Christmas Eve, 2018. I was sitting alone in a hostel, broke, newly single, and thoroughly burnt out. Although I say alone… I did have bed bugs, so technically there was company.
A year earlier I’d released my first online course. Thirty-two hours of it. I was completely convinced it was going to be a huge success, and instead it was more or less the beginning of the end of that particular chapter of my life.
What follows is everything I’ve learned since. I’ve now taught more than 240,000 students across 160 countries, and the thing I’m proudest of isn’t the number… it’s that my average completion rate on Skillshare sits somewhere around 40 to 50%, when the industry average is about 3%.
This is the advice I wish someone had handed me at the start. It would have saved me years of trial and error and tens of thousands of dollars.
The short version
Most online courses fail because the creator focused on selling and hosting instead of teaching. Being an expert in something does not make you good at explaining it, and nobody in the course-creation world teaches you the difference. Fix the teaching first: build around what students will DO rather than what you know, price on the transformation rather than the hour count, and start smaller than feels right.
- Roughly 3% of people finish the average online course. Mine now run at 40 to 50% on Skillshare, and the change came from teaching skills, not marketing skills.
- Adults don’t learn by being told things. They learn by doing, so every section needs something to apply.
- Your price reflects the transformation you deliver, not how many hours of video you’ve crammed in.
- Don’t run ads until you’ve made some organic sales, or you’re just paying to push people into a broken process.
- A good course with bad production is still a good course. A bad course with beautiful production is still a bad course.
You can watch the whole thing below, or keep reading, both cover the same ground. Fair warning, the video is a long and fairly unfiltered one, so it’s a grab-a-cuppa job.
Jump to a bit of the video:
1:08 Teaching skills beat expertise
3:12 No one gets taught how to teach
5:40 Stop doing the easy thing
7:17 Creating bingeable content
9:32 Grown up brains learn differently
13:32 What is your course worth?
17:23 Your slides should make NO sense
19:33 A good course with bad production
21:46 Embrace your weirdness
25:37 Why will you fail?
28:11 Personal reflections
Why doesn’t being brilliant at something make you good at teaching it?
Because teaching is a completely separate skill from doing, and almost nobody trains for it. Expertise tells you what the right answer is. Teaching is about getting that answer into someone else’s head in an order that makes sense to them.
Think about how many genuinely terrible managers you’ve had in the corporate world. They usually got promoted because they’d been there a long time and they were excellent at the actual job… and then nobody taught them how to lead people. Same pattern, different setting.
In the course world this shows up as very famous people selling very expensive courses with almost no substance in them.
Here’s my honest position on myself. I’m not the best marketer in the world. I wouldn’t even claim to be a great one. But for this particular job I am much better off being a good marketer and a great teacher than the other way round. If you have to be world class at one of the two, pick teaching.
Why does nobody teach course creators how to teach?
Because it isn’t what anybody searches for. When you decide to make a course you Google “how do I sell my course” and “which platform should I host my course on”. You do not Google “how do I become a good teacher”.
And that’s the gap, because look at what you can and can’t outsource. You can hire someone to market your course. You can hire an editor. You can hire a strategist, a designer, a funnel person. You cannot hire someone to make your course engaging and watchable, because that comes from you and your stories.
Worth saying too: being good on YouTube doesn’t automatically make you good at courses. They’re different formats with completely different jobs.
So I invested properly in learning to teach, and that one decision is what moved my completion rate from around the 3% average to 40 to 50%. On courses that in some cases run to twelve hours.
What’s the easy thing you should stop doing?
Whatever you’re avoiding, honestly… for me it was two things: getting on camera, and doing the work of connecting my lessons together.
For years I hid behind slides and a voiceover because I wasn’t confident enough to be on screen. And I skipped the small linking work between lessons, the “here’s what’s coming next and why it matters” moments, because it’s fiddly and nobody makes you do it.
Both of those are the easy option. And the easy option is almost never the thing that has the biggest impact on your students.
There’s a related one worth naming: self-promotion versus idea promotion. I used to run webinars that were basically about how great I was… nobody cares. It’s not about you, it’s about them and what they’re going to be able to do afterwards.
How do you make a course people actually finish?
Give them a reason to start the next lesson before they’ve finished the current one, the same way a good series does it.
Two things I now do without exception. At the end of every lesson, I tease what’s coming in the next one. At the end of every section, I recap what they’ve just learned so they can feel the progress.
It sounds almost too simple to matter… and it matters enormously. Completion is not really about willpower, it’s about momentum, and momentum is something you design in rather than hope for.
How do adults actually learn?
By doing, not by being told. This one has a proper name, andragogy, and it’s the single biggest difference between teaching adults and teaching children.
A child will sit through a lesson on something that seems completely pointless because an adult told them to. An adult will not. An adult wants to know what they can do with this, today.
I got this wrong for years. I’d cram my courses full of information and assume that delivering it was the same as teaching it. But watching a hundred hours of cooking shows will not make you a good cook. Stop teaching what you know and start teaching what you do.
What changed it for me was Skillshare, oddly. Every class there has a project component, which was designed for creative subjects. Being forced to build a project into a marketing course made me completely rethink the structure. Now every one of my courses has practical exercises after each section, and it has been a genuine game-changer for both engagement and results.
What should you actually charge for a course?
Price on the transformation you deliver, not the amount of information you’ve included. This is the one I see people get wrong most often.
My biggest early mistake was assuming that longer meant more valuable, so I could charge more… which is exactly backwards. A short course that helps someone land their first client is worth considerably more than a thirty-hour course full of fluff.
And here’s the part that surprises people: don’t overstuff your programme with bonus guides, extra workbooks and freebies. It feels generous… what it actually does is delay the outcome, which delays the testimonial, which is the thing you actually need.
One more business note while we’re here, because it doesn’t fit neatly anywhere else. Don’t run paid ads until you’ve made at least a few organic sales. You need to see where people get stuck in your sales process first. Putting money behind a broken funnel just means you’re paying for the privilege of watching people drop off.
Why shouldn’t your slides make sense on their own?
Because if the slide says everything, you’re redundant. If someone can read your slide and get the whole point, they don’t need to listen to you… and at that point you’ve made a PDF with a voiceover.
I used to pack my slides absolutely FULL. Every point, every sub-point, all of it up there in eleven-point type.
Look at how Seth Godin does it. His slides make almost no sense without him standing next to them, because they’re there to support what he’s saying rather than to replace it. That’s the goal.
It has genuinely taken me years to get slides right, and I still tinker. If a proper tutorial on this would help, let me know and I’ll make one.
Does production quality actually matter?
Less than you think, and never more than the teaching. A bad course with beautiful production is still a bad course. A good course with rough production is still a good course.
When I started teaching on Skillshare in 2020, my setup was a webcam, a stack of books to prop it on, and a thirty dollar microphone… that was it. And people took those courses, finished them, and left reviews.
So please don’t let the gear be the reason you haven’t started. Improve the production later, once you know the course works.
Should you teach in your own style, even if it’s a bit much?
Yes, because you’re going to do it anyway. You will teach the way you like to learn, and there’s no point fighting it.
I’m a visual, emoji-loving, story-telling sort of person, so my courses are full of colour and GIFs and daft asides. I need the fluff to function. My husband is all business and would teach in a completely different way, and his students would love him for it.
Not everyone will love your style. That’s fine, because you’re not trying to be for everyone… you’re trying to be unmistakably for your people. And the ones who like how you explain things will stay for every course you ever make.
What’s the question to ask before you build anything?
“Why might this fail?”
Not “why will this succeed”, which is the question we all naturally ask when we’re excited. Deliberately go looking for the ways it could go wrong, and you’ll fix half of them before you’ve filmed a thing.
I never asked it about my first course. I was SO convinced it was going to be huge that I didn’t leave any room for the possibility that it wouldn’t be.
So what actually went wrong with that first course?
Four things, and they compounded.
- It was thirty-two hours long. Far too much for a beginner to ever finish, which meant almost nobody got a result from it.
- The product-market fit was wrong. I built it for aspiring digital nomads, but most of them wanted the lifestyle, not the graft of learning digital marketing. And honestly, who can blame them.
- There were no actionable exercises. Just information, endlessly delivered.
- I ran it on a launch-and-close model, which meant a huge push, then a crash, then another huge push. That’s the bit that burnt me out.
If I were starting again I’d think about it like learning to drive. You don’t go from walking straight to driving on a motorway. You watch other people drive, then you drive with someone beside you for a year, then you go out on a restricted licence, and only then do you get the full thing.
Courses should work the same way. Make something small. Learn from it. Iterate. Then grow.
One more thing, on empathy
You genuinely never know what someone is dealing with.
My first few courses had slides that were completely inaccessible, and I didn’t even know to check. That’s on me and I’ve fixed it since.
But the bigger lesson came from students telling me why they hadn’t finished. A sick child. A parent needing care. A job falling apart. Reminder emails to keep going are fine and I still send them, but it has been genuinely humbling to be reminded that a course sits somewhere quite far down most people’s list of urgent things, and that’s completely as it should be.
Where would I start if I were you?
Make something small, on a topic you can genuinely teach rather than just perform, with something for the student to DO after every section. Then ask a few people what confused them, and fix that.
That’s it. That’s the whole thing, and it took me about eight years and one very memorable Christmas to work it out.
If you’re weighing up where to actually put your course, I’ve written a full breakdown of whether Skillshare and Udemy are worth it, with my real earnings figures in it.
And if you’re staring at a blank page wondering what to make at all, my free personalised marketing plan will give you a proper starting point built around your actual business.
Trust your gut on this one. Your first course won’t be perfect, and mine certainly wasn’t. Just make it smaller than you think it should be, and start.
Originally published December 2024, updated August 2026.
