Loading page…
A course community that stopped sending its students elsewhere to sell and built the checkout itself — a digital-products-only marketplace that did ฿731,000 of GMV in its first 90 days, seeded entirely by people who had already paid for the course
top seller ฿477,981 in 90 days, second ฿215,550 · course sells at ฿5,990 to a community of nearly 2,000 · AI tools at ฿499/mo
Most creators teach people to build a product and then hand them off to someone else's checkout. He built the checkout, bundled it free into the ฿5,990 course, and paid sellers out every Monday — ฿731,000 of GMV in 90 days from an audience he had already sold to once.
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
The operator behind this is a Thai marketer who went solo around 2014–2015, back when running a one-person online business meant buying ads and selling products yourself. He trained in IT, worked in marketing, and spent ten to fifteen years as a consultant and coach to business owners. He started using AI in 2022, and he is unusually honest about the timeline: 2022 and 2023 were trial and error — try something, decide it is not quite right, disappear for three months, come back when a new tool appears. The turning point where AI actually produced income was 2025. Four years, two of them spent flailing. Today the business has three layers. The teaching layer is Khlang Saeng, a course community priced at ฿5,990 with close to 2,000 members, organised around a framework of Create, then Work, then Build, and a course line called The Machine Series — AI Traffic Machine, AI Avatar Machine, AI Money Machine. The tools layer is Hero AI Engine, now on version 3, for image generation, and Hero AI Creator Studio for editing clips with subtitles, voice, music and b-roll, sold at ฿499 a month and given free for a year to course members. The third layer is the one with the numbers: Pizza, a marketplace exclusively for digital products, which Khlang Saeng members can use to open a store and sell. Sign-up is free, sellers get paid out every Monday, and in its first 90 days it did roughly ฿731,000 of gross merchandise value. He says he was surprised by it himself.
Pizza never had a cold start, and that is the whole design. The marketplace was launched into a paid community of nearly 2,000 people who had already spent ฿5,990 to learn how to make digital products and then had nowhere obvious to sell them.
The quick-reference card — acquisition channels, replication playbook, risks & traps, and our verdict — is members-only.
Join the Unicorn Club — $5/moCancel anytime, in one clickPaste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea — a full build spec and an SEO growth plan — plus 10 tailored build specs a week. Members only.
Join the Unicorn Club — $5/moCancel anytime, in one clickEvery claim traces back to a listed source — open any of them and check the number yourself.
Data credibility: Founder-reported throughout. The creator is describing his own platform's back-office figures on a live stream and says explicitly that he is showing as much of the internal data as he is able to — nothing is third-party verified and no dashboard can be confirmed from an audio transcript. One point in their favour: the numbers are internally consistent. ฿477,981 plus ฿215,550 plus the third seller's roughly ฿24,750 (25 orders at ฿990) comes to about ฿718,000 against a stated 90-day GMV of roughly ฿731,000, which is what a real top-three board should look like. Figures are left in baht and not converted, because the transcript provides no exchange rate. The country is not stated outright in the segment and is taken here from the currency and language rather than from the group file's region tag. Several names come from a Thai auto-caption and may be spelled differently in reality: the community brand is rendered variously as คลังแสง, คังแสง and คางแสง and is transliterated here as Khlang Saeng; the marketplace is spoken as "Pizza" with a URL that the caption garbles; and the featured veterinary store appears as something close to "BuffNista Academy." A third price point for the ฿100,000 arithmetic is garbled in the caption and is omitted rather than guessed. Community size is given as "nearly 2,000" with no statement that all of them paid the ฿5,990 price, so no revenue is inferred from it.
He does the arithmetic on camera, and it is the most useful thing in the video. To reach the first ฿100,000 you sell 17 people a ฿5,990 course, or 101 people a ฿990 product.
The loop closes because every layer feeds the next one. Members pay ฿5,990 to learn the Create–Work–Build method; they publish content; they open a Pizza store; their sales become the case studies he shows in the next live stream, which sells more memberships, which adds more sellers to the marketplace.
He describes the dead years directly. Four years of using AI, of which roughly two were spent going in circles: try a tool, conclude it is not quite it, vanish for three months, return when something new launches.
Three products that only make sense together. Khlang Saeng is the paid community and course library at ฿5,990, built on a framework of Create, Work, Build, with three courses in what he calls The Machine Series: AI Traffic Machine, which is about finding the content gap before you write anything; AI Avatar Machine, for faceless presenting; and AI Money Machine. Hero AI Engine and Hero AI Creator Studio are the tools — image generation on version 3, and a clip editor that adds subtitles, voice, music and b-roll to footage you upload, whether or not it features an avatar — at ฿499 monthly, free to members. Pizza is a marketplace restricted to digital products: sellers list an ebook, template pack, sheet or course, buyers pay and receive the file immediately, and payouts run every Monday. He describes himself plainly as the intermediary. Our read: the restriction to digital products is the smart part. No inventory, no shipping, no returns logistics, and instant fulfilment means the payout schedule is the only operational commitment he carries. The featured store he shows — a veterinarian selling niche nutrition material for pets with kidney disease — is exactly the profile this works for: a professional with narrow expertise, no e-commerce ability, and no interest in building a storefront.
The disclosed money is on the course and the tools. Khlang Saeng sells at ฿5,990, and he states that 17 sales clear ฿100,000; the community sits at close to 2,000 people, though he never says all of them paid that price, and a ฿990 tier is discussed as an alternative price point. Hero AI Creator Studio is ฿499 a month billed monthly or ฿480 a month on the annual plan, free for a year for members and free for life for lifetime members, with an image-generation quota above which users buy credits. A founder package at 50% off is capped at the first 100 buyers and he reports somewhere between ten and fourteen sold at the time of the stream. On Pizza he gives GMV, not his own revenue: roughly ฿731,000 across 90 days, with the top store at ฿477,981 at around ฿29,000 per order, the second at ฿215,550, and a member's ฿990 starter set taking 25 orders for about ฿24,000. Our read: no take rate is disclosed anywhere, so the marketplace's contribution to his own income is unknown and should not be guessed at. The more likely economics are indirect — the marketplace is a retention and proof mechanism for a ฿5,990 course, and its GMV is the sales material. That is a legitimate model, but it means the ฿731,000 is his students' money, not his.
This is a one-person operation running three products, which tells you most of what you need about the cost structure. He publishes a weekly live stream, one hour, every Wednesday at 16:00, which is the entire content calendar for the business; he notes that his genuine on-camera appearances amount to under 10% of what is on his channel, with the rest produced. Distribution runs through TikTok, YouTube and Facebook, with a LINE bot named Pinky handling lead capture — viewers message the word for "plan" and receive a PDF, which is a cheap and very effective list-building mechanism in that market. The tools are built in-house rather than resold, which is why they can be given away free to members: Hero AI Engine for images, Creator Studio for editing, Pizza for checkout and payouts. Our read: building your own tools instead of reselling someone else's is the decision that makes this whole structure work economically. A creator who bundles third-party software into a course pays for every member forever; a creator who owns the software has a marginal cost near zero on the bundle and a second revenue line from non-members at ฿499 a month. The trade is that he now carries three products' worth of maintenance alone, and he is explicit that he wants feedback specifically to keep developing the system.
Copy the order of operations, which is his central teaching and also what he actually did. The most common mistake he sees is starting at Build — making the ebook, the course, the system, the storefront — and then discovering nobody is buying, because building is the fun part now that AI makes it possible. Create first until people know you, use AI to Work so you can produce consistently, and only Build once people are already messaging you. Copy the pricing arithmetic too: pick the revenue target, divide by a believable number of buyers, and let that set the price rather than guessing at a number. And copy the structural move — if you teach people to make something, owning the place they sell it is worth more than the affiliate commission you would earn sending them elsewhere, because it converts your students' success into your marketing. Change the timeline expectation. He took four years, two of them unproductive, and the marketplace worked in 90 days only because 2,000 people had already paid him. Our read: without an existing audience this is a two-to-three-year sequence, and anyone starting the marketplace first has the dependency exactly backwards.
Concentration is severe and visible in the numbers. Two stores account for roughly ฿693,500 of the ฿731,000 of 90-day GMV, meaning the marketplace is really two successful sellers plus a long tail. If either leaves — and there is nothing holding them, since a digital seller can move to any checkout in an afternoon — the headline number collapses. The second risk is that the whole structure rests on one person's attention: one weekly live stream, one face, one LINE list. The third is a dependency loop worth naming plainly. Students pay ฿5,990 partly on the strength of case studies drawn from other students, and the most visible case studies are people who sell courses about doing this. A creator-education economy that mostly sells creator education is fragile, and the strongest example he shows — a TikTok channel grown from zero to nearly 300,000 selling an avatar storytelling course, then a second course about monetising the first — sits squarely in that pattern. Our read: the healthiest signal in the whole video is the veterinarian selling pet kidney-disease nutrition and the member with a day job selling a ฿990 professional starter set, because those are outside the loop. Whether the marketplace can keep recruiting sellers like them, rather than more people teaching AI, is the question that decides if this compounds or plateaus.
The four founder playbooks — first customers, first dollar, the flywheel, the silent stretch — are in the Unicorn Club, on this idea and on all 1,403 others. The story, the numbers, the sources and every later section stay open to everyone.
Join the Unicorn Club — $5/moCancel anytime, in one click