How to build an online course from scratch: structure before filming

The question always arrives in roughly the same shape: I filmed hours of video over a few weekends, I have opened my academy dashboard, so where do I start organising it? The honest answer is that organising does not come after filming. Organising is what decides what you film.

The most expensive mistake in building an online course is not a cheap camera or bad lighting. It is pressing record before you have a written structure. Every change to the structure after filming means filming again, and every reshoot pushes the launch further out and burns some of the enthusiasm you started with.

This article is about the work that comes before the camera: turning what you know into a path with a beginning, an end and proof of arrival, and laying that path out inside the platform before you record a single minute.

Why filming first lets you down

Start with recording and the structure gets shaped by the footage instead of the footage being shaped by the structure. You end up with a forty minute video carrying five ideas. Your learner opens it, cannot tell where to begin, puts it off once, and never comes back.

You lose three things at once:

• Visibility. One long file is a single progress point: watched or not watched. A separate lesson records its own progress, so you know exactly which lesson people stop at.
• Version control. Every revision spawns another copy of the same file, until you are staring at a folder wondering which one is current.
• Your evenings. Every update means a new link sent to everybody, and whoever missed the email stays on an old copy forever.

Notice that the material is already yours: old recordings, a deck, exercise files, written notes. What is missing is not content. What is missing is order. That is good news, because ordering is desk work you finish in two sittings, while filming from scratch is a project measured in weeks.

Start from the transformation, not from what you know

Most trainers open with the wrong question: what do I know? It produces a long course with no clear ending, because what you know never runs out, so the course ends when you get tired rather than when your learner arrives.

The useful question is this: the learner who pays today, what will they be able to do in six weeks that they cannot do now?

Write the answer as one sentence in this shape: I take this person from here to there, and the proof they arrived is this concrete piece of work. For example, I take an employee with no commercial account to a storefront that has taken its first real order, and the proof is a screenshot of that order.

That sentence is your ruler. Every lesson you think of afterwards gets measured against it: does this move the learner closer to that state, or does it simply display what I know? Anything that does not move them leaves the main path or becomes an optional module at the end. That test alone strikes lessons off your list before you film them, which saves you the shooting and the editing that would have followed.

Its second benefit is that it gives your course an ending. A learner who can see where the path leads keeps walking it. A learner working through open-ended material with no finish line stops halfway and cannot say why.

A module is a milestone, a lesson is one idea

The structure that works is almost disappointingly simple: a course is modules, and a module is lessons. All the craft is in where you draw the boundaries.

A module is a milestone the learner can name. Ask them when they finish it what they now know, and they should answer in one clean sentence: I know how to choose a product. If they cannot name it, it was never a milestone, just a pile of lessons under a heading.

A lesson is one idea and one action. The working test is to summarise it in a sentence. If you need two clauses joined by "and", you have two lessons, so split them. Splitting pays twice over: the learner digests it, and you see their progress precisely, because progress is recorded lesson by lesson.

Before you film anything, write four lines for each lesson:

1. The sentence: after this lesson the learner can do X.
2. What I show: the example, screen or case that makes the idea visible.
3. What they do: an exercise, a file they fill in, or a decision they make before moving on.
4. The proof: how I know they actually understood before they continue.

Twenty lessons at four lines each is eighty lines, written in a sitting or two. That is your course, complete, before you have bought a microphone.

Not everything you want to say deserves video

Video is the most expensive thing you make and the hardest to fix. And a lesson is not one single thing anyway: you explain on video, attach the exercise file, write the takeaway, then flag the warning nobody can afford to miss. Choosing the right element is not a matter of taste. It is a decision that saves you hours of filming.

What you need to convey	The right element	Why
A performance or on-screen motion	Video, uploaded or embedded	Words cannot carry it; it has to be watched
Steps they come back to repeatedly	Formatted text	Rereading a line beats scrubbing a timeline
A template or worksheet	A file or PDF	They open it, work in it, and keep it
A comparison or a diagram	An image	Grasped in one look; video scatters it across minutes
A warning they must not miss	A callout	It breaks out of the paragraph instead of sliding past
A long lesson that needs air	A divider	Cuts a heavy stretch into pieces that digest

When you get to uploading, know the limit up front: 50MB per uploaded file. Long video is not meant to be uploaded at all. Embed it from YouTube, Vimeo or Google Drive and it plays inside the lesson exactly as an uploaded file would, using none of your storage. Everything you do upload lands in one media library you reuse across any course or page, so the intro you filmed once never gets filmed twice. Storage itself is 50GB on Basic and 100GB on Pro.

Lay the structure out in the platform before you write the script

Here is the step nearly everyone skips: create the course while you own zero seconds of video. In the course and content builder you create the course, add its modules and then its lessons as titles only, drag everything until the sequence settles, and keep it as a draft.

Then do the thing that exposes everything: preview it as your learner sees it. What you see in the preview is what they actually move through, with the module list to one side, the current lesson in the middle, and completion marked automatically as they go. That is why an empty skeleton seen through their eyes gives away what a notebook outline hides: a module with nine lessons sitting next to one with two, three consecutive titles saying the same thing in different words, a lesson stranded at the end that serves no part of your transformation sentence.

Fix all of that now, while fixing means dragging a title. Fix it after filming and it costs you a reshoot.

After publishing, editing stays direct: any change to a lesson reaches everyone who bought, with no new links to send. You can also publish what is ready and add modules later, and buyers get the new material automatically. A course is not a book printed once.

Who sees what, and when

Once the structure settles a new question appears: does the learner get everything on day one? Access already follows the purchase or a grant you give, never a link passed around. On top of that you can release the content in stages. Drip release unlocks modules as time passes from each learner's own enrolment date rather than one fixed date for the group, which protects the pace of the course and stops today's buyer from taking everything and disappearing.

And because your structure is cut into separate lessons, progress is saved per lesson, so learners resume where they stopped and learner progress reports show you who finished and who stalled halfway. When ten people stall at the same lesson, the problem is the lesson, not the learners.

Be precise about one thing: drip release measures time, not understanding. It does not wait for anyone to complete module one. If you want a gate on comprehension itself, that is a quiz that checks understanding before moving on or an assignment they submit and you grade, and those, along with a certificate issued at completion, start on the Pro plan.

What the builder will not do for you

This section matters more than everything above it, because a trainer who decides on the strength of a half promise finds out the rest at the worst possible moment.

• It will not write your curriculum. It gives you modules, lessons, drag ordering and a preview. It will not tell you that your third module is redundant or that your transformation sentence is vague. That judgement is yours, and it takes two sittings.
• Reports tell you where they stopped, not why. You will know ten people left at lesson seven, and you will not know the reason until you ask them or watch the lesson through their eyes.
• Drip release is time based. It counts from each learner's enrolment date. It is not a completion requirement.
• The upload limit is real. 50MB per file, so long video gets embedded, and an embedded video still lives with its provider, which means you set its sharing options there too.
• Preview shows what the learner sees, not what they feel. Nothing replaces a few real people walking through the first version before you launch.

Four mistakes that keep repeating, and where to start this week

One: filming everything before the first learner. Twenty finished lessons before anybody has bought is a bet of several weeks against a guess. Publish a module or two, watch where people stop, then film the rest knowing what you are filming.

Two: the overstuffed first module. The natural pull is to put everything you know at the front to prove you are worth the price. The result is that your heaviest module sits exactly where motivation is weakest. Make module one the shortest, and end it with a small, concrete win.

Three: waiting for perfect. A course that never ships teaches nobody. Publish what is ready and add modules later; whoever bought gets the new material automatically.

Four: postponing the ordering. "I will upload the files now and organise them later" is the sentence that ends in a course nobody ever organises. Ordering is the cheapest thing you do before filming and the most expensive thing you do after.

So, practically: start with the transformation sentence, break it into milestones, break each milestone into single ideas, then write all of it as empty modules and lessons inside the platform and look at it through your learner's eyes. What remains is filming and editing, and that is the easy part, because you know exactly what to film.

The course and content builder is on every plan, with no cap on how many courses you publish or how many learners enrol, while what sits on top of a course, the quizzes, assignments, certificates and live sessions, starts on Pro. If the structure is not built yet, build it this week: fourteen days free with no card is enough to see your whole course on screen before you shoot a single frame.