The instinct when you set out to create a cohort-based course is to build the course first: all the sessions, all the material, finished, before anyone enrols.
You cannot, and the attempt is what kills most first cohorts before they start. The first run is a research project you charge money for. What has to be finished in advance is the arc. The material gets made a week at a time, while it is happening.
Build the arc completely, the material barely
Two things are easy to confuse. The arc is the path: the end state, and what each week delivers on the way there. The material is the slides, the exercises, the recordings.
When you create a cohort-based course, the arc has to be complete before you sell anything, because the arc is what you are selling. The material does not, because until real people are in the room you are guessing about which parts they need.
Everything below follows from that split.
Eight weeks out: write the weekly waypoints
One line per week, in this exact form: by the end of week three, each person will have ______.
Something they possess, not something they understand. A priced offer. A written positioning statement. A completed audit. A recorded pitch.
If you cannot fill in every line, you are not ready to sell, because that list of lines is your sales page. It converts better than any description of content, since it describes outcomes in sequence and the buyer can see themselves arriving.
Decide one more thing here: roughly how much of each week is you teaching a planned topic, and how much is you working on participants' own situations. Both are fine. Unplanned mixing is what makes sessions feel shapeless.
Six weeks out: fix the dates, then open sales
Dates before copy. Put every session in your calendar, including the two you will later be tempted to move, and treat them as immovable from now on.
Then start selling. Six weeks is enough runway for a first run sold to people who already know you. It is not enough for an audience that does not exist yet, so if you are starting cold, extend the runway and plan a smaller first cohort rather than a later one.
One decision to make now rather than later: the minimum number of people at which you will still run it. Write the number down. Three days before the start, with four people enrolled, is the worst possible moment to be deciding this, and it is how people end up delivering a cohort they resent for six weeks.
Two weeks out: build week one and nothing else
The urge to build everything will be strong. Build three things.
Week one's session, properly. This one deserves real preparation.
The pre-work. Twenty to thirty minutes of something, due before the first session.
The welcome message. What to expect, when, and what you need from them.
The pre-work matters far more than its size suggests, and it does two jobs. It tells you who is genuinely engaged before you have invested six weeks, because the people who skip pre-work are the people who will drift in week three. And it turns session one from a lecture into a conversation, because everyone arrives having already done something.
Week one: the session that sets everything
Almost all of a cohort's behaviour is established in the first ninety minutes, and very little of it is established by your content.
Have everyone speak inside the first fifteen minutes. Anyone silent through session one tends to stay silent for the whole run. Thirty seconds each is enough.
Show the entire arc. All the waypoints, in order, so people can see the shape of where they are going. A programme whose shape you cannot picture feels endless.
Have each person state their commitment out loud. One sentence about what they intend to walk away with.
End with the first piece of work and a real deadline. Something due before session two, visible to the group.
Deliberately teach less than you want to in week one. The job of the first session is to establish that this is a thing people participate in rather than watch.
The middle: plan for the sag
Attendance and energy drop in the middle third of every cohort that has ever run, including excellent ones. Treating it as a signal that something is wrong leads to exactly the wrong response.
Three things that help, and one that does not.
Put a deliberately lighter week in the middle. Less to do, not more. The sag is usually accumulated workload rather than lost interest.
Contact anyone who misses a session the same day. One line, privately, no guilt in it. This single habit does more for completion than any content improvement, because people who feel noticed come back and people who slip away quietly do not.
Show the group a win from inside the group. One participant's actual result, shared with their permission. Peer proof lands differently from encouragement.
What does not help is adding material. The instinct when a cohort sags is to give people more value, and more value in the middle third accelerates the drop rather than slowing it.
The final week: harvest rather than conclude
Most final sessions get spent summarising, which is the least valuable thing you could do with them.
Have each person show what they produced. This is where your testimonials come from, and they will be specific in a way that no feedback form produces.
Collect the artefacts. With permission. These sell cohort two better than anything you can write.
Sell the next thing while the group still exists. An alumni tier, the advanced version, individual work for the two or three who want depth. The final session is the single highest-conversion moment you will have with these people, and three weeks later it is gone.
Ask where they nearly dropped out. Not "how was it". That single question produces the most useful information you will get all run, and the answer usually points at a week you thought went fine.
Why cohort two costs a fraction of cohort one
This is the part worth holding onto while the first run feels rough.
After one delivery you have recordings of every session, the real curriculum rather than your guess at it, testimonials with specifics in them, an arc that has survived contact with actual people, and a waitlist of those who could not make the dates.
The second time you create a cohort-based course from the same material, it takes a fraction of the effort and can be priced higher, because the first-run discount has done its job. Cohort three is where it becomes a product rather than an event.
The first run is not supposed to be polished. It is supposed to produce the asset.
How to create a cohort-based course on SKEP
The sequence above has a practical requirement: the dates, the pre-work, the live sessions, the group and the credential all need to exist in the same place before week one, and they need to not interfere with the cohort you run after this one.
On SKEP, once the dates are fixed you publish the cohort as a product and share one link, so sales can open while the material does not exist yet, which is exactly the order this post argues for.
Pre-work and any recorded segments sit in course modules, so your live hours stay for the things that need you present. Sessions run as cohort sessions with the video built in on your fixed slot.
Each cohort gets its own space, which is what makes the next run clean: the February group's conversation does not bleed into April's, and people can see who is in the room with them. Checkpoints and the end-of-run credential are part of the same product. And when you make the alumni offer in the final session, member pricing lets you give graduates a genuine reason to stay rather than a discount code you have to remember to issue.
Try this yourself
1. Write the weekly waypoints before anything else. One line per week, each naming something the person will possess. This is both your design document and your sales page.
2. Put the sessions in your calendar today, then decide your minimum viable number. Both before you write a word of marketing copy.
3. Build week one and the pre-work only. Then stop building and start selling. Everything else gets made the week before it is needed.
4. Diarise a check-in for the middle week now. You will not remember to plan for the sag once you are inside the run.
To create a cohort-based course, finish the arc and leave the material unfinished. Sell the sequence, deliver it one week at a time, and let the questions people ask become the curriculum you could not have written in advance.
The first run buys you the asset. Run it before you feel ready, because feeling ready is not a stage that arrives.
Start building it. SKEP is free to begin. No card needed, and everything in this post runs on it. See the plans for coaches and trainers and pick the one that matches where you are today.
© SKEP
The Launchpad for coaches, trainers, consultants & subject matter experts
support@skep.in



