There is a version of this guide that starts with camera gear. This is not that guide. If you want to create and sell an online course, almost everything that decides whether it works happens before you record a second of video — who it is for, what it promises, whether anyone will pay, and how the buyer gets from a link to a lesson without emailing you for help.
What follows is the whole path in order: validating the idea, outlining it so students finish, recording it without a studio, pricing it, choosing where to host and sell it, launching it, supporting students, and then improving it. Along the way we will be specific about the parts people usually skip, because those are the parts that quietly kill first courses.
This is written for someone selling their first or second course — a freelancer, a coach, a tradesperson, a teacher, someone with a skill and an audience of a few hundred people. If you have a hundred thousand followers, the same steps apply with bigger numbers.
Validate the course idea before you build anything
The most expensive mistake in online courses is spending six weekends recording something nobody asked for. Validation is simply finding out, cheaply and early, whether people will pay. It is not asking friends whether your idea sounds good — everyone says yes to that question, because saying no is rude and costs them nothing.
The only honest signal is money or a commitment that feels like money. That means a pre-order, a paid waitlist deposit, a live pilot round at a reduced price, or a filled booking calendar for a workshop version of the same material. In our experience, a small warm audience that will not buy at half price is telling you something clear, and it is worth listening before you buy a microphone.
There is a softer signal worth collecting too: the questions you already get asked. If five people this month asked you the same thing in DMs, emails or client calls, you have found a topic with demonstrated demand. Write those questions down verbatim — they become your module list and later your sales page copy.
- Pre-sell at a discount to a small list and set a minimum number of buyers you need before you record
- Run the material once as a paid live workshop, then edit the recording into the course
- Open a waitlist with one question: what are you stuck on right now?
- Count the questions you already answer for free every week
- Refund everyone without argument if you decide not to build it
Choosing an online course topic people will actually pay for
Paid courses solve a problem the buyer already knows they have. That is the difference between a topic that sells and a topic that merely interests people. Nobody wakes up wanting to learn colour theory; plenty of people wake up wanting their product photos to stop looking cheap. Same knowledge, different framing, very different conversion.
The test is whether you can finish this sentence with a specific person and a specific result: this course helps ____ get ____ in ____ time. If the blank is filled with a broad audience like anyone or a vague result like get better at, keep narrowing. Narrow topics feel like a smaller market and behave like a bigger one, because the promise is believable.
You also need standing to teach it. That does not mean credentials — it means evidence. Screenshots, before-and-afters, client results, your own work, students you have already helped one to one. Buyers are not looking for the world expert; they are looking for someone a few steps ahead who can prove it.
- Good: help small cafe owners shoot menu photos on a phone that look professional
- Weak: a photography course for beginners
- Good: help new freelancers write proposals that win the job at a higher rate
- Weak: freelancing tips
- Good: help HR managers run a hiring process that fills a role in three weeks
How to outline an online course students actually finish
Start at the end. Write the outcome, then work backwards through the steps required to reach it, then delete every step that is interesting but not required. Most first courses are twice as long as they need to be because the creator taught everything they know instead of everything the student needs.
Structure it as modules of related lessons, with each lesson doing exactly one job and each module ending in something the student has produced. A module that ends with a finished thing — an outline, a price list, a first draft, a set-up account — creates momentum. A module that ends with more theory creates drop-off.
Keep lessons short. Five to twelve minutes is a comfortable range for most material, and it makes editing far easier, because a fluffed line costs you one short retake. Long lessons also hide structural problems: if you cannot say what a lesson is for in one sentence, it is probably two lessons or none.
- Write the outcome first, in the student's words
- List every step between where they start and that outcome
- Cut anything that is background rather than action
- Group steps into modules of three to six lessons
- End each module with a deliverable the student has made
- Add one worksheet, template or checklist per module — these are often what buyers value most
What to record on video, and what to write instead
Video is not automatically the best format, it is just the default. Reference material — checklists, templates, prompt lists, swipe files, spreadsheets — is better as a document, because students need to scan it, not watch it. Putting reference material on video is one of the most common reasons students report that a course felt long.
Use video where watching matters: demonstrations, walkthroughs where you do the thing on screen, critiques, and anything where tone and judgement carry the lesson. Screen recordings tend to be more useful than talking-head footage in skills courses, and they are dramatically easier to produce well.
A practical mix for a first course is roughly seventy percent screen recording, twenty percent talking head for the framing lessons, and a set of downloadable documents that carry the reference load. That mix also lowers the cost of updating the course later, because a changed template is an upload, not a reshoot.
How to record an online course without a studio
Audio is the thing people notice and gear reviews under-sell. A modest USB or lapel microphone in a soft-furnished room will outperform an expensive camera in an echoing kitchen every time. Curtains, a rug, a wardrobe of clothes behind you — soft surfaces are free acoustic treatment.
Record in a consistent set-up so lessons cut together: same time of day if you rely on window light, same microphone position, same screen resolution, same browser zoom for screen recordings. Inconsistency between lessons reads to students as sloppiness even when the teaching is excellent.
Do not script word for word unless you are unusually good at reading naturally. Use a bullet outline on a second screen, record in one take per lesson, and accept small stumbles. In our experience, courses recorded conversationally get better completion rates than perfectly polished ones, because students feel taught rather than presented at.
Batch it. Set up once, record four or five lessons in a session, and do all your editing in a separate block. Creators who record one lesson per week almost always stall around lesson four.
- A quiet room with soft furnishings beats a fancy camera
- Test your first two minutes and listen back on phone speakers, not headphones
- Record lessons short so retakes are cheap
- Keep a consistent set-up so lessons feel like one course
- Batch recording days; edit on a different day
How to price an online course
Price against the outcome, not the runtime. If your course helps a freelancer raise their rates or win one extra client, its value is measured in that client, not in your video hours. Buyers who are solving an expensive problem are not comparing you to a cheap course; they are comparing you to living with the problem.
That said, first prices are usually best set slightly low and raised deliberately. A common pattern that works: a pilot cohort at a clearly-labelled founding price, testimonials collected, then a permanent price two or three times higher for everyone after. The founding buyers get a genuine deal, you get proof, and the price rise has an honest story behind it.
Avoid the middle-of-nowhere price — cheap enough to feel disposable, expensive enough that people think about it. Very low prices attract buyers who never open the course, which hurts your completion rate and your testimonials. If you want a cheap entry point, make a small, genuinely small product rather than discounting the real one.
Payment plans raise conversion on higher-priced courses and raise your admin slightly. Two or three instalments is usually enough; anything longer creates chasing.
Where to host and sell an online course
You need four things: somewhere the video lives, a page that sells the course, a checkout that takes payment, and reliable delivery to buyers. A dedicated course platform bundles all four plus progress tracking and drip scheduling, and charges you monthly or takes a cut. It is a good fit if you plan to run a large curriculum with many students.
The lighter route is to host video with an unlisted player, sell through a storefront page with a payment form, and deliver the lesson links automatically on purchase. This is how a surprising number of profitable courses actually run, and it is the fastest way to sell an online course without a website. The trade-off is that you manage the structure yourself instead of getting it from a template.
Whichever you pick, judge it on three unglamorous things: does the checkout work on a phone, does the buyer get access instantly without you touching anything, and can you email all your buyers at once. Those decide the customer experience far more than the feature list does.
- Video hosting that allows unlisted or private embedding
- A sales page you can edit without a developer
- A checkout that works on mobile and supports your currency
- Automatic delivery on payment — no manual sending
- A way to email every buyer, and every buyer of one specific course
- A refund route that does not require support tickets
The sales page that does the selling for you
A course sales page has one job: to make a specific person feel understood and then show them exactly what they get. Open with the problem in their language, not with your bio. The bio matters, but it belongs after you have earned the right to be believed.
Then answer, in order, the questions every buyer has: who is this for, what will I be able to do afterwards, what exactly do I get, how long will it take me, who are you, what if it does not work, and how do I buy. Almost every under-performing course page is missing three of those seven, and the missing ones are usually what exactly do I get and what if it does not work.
Show the course, do not just describe it. A module list with lesson titles, a screenshot of the templates, a two-minute preview lesson — these reduce the buyer's risk more than another paragraph of persuasion. Concrete beats convincing.
How to launch an online course for the first time
A launch is not a single announcement; it is a sequence with a deadline. The deadline is what converts, because most people who intend to buy will buy at the end. Without one, your sales page becomes a permanent maybe.
A workable first launch runs about seven to ten days. Before it opens, warm the audience for a week or two with content that touches the same problem — not teasers, actual useful material. Then open with a clear announcement, follow with an email or post that handles the biggest objection, share proof mid-window, and close with an honest final reminder that says when the doors shut and what happens after.
Sell to the people closest to you first. Your email list, your past clients, the people who replied to your posts, the waitlist. Cold traffic is a later problem; the first launch exists to prove the course works and to produce testimonials for the second one.
- Week minus two: publish free content on the same problem, open a waitlist
- Day one: announce, explain who it is for, link the page
- Day three: answer the biggest objection honestly
- Day five: share a preview lesson or a student result
- Day seven: last call with the exact closing time
- After: thank buyers, and tell non-buyers what happens next
Supporting students without living in your inbox
Support expectations are set by your sales page, so set them there deliberately. If the course includes email support, say what response time you offer. If it does not include support, say that plainly and price accordingly. Ambiguity here is what turns a good course into an unpaid job.
Most support requests are structural rather than personal: a link that is unclear, a step that assumed knowledge, a template that opened wrong. Log every question you receive for the first two cohorts. When the same question appears three times, the answer belongs in the course, not in your replies.
Group formats scale better than one-to-one. A monthly live call, an occasional office-hours session, or a simple community thread handles the same volume of questions in a fraction of the time, and students often answer each other. If you want the higher-touch version, sell it separately as a coaching upgrade rather than bundling it in.
Refunds, guarantees and the awkward parts
A clear refund policy sells more than it costs. Buyers are risking money on someone they have never met, and a plain guarantee — thirty days, no argument — removes the largest objection on the page. Refund rates for honest courses tend to stay low, and the sales you gain generally outweigh the ones you return.
Be equally plain about what the course is not. If it will not work without existing tools, an audience, or a certain amount of time per week, say so before purchase. The refund you avoid by being honest costs you nothing; the refund you cause by overselling costs you a review and a referral.
Handle the rare bad-faith case without drama. Refund, remove access, move on. It is cheaper than the argument, and nothing on your sales page has to change.
The numbers worth watching after launch
Four numbers tell you almost everything: how many people saw the sales page, how many bought, how many finished at least the first module, and how many asked for a refund. Conversion tells you whether the page and price fit. First-module completion tells you whether the promise and the opening lesson match.
Low traffic with high conversion is a marketing problem and a good one to have — make more of whatever brought those people. High traffic with low conversion is a page or price problem, and the page is usually the culprit. High refunds with high completion means the outcome did not land; high refunds with low completion means people bought and never started, which is usually an onboarding failure.
Ask every buyer one question a week after purchase: what nearly stopped you from buying? The answers rewrite your sales page for you, in the buyer's own words, which is the only voice that reliably converts other buyers.
Version two: why the second course sells better
Treat the first run as a paid draft. You now have real questions, real drop-off points, real testimonials and real language. Re-record the two lessons that generated the most confusion, add the resource everybody asked for, tighten the module that dragged, and raise the price with a straight face.
Then decide what the course leads to. A course is often a middle product, not an end: it can feed a coaching offer, a template pack, a community, or done-for-you work. Buyers who finished and got the result are the warmest audience you will ever have, and most creators never sell them anything again.
Keep the whole thing in one place — the sales page, the payment, the buyer list, the follow-up emails. When your course, checkout and audience live in five different tools, the friction is not just yours; buyers feel it too. You can build the page, take payments and deliver a course to buyers free at onesol.io.