How to Create an Online Course in 2026: A Step-by-Step Guide for Experts

You've built an audience. People ask you the same questions every week. At some point, someone says it: "You should turn this into a course." Here's how to do that without spending three months filming something nobody buys.

The short answer

To create an online course, follow these seven steps:

1. Test the topic. Pre-sell it before you build it.
2. Pick a format: self-paced, cohort or hybrid.
3. Plan it. Start with the outcome, then build the modules.
4. Record with gear you already own. Upgrade later.
5. Set a price.
6. Choose where to host and sell it.
7. Launch it, and plan your first 30 days.

Most guides go straight to cameras and software. This one spends more time on the parts they skip: checking that people will pay, picking the right format, and the money side of it all.

Pick a topic people will pay for

Here's the rule: get money, or at least a firm yes, before you record a single lesson.

It sounds backwards. A lot of experts want the course finished and polished before they ask anyone to buy it. But a finished course nobody wants is the most expensive mistake in this whole process. You lose weeks of work, and you learn nothing about what your audience actually wanted.

What makes a good course topic

Look for a problem people already spend money or time on. If they're buying books, hiring freelancers or watching hours of YouTube to fix it, that's a good sign.

You also want a clear before and after. "You can't run Facebook ads" turns into "You can launch a campaign that pays for itself." If you can't describe the change in one line, the topic is probably too vague.

Still stuck? Look at your inbox. The question people ask you most often is usually the best place to start.

Three ways to test the idea

Pre-sale. Sell the course before it exists. Offer a founding price and a set start date. If nobody buys, you don't build it, and you refund anyone who did. That stings a little, but it's much cheaper than filming a flop.

Waitlist. Put up a simple sign-up page that explains the course. Share it and watch how many people join. Then email that list a pre-sale offer. A big list that won't buy is useful info too.

Paid workshop. Run a 60 to 90 minute live session on one small slice of the topic. If people pay for a slice, they'll likely pay for the whole thing. You also get to hear the questions people ask, and those questions make a great outline.

Self-paced, cohort or hybrid

This choice shapes your price, your workload and how many students actually finish. Here's what each one means.

• Self-paced: recorded lessons. Students start whenever they want and go at their own speed.
• Cohort: a group starts on the same date and moves through together. There are live sessions and deadlines.
• Hybrid: recorded lessons plus something live, like a weekly call or a group chat.

Which format do students finish?

Big free courses have a completion problem. Researchers Justin Reich and José Ruipérez-Valiente looked at six years of Harvard and MIT courses on edX. In "The MOOC Pivot" (Science, 2019), they found completion rates stayed low and never really improved.¹

Paid courses from independent creators look different. Course platform Ruzuku studied its own data, covering 27,156 courses and 1.31 million enrollments. This is one platform looking at its own users.

Source: Ruzuku completion study²

The sweet spot was cohorts of 11 to 25 students. That's big enough for good discussion, but small enough that nobody hides.

What each format lets you charge

Ruzuku says cohort courses sell for 2 to 5 times more than self-paced video courses.³ On its platform, the median one-time price for a self-paced course is $150, while live cohort courses routinely charge $497 to $1,898.

The gap makes sense. With a cohort, students get your time and a group to keep them going. That's harder to find for free.

Which one fits you?

Pick self-paced if you have a large audience and a topic that won't go out of date. You build it once and sell it for years.

Pick cohort if the skill needs feedback, or your students are the type who need deadlines to get things done.

Pick hybrid if you want some of both. Record the lessons, then add a weekly call.

Plan the course

Start at the finish line. Before you outline a single module, write one sentence about what a student can do when they're done.

For example: "By the end, you'll have a working Shopify store with your first five products listed." That's much better than "you'll understand e-commerce." Understanding is hard to see. A working store isn't.

That sentence becomes your filter. Every module has to move the student closer to it. If a lesson doesn't, cut it, even if it's your favorite part. You can always turn it into a bonus or a blog post.

Build the modules backward

Work back from the outcome. Ask yourself what a student needs to have done right before the finish. Then what comes right before that. Keep going until you reach where a beginner starts.

Each module is one milestone. Each lesson inside it is one small step toward that milestone. If a lesson needs two steps, it's probably two lessons.

How long should it be?

Short enough to finish. There's no magic number, and anyone who gives you one is guessing.

Here's one real data point, though. Ruzuku's report on yoga courses found a median of 7 modules and 28 lessons.⁴ That's just one niche, so treat it as a rough guide.

Give students a win in week one

This might be the most useful number in the whole guide. In Ruzuku's business and professional course data, students who finished at least one lesson in their first week completed the course 73.9% of the time.⁵ Students who didn't? Just 14.3%.

So make lesson one short, and make it pay off fast. Give them something they can use that same day. That early win is what brings them back for lesson two.

Make it

Start with what's already in your pocket. Your phone already shoots good enough video for a first course.

Course creator Philip Hofmacher, who has made multiple six figures from his courses, recorded his first one with an old phone, no microphone and no lighting. As the Write Build Scale episode with him (January 2026) puts it:⁶ "Your first course does not need a professional setup, studio lighting, or fancy gear."

If you'll only spend money on one thing, spend it on sound. A lot of creators will tell you students forgive grainy video. Muddy, echoing audio makes them click away.

How much does it cost to create an online course?

Less than most people think. Here are three setups, from free to pro, at list prices checked in October 2026.

Prices: Shure MV6 and MV7+,¹³¹⁴ RØDE PodMic USB,¹⁵ Logitech C920s,¹⁶ Elgato Key Light Mini¹⁷ and Descript.¹⁸

Platform fees are separate. See our breakdown of Teachable pricing for one example.

Teaching software? Record your screen

If your course teaches a tool, like Excel, Canva or a coding language, you may not need a camera at all. Screen recording with a voiceover is often clearer than a talking head. Students mostly want to see your clicks.

A simple order to follow

1. Record one test lesson with your phone.
2. Watch it back and listen closely.
3. If the sound is the problem, buy a mic first.
4. Only upgrade the camera or lights once students start asking for it.

Price it

Pricing is where a lot of experts sell themselves short. They look at a cheap course on a marketplace and assume theirs has to compete with it. It doesn't.

Here's what real courses sell for on one platform. All of these numbers come from Ruzuku's own data, so they show one corner of the market.

Sources: Ruzuku reports on coaching, education and business courses.⁷⁸

For context, the median across Ruzuku's 11,691 paid courses is $150, and the middle half sell for $49 to $357.⁹ Coaching courses sit far above that line.

Heartbeat's own data lands in the same place. Across 716 one-time products sold on Heartbeat (courses, workshops and programs), the median price is $149, the middle half sell for $47 to $462, and about one in eight charge over $1,000. If you plan to bill monthly instead, our membership pricing calculator suggests a price from what similar communities on Heartbeat charge.

Why access costs more than content

Think about it from the buyer's side. Recorded videos are easy to compare. A student can find something similar on YouTube, or a cheaper course that covers the same ground.

What they can't easily find is you. Live answers, feedback on their actual work and a group of people stuck on the same problem are all hard to copy. That's why Ruzuku sees cohort courses selling for 2 to 5 times more than self-paced ones.

So if you want to charge more, add more of yourself.

Three ways to raise your price

• Add live calls. Even one group call a week changes how buyers see the value.
• Add feedback. Review homework, projects or recordings.
• Add a community. If you want ongoing income, a paid group works well alongside a course. A RevenueGeeks study of Skool, which we covered in our guide to online communities, found trending paid groups charged a median of $39 a month in September 2026.¹⁰

Start with a founding price

Your first round of students is helping you test the course. Give them a lower founding price, and tell them the price goes up after launch. They get a deal, and you get feedback and testimonials to justify the higher price next time.

Where to host and sell it

You've got three main kinds of places to choose from. Marketplaces bring buyers to you, which is handy when you're starting out. The trade-off is that they control your price and keep your customer list. Course platforms you run yourself flip that around. You own the site, the email list and the price, but you have to bring the buyers yourself. Community and cohort platforms are built around live sessions and groups, so they suit cohort and hybrid courses best.

If you already have an audience, your own course platform or a cohort platform usually makes more sense than a marketplace. You've done the hard part of finding people. There's no reason to give away your list or your pricing. For a side-by-side look, see our comparison of the best online course platforms.

Launch it

Think of your launch as a whole month. Here's a simple plan you can copy and adjust to fit your audience. For where to sell it, what each option keeps from a sale and the marketing after launch, see how to sell online courses.

Days 1 to 7: open the doors

Email your waitlist first. They raised their hands early, so they get first dibs. Open enrollment with your founding price and a clear deadline. A deadline matters more than you'd think. Without one, "I'll sign up later" quietly turns into never.

Days 8 to 14: show up live

Run a free live session or Q&A. Teach one useful thing, then answer questions about the course. People who are on the fence usually just need to hear you talk. Then send a few reminder emails before the cart closes, including one on the last day.

Days 15 to 21: welcome your students

Close enrollment and say hello. A short welcome video or email goes a long way. Then push hard on one thing: getting everyone through lesson one in the first week. Remember those Ruzuku numbers. Students who finished a lesson in week one completed the course 73.9% of the time, compared with 14.3% for everyone else.

Days 22 to 30: listen and fix

Ask students what's working and what's confusing. Collect testimonials while the excitement is fresh. Find your weakest lesson and fix it. Then pick a date for round two, so the next launch is already on the calendar.

You don't need to be everywhere

It's easy to think a launch means posting on every social channel, every day. Web designer Paige Brunton, whose blog said in 2022 that her courses had passed $2 million in revenue, told SquareKicker she tested a launch without promoting it on Instagram:¹¹¹² "We saw no difference in the launch."

Put your energy into the channel where your buyers already listen to you, which for a lot of experts is email.

Where Heartbeat fits

Heartbeat is built for courses that come with people attached. You can run a course evergreen or as a cohort, drip lessons on a schedule, and add discussions, assignments and documents next to each lesson, so the group you sold is in the same place as the material. Live sessions run as events with Zoom set up for you, attendance tracking and recordings that come back to the course.

Checkout handles one-time prices, installment plans, subscriptions and bundles, and you can offer free teaser lessons before someone buys. Heartbeat has a 14-day free trial, with no card required to start.

Frequently asked questions

Can ChatGPT create a course?

It can help. ChatGPT is good at drafting an outline, writing quiz questions or roughing out a lesson script. What it can't give you is your experience, your stories or your face on camera, and those are the reasons people buy from you. Use it for the first draft, then check every fact yourself.

Is there a free way to create online courses?

Yes, at least at the start. Your phone, a window for light and your phone's built-in editor will get you recording. Many platforms also have a free or low-cost way to sell. Look at the $0 row in the cost table above.

How much does it cost to build an online course?

It can cost nothing if you use what you already own. A mid-range setup with a USB mic, webcam and light runs about $299 in gear, plus editing software from $16 a month. See the full breakdown in the cost table.

Do online courses make money?

Some do very well. Paige Brunton said on her blog in 2022 that her courses had brought in more than $2 million. Others don't. The courses that sell tend to be tested first, priced with care and launched to an audience that already trusts the creator. For a sense of typical prices, Ruzuku's data puts the platform-wide median at $150.

How long should an online course be?

As short as it can be while still getting students to the result. Start with the outcome and cut anything that doesn't serve it. One Ruzuku report on yoga courses found a median of 7 modules and 28 lessons, but that's one niche.

Should I pre-sell my course?

Yes, if you can. A pre-sale tells you people will pay before you spend weeks building. If it doesn't sell, you've saved yourself the work.

Is self-paced or cohort better?

Cohorts get more students to the end. In Ruzuku's data, they had a 53.4% average completion rate, compared with 41.9% for open self-paced courses, and they tend to sell for more. Self-paced is easier to scale.