Because most buyers never watch it to the end. By the numbers I know, only 3 to 12 out of 100 buyers actually finish an online course, and only 48 even start. Anyone selling knowledge in 2026 should plan it live, because that is where people actually show up.
I specialised in online courses for a long time and called myself an online course ghostwriter, the way there are ghostwriters for books. I built courses for entrepreneurs with a process that takes little effort from the client, so I know these numbers from practice. Today I run my AI agency from Dubai and work with entrepreneurs in the German speaking market. I let the course business run out at the beginning of the year, because I saw that the number of online course clients was shrinking. Below are the numbers behind that and the reason a course can still bring in leads. After that I describe how I built courses for busy entrepreneurs and why I put my own offer on live sessions.
Of 100 buyers 48 start, and only 3 to 12 see the course through
I have a strong opinion on this: even though online courses are in my blood, I would never make a recorded course again myself. How many out of 100 buyers really watch a course to the end? My answer is 3 to 12 percent, and I know these numbers from my own work. The number of people who start at all is even starker, because it is only 48 percent, so half do not even begin. That is the rule, and it is why I left this business. The dream of recording once and selling for years was still a thing during Covid or just after it, and then online courses went downhill. In practice that dream no longer exists in this form, and I saw this happen in my own business.
Selling a recorded course means selling a name, a landing page and a promise
If half of the buyers never even start, the content is not what gets sold. What gets sold is the name, the landing page and the promise. The few percent who finish a course need hardly any attention. Some buyers pay several thousand euros and still do not open the course, and in my experience that is the rule. Why we really made online courses was lead generation: if someone watched the course it was a bonus, and we used the course as a paid lead magnet. On top of that, we now get knowledge with one question to Claude, which is why I see everything recorded dying out. I no longer think anything of online courses in 2026 and would not advise anyone to make one, even though I built them myself.
Entrepreneurs invested 4 to 6 hours in their course instead of 40
My clients were mainly entrepreneurs, and I have a very good connection to companies, better than to self-employed people. For them it made sense to invest 4 to 6 hours in an online course instead of 40, because their hourly rate is worth so much. For self-employed people an offer where they build the course together with me fitted better. For busy entrepreneurs I simply made it for them. Over the years I developed a process that draws the knowledge out of the client with little effort: the table of contents from an onboarding form, then a questionnaire in interview style, then the script in the client's voice, and finally the slides he could record. The offer fitted well, and the industries ranged from health to real estate.
Live is the new gold, and without a date in the calendar nothing happens
When I saw that online course clients were becoming fewer, I faced two ways: push the course area once more with sheer effort, or start something new right away. I let it run out at the beginning of the year and began with AI. I did not want to cling to a business that was getting weaker. Live is simply the new gold for me. I am planning a bootcamp for entrepreneurs around Claude myself, and I am even considering not handing out the recording at all. If you have to be there live, you commit, because as I see it: "If it is not in the calendar, then you do not do it." I am an exception here, since I like watching recordings. I watch them like a podcast, while brushing my teeth or eating breakfast.
Summary: anyone selling knowledge in 2026 plans a live session
So why do I no longer recommend a recorded online course? Because only 3 to 12 out of 100 buyers actually watch it to the end, only 48 even start, and the content is quick to get with one question to Claude anyway. A course then sells a name, a landing page and a promise, and it works best as a paid lead magnet, the way we used it back then. Even my clients, the entrepreneurs, invested only 4 to 6 hours in their course, because their time was worth too much. For my own bootcamp on Claude I am considering not releasing the recording at all. If you want to pass on knowledge in 2026, ask yourself before the first day of recording whether the same effect can be reached in a live session that your participants put firmly in their calendar.