Selling a certification course means selling your vouch. A credential is worth exactly what the name behind it is worth, and when you certify a graduate you stake your reputation on their competence... that transfer of trust is the thing they're actually buying, and the thing you have to protect. AI has pushed the price of the information inside most certification programs toward zero, which leaves the vouch as the scarce ingredient: a machine can teach your material, and it has no name to put on the line when your graduate goes to practice it. So the honest answer to how you sell one: prove the method across real clients, write it down completely, price the credential against what it changes for your graduates (the category runs from about $1,000 to $25,000 and up), and defend the standard like it's your own reputation. Because it is. Since 2010 we've watched hundreds of thousands of experts build businesses here (more than $11.7 billion earned by experts on Kajabi so far), and the certifiers who last are the ones who treat that defense as the whole job.
This post is part of a series on what expert businesses sell now that information stopped being a moat. The short version of that argument: knowing things used to be defensible because knowledge was hard to reach. Now everyone carries a model in their pocket that has read more than any expert alive. When the how-to is free, experts sell the four things AI can't fake: accountability, ownership of outcomes, vouching, and judgment. Certification is the corner of the expert economy where that shift bites hardest, and pays best... and to sell the vouch, you first have to understand what your buyer believes they're purchasing. Start there.
What are you actually buying when you get certified?
A reputation you haven't had time to build.
Think about what the certifier spent to create the thing you're paying for. Years of client outcomes. Public results. Graduates who performed well, in front of people who were watching. Every one of those compounds into a name that means something in a specific field. When that expert certifies you, you get to borrow the name. Their trust becomes your trust, on day one, before you've earned any of your own.
One expert we've followed for years trains end-of-life guides... people who sit with families through the worst week those families will ever have. She stopped doing that sitting herself a long time ago. She built a training program, set a standard, and credentialed the people who now do the work. Her method shows up at bedsides she's never seen, in the hands of practitioners she tested and signed off on.
Here's the part that matters for this post: the families hiring her graduates see the credential, recognize who issued it, and trust the person in front of them because someone they trust already did. The curriculum and the training hours stay invisible the whole time. That's the entire transaction. The training was the cost of the vouch, and the vouch is the product.
Do certifications matter now that AI can teach anything?
More than they did. And precisely because AI can teach anything.
A buyer used to screen for knowledge. Ask a few questions, see if the person sounds like they know their field. That screen is dead. The fakes have read the same material the real practitioners have (faster, honestly), and generated confidence is indistinguishable from earned confidence on a sales call. So buyers moved to the screen that still works: backing. Who trained this person? Who tested them? Whose name is attached to them... and what happens to that name if this person fails me?
That last question is the whole game. A vouch is only worth what the voucher stands to lose.
Run the test on AI. A model can deliver a curriculum, quiz you on it, simulate hard cases, even grade your practice sessions. Genuinely useful, all of it. Then ask what the model loses when its student botches a real client. Nothing. No reputation drops. No future students think twice. No skin, no vouch. Which means every piece of a certification program is now cheap to produce except one, and that one piece requires a human with something at stake.
"So a certification is just a piece of paper."
Some are. Specifically, the ones handed to anyone with a credit card. The name is the product and the paper is the receipt, so when the name behind a credential stops meaning anything, the paper follows it down. That's an argument for picking certifications carefully, and it's the same reason the good ones cost real money.
"Couldn't an AI just issue certificates?"
It can print them today. Nobody cares, for the reason above: a certificate from something that risks nothing certifies nothing. You might as well frame a fortune cookie.
Why would an expert sell a certification?
Because it's the one model that sells the method without selling the expert's hours, including the trust.
Look across the five business models for selling your expertise and certification is the odd one out: every other model keeps the expert in the delivery loop. Coaching needs your calendar. Consulting needs your calendar and your weekends. Courses scale the content but the results still lean on you showing up. Certification packages the whole system... the steps, the judgment calls, the standard... and hands it to other practitioners who deliver it in places you'll never go.
The economics explain why serious experts eventually look at this model. Take a modest program: a $5,000 credential and 50 graduates a year is a $250,000 training business. Now look past the revenue to the reach. Fifty graduates, each working with a hundred clients: that's 5,000 people a year getting the method while the expert who built it is at home revising the curriculum. Pricing across the category runs from about $1,000 at the shallow end to $25,000 and up for deep professional programs, and the spread mostly tracks one variable: how much the credential changes what a graduate can charge.
Fair warning before anyone gets excited: this is an advanced model, and most experts aren't ready for it. You need a method that's proven across many clients, written down well enough that someone who's never met you could run it, and a genuine willingness to police the people who carry your name. Which brings us to the ugly part.
What makes a certification worthless?
Your weakest graduate. They set the market price of your credential, and they set it in public.
Picture it from the buyer's side. Someone pays for a practitioner certified in your method, gets bad work, and tells everyone. The blame lands on the credential, because the credential is what the buyer trusted. The practitioner was just the delivery mechanism. One sloppy graduate can burn years of accumulated vouching, which is why the certifiers who last treat quality control as the actual job:
- A standard that means something. Written down before the first student enrolls. What does a graduate have to be able to do, and prove, to carry your name?
- Real assessment. Exams, observed sessions, case work. If everyone passes, nothing was tested.
- The willingness to say "not yet." To someone who paid you. This is the part that separates certifiers from certificate printers.
- A revocation policy you'll actually use. Pulling a credential from a paying graduate hurts. Watching the market decide your credential means nothing hurts more, and takes longer.
Price is part of the same signal, by the way. Put a $500 tag on your credential and you've told the market what your name is worth. Cheap certifications attract casual practitioners, casual practitioners produce casual results, and casual results price the credential right back down to what you charged for it. The loop closes fast.
The blunt way to say all of this: a certification program where everyone passes is a merch business. The revenue is real. The vouch isn't.
How do you know if your methodology is certifiable?
If you're an expert reading this and doing math on your own method, here's the honest screen. A methodology can carry a credential when it passes four tests:
- You can write it down. The steps, the order they run in, and the calls a practitioner has to make along the way... captured completely enough that someone new could work from the page.
- You can teach it to someone wired differently than you. Your gut doesn't transfer. Principles do.
- You can test readiness. Some observable bar that separates "took the training" from "ready for real clients."
- Other people get comparable results without you in the room. Comparable, not carbon copies. If results collapse the moment you leave, you have a personal practice, and that's a different business.
And the disqualifier, which we mean as a description rather than an insult: "My method won't survive a manual. I read the room and respond." Plenty of brilliant practitioners work exactly this way, off signals they'd struggle to describe. That gift monetizes beautifully through coaching and consulting, where your instinct does the work directly. But you can't put your name on what you can't inspect. Certifiable methods are inspectable methods, and turning instinct into an inspectable method is usually the longest part of building a program.
Methodology Certifiability Test: a 12-question self-assessment that pressure-tests your method across four areas (methodology, market, business, quality) and tells you plainly whether you're ready to certify others, close but missing pieces, or better served by a model that uses your intuition as-is.
The thing machines can't sell
The pattern across this whole series is that AI ate the parts of expertise that were information and left the parts that were human. Vouching might be the purest case. A reputation takes years to earn, seconds to spend, and can't be generated at any price... which is exactly what makes it worth buying, and worth selling with care.
Vouching is one of a handful of durable goods left on the shelf. The others we've covered so far: accountability, the thing coaches actually sell and judgment, the filter buyers pay experts for.
AI will keep getting better at teaching. Names still have to be earned. Spend yours carefully.
