Most online courses never sell because the experts behind them spend months building and almost no time selling. We've watched this since 2010 at Kajabi, where experts have earned more than $11.7 billion, and the pattern is blunt: among experts who make at least one sale, the typical first sale lands 67 days after signup, and 30% get there inside a month. The people who sell start selling almost immediately. Meanwhile, across 36.2 million customer relationships on our platform that began with something free, 4.0% ever turned into a purchase... so a course launched to a tiny audience fails on arithmetic before anyone gets around to judging the content.
That's the diagnosis. The rest of this is the autopsy, the math, and the fix.
The autopsy always looks the same
We've watched thousands of course launches from the platform side. When one dies, the story barely changes.
An expert with real knowledge... years of client work, actual results, the kind of experience people already pay for... decides to package it. Four months go into curriculum. Sixty videos get recorded. The intro animation gets redone twice. The welcome sequence gets a ninth rewrite before a single stranger has joined the list it will send to.
Then the launch. To an email list that would fit around a Thanksgiving table.
Three sales. Two are friends.
Here's the part that stings: the course was usually good. From where we sit we can see completion rates, refund rates, reviews. The content holds up fine. Open the expert's calendar instead and you find the actual cause of death: 120 days of building, zero days of audience-building. The product got finished. The business never got started.
"Build it and they will come" is the priciest tuition this industry charges, and experts keep paying it in four-month installments. (The passive-income ads that sell this dream deserve some blame here. Historically, ours included.)
Is marketing actually 80% of the job?
Yes. And for year one, we'd call that conservative.
A course business is a marketing operation that happens to ship a course. Expect to spend roughly four hours promoting for every hour creating, and expect that ratio to hold long after launch. Self-paced courses get almost no help from word-of-mouth: people rave to their friends about their coach or their community, and say nothing about the video library they bought in March. So every sale has to be manufactured. An audience built, trust earned, an offer made. Then made again.
"But I'm a teacher, not a marketer."
Then a self-paced course may honestly be the wrong container for your expertise. A course is one of five models for selling what you know... coaching, community, consulting, and certification are the others (and a newsletter makes a fine front door to any of them)... and each carries a different marketing load. Coaching and community businesses spread through referrals and relationships in ways courses never will, and choosing among the five business models with your appetite for marketing in mind beats forcing yourself into the heaviest one. There's no shame in that trade. There's a lot of shame in recording 60 videos and then discovering you hate the other 80% of the job.
What the data says about who actually sells
Two numbers from our platform, and they tell one story.
The typical first sale takes 67 days. Among experts on Kajabi who make at least one sale, the one in the dead middle (line everyone up, fastest to slowest) got there 67 days after signing up. 30% did it inside 30 days. 58% inside 90.
Sit with what that implies. The experts who ever sell anything mostly sell within their first two or three months. They put an offer in front of people while the course was still ugly, unfinished, or in some cases nonexistent. The sellers sell early. The perfectionists mostly never sell at all.
Free-to-paid conversion runs 4.0%. We counted 36.2 million customer relationships on Kajabi that started with something free... a lead magnet, a freebie, a $0 opt-in. 4.0% of them later bought something. That's what a functioning free-to-paid engine looks like at scale: 1 in 25. As far as we know, nobody else can print that benchmark, which is exactly why so many experts plan their launch around imaginary conversion rates instead.
Now run the math almost nobody runs before building:
Say your course is $500 and you want $50,000 from it. That's 100 buyers. At a 2% launch conversion rate on an email list (a common planning assumption, and a reasonable one), 100 buyers takes roughly 5,000 subscribers. And the 4% benchmark says that as you build that list with free content, the overwhelming majority of people you attract will never buy anything, no matter how good the paid thing is. The funnel is leaky by nature. You beat it with volume and time, and you can't manufacture either one in launch week.
That math is why "my course flopped" is usually a misdiagnosis. The course did its job for the twelve people who saw the sales page. The audience was missing. A great course launched to 150 people performs like a great concert played in an empty parking lot.
"Okay, but my topic is different. Once people find it, it'll spread."
We hear a version of this weekly, and we understand the hope. But we've had a front-row seat since 2010 to experts trying to sell knowledge in every niche you can imagine (and several hundred you can't), and topic has never rescued a launch with no audience behind it. Buyers arrive through distribution you built, and by no other route. The experts in the weirdest, smallest niches who win are the ones who did the audience work first, same as everyone else. Your topic buys you attention once someone is looking. It does none of the looking for you.
Why building feels safer than selling
Because building is private and selling is exposed.
Inside your course builder, you're productive every single day. Polish the slides, re-record module four, adjust the thumbnail colors. Nobody rejects you in there. It counts as work because it is work... it's just work with zero customer acquisition attached to it.
Marketing means asking strangers for attention and mostly getting ignored, which is its own specific flavor of miserable. So experts hide in production. The course gets a professional intro video while the email list sits at zero. And courses only make the hiding easiest: the consultant keeps tightening the deck, the certification builder rewrites the exam for a fifth time before a single candidate applies, the coach refines an intake questionnaire for clients she hasn't gone looking for. Same disease, different container. We've come to read heavy polish on an unlaunched course as a warning sign... the business equivalent of cleaning your room to avoid studying for the exam.
Perfectionism gets billed as quality. Most of the time it's fear with better lighting.
What the experts who sell do differently
Three moves, in order. (For the full walkthrough of the model end to end, see our complete guide to creating and selling online courses, coming soon.)
They build an audience before they build a product. Weeks or months before there's anything to buy, they're publishing genuinely useful things for one specific kind of person and collecting email addresses. Email matters more than follower counts here, because you can reach a subscriber directly instead of renting access from a platform feed. We laid out the math in [INTERNAL LINK: "why 2,000 email subscribers beat 10,000 followers" -> email-subscribers-beat-followers].
They sell before they finish building. Pre-selling an unbuilt course sounds scandalous the first time you hear it and obvious forever after: money is the only validation signal that can't lie to you. Announce a start date, take the money, and stay ahead of your first cohort as you build. If nobody buys, you just saved yourself four months. We wrote up the whole play in [INTERNAL LINK: "how to sell your course before you build it" -> sell-your-course-before-you-build-it].
They ship small. A tight course that delivers one clear result beats a 200-lesson monument that overwhelms everyone who enters it. Less time in production also means more weeks of the year spent on the 80% job. Here's [INTERNAL LINK: "the minimum viable course" -> minimum-viable-course] if you want the spec.
Notice what's missing from that list: better cameras, better editing, better fonts. We've seen courses cross seven figures with footage shot at a kitchen table on a laptop webcam. We've also seen cinematic productions that never cleared ten sales. Production value shows up nowhere in the failure data. The calendar shows up everywhere.
The calendar test
Pull up your calendar and count two numbers for your course project, current or planned: days spent making the thing, and days spent putting your ideas or an offer in front of strangers.
If the first number is 90 and the second is 0, you already know how the launch goes. And the fix costs nothing: invert the calendar. Start the audience work on day one... publish, collect emails, float the offer... and let the course get built in the gaps. The data says your buyers show up for you long before your curriculum is done. Around day 67, typically. Be selling something by then.
If you're sitting on a half-built course right now, read how to sell your course before you build it next. It will either save the project or save you four months. Both outcomes beat launching to a parking lot.
