as long as the result needs.

Do I need a big audience to sell an online course?

No fixed number exists, and anyone who quotes one is guessing. What you need is a way to reach people who have the problem: an email list, a following, clients or a community you belong to. A small, warm audience is enough to pre-sell a first run and learn from it. Thinkific's own pre-selling guide suggests starting with warm leads and widening only once the message works.

How much should I charge for an online course?

There is no reliable typical price. Set it from the problem you solve, what buyers already pay for alternatives such as books, coaching or other courses, and how much of your time the offer includes. Then subtract the platform fees at your expected sales. On a $197 course, a Gumroad sale leaves about $170.79 and a Thinkific Basic sale about $190.99 before the $54 monthly plan. Checked October 2026.

When does a monthly course platform beat a per-sale fee?

When the plan costs less than the per-sale fees it replaces. On a $197 course, Gumroad's 10% plus $0.50 is $20.20 a sale, so Thinkific Basic at $54 a month comes out ahead from three sales a month, Teachable Builder at $89 from five and Kajabi Basic at $179 from nine. Whop has no plan and a 2.7% card rate, so on fees alone it stays lowest at any volume. Checked October 2026.

Free · no card needed

Start with the free Starter Kit.

A pricing worksheet from Digital Products, plus four more working files.

Starter Kit5 files · free
All Access · all four programs · $99/mo

Pre-sale met its goal? Now build the business around it.

The Digital Products program, included in All Access, goes lesson by lesson through the pre-sell method, course recording, pricing a course, the five-day email launch and the funnel that keeps selling afterwards. The other three programs, live coaching and the private community come in the same subscription.

See All Access: $99/mo →30-day money-back guarantee
Free

Run your own fee scenario

The free break-even calculator shows what each platform leaves at your price and volume. The free Telegram channel posts platform pricing changes as we spot them.