Samuel Thompson has launched 100 companies in 10 years.
One of them sells sugar water to hummingbird owners. Another one is a divorce book written by ChatGPT. During COVID he moved $800,000 of a teeth whitening product he found on AliExpress.
None of these are billion dollar ideas. That's the point.
Sam builds the same thing over and over: a one product Shopify store, one offer, Meta ads pointed straight at the product page. When the math works, he scales the spend. When it stops working, he shuts it down and launches the next one.
Greg Isenberg had him on the podcast and made him build one live, start to finish. Ebook, store, copy, ad creatives, all of it. Sam says the whole thing takes about an hour once you know the moves.
I watched the full episode and pulled out the playbook. Here's the video, then the numbers.
The rigged slot machine
Sam's whole philosophy fits in one sentence from the episode:
"All I'm ever trying to do on ecom is create a rigged slot machine where I know I can put a dollar in and get $1.30 out."
He's not building a brand or chasing some three-year LTV model. He finds an offer where cost of traffic times conversion rate times average order value comes out positive, then he feeds that machine as many dollars as it will take.
And when the math stops holding? He quotes it straight: that's why he's done 100 companies. He just goes and does another thing that gives him that return.
He learned this as a 19 year old dropshipper with $1,000 to his name. When that's all you have, you can't afford a marketing angle that doesn't pay for itself. The $800K teeth whitening run during COVID came out of that exact discipline.
→ The takeaway: validate that you can profitably acquire customers before you build anything else. Sam says that's all any entrepreneur is really testing at the start. If you can't acquire customers at all, what are you doing?
The offer: a ChatGPT divorce book
Sam is getting married in a few months. While planning the wedding he noticed the industry is enormous, then looked up the other side of it: roughly half of marriages end in divorce.
Huge market. Terrible competition for attention, because the only people advertising to divorcing couples are family law attorneys.
So he wrote a book. Sort of.
The actual prompt was 30 seconds long: "I want to write a book called The Divorce Bible: How to Win Your Divorce." ChatGPT produced the structure, then he went chapter by chapter with "please write chapter 2." The first version was maybe 70 pages. Just enough to sell.
Once real sales came through, he ran a deep research report, rewrote the prompts with word counts and case study requirements, and the book grew to 197 printed pages.
Then the unglamorous assembly line:
- Paste the chapters into a Google Doc, format at 5x7, export as PDF
- Design the cover in Canva (a stock template with the copy swapped)
- Merge front cover, book, and back cover with a free PDF merge tool
- Pull product mockups from Envato Elements, about $15 a month
That's the product. 300 printed copies showed up at his front door and his fiancée was, in his words, jokingly concerned.
(Launching a divorce book while planning your own wedding is a special kind of founder behavior.)
The funnel math: $29 book, $42 cart, sub-$45 CAC
This is the part most people skip, and it's the only part that matters.
Sam sold the book at $29 with a one-click bump offer at checkout, a $17 or $18 add-on. Enough buyers took it that his average order value landed around $42 to $43.
His cost to acquire a customer on Meta: under $40 on good days, about $45 on bad ones.
So the funnel settled at break even. Which sounds like failure until you hear what he does with it. Break even means he can print customers for free while he pushes the cart value up. That's exactly why he was rebuilding the store on Shopify during the episode: bundle the book into a "starter kit," list it at $58, show a compare-at price around $99, and let the same traffic throw off margin.
His pricing test method is just as blunt. Run $1,000 of ads at one price. Read the metrics. Change the sale label, run it again at another price, and watch what conversion rate does against order value. If you can raise price 20% and conversion only drops 10%, you take the trade every time.
Building the one product Shopify store
Sam doesn't design stores from scratch. He buys conversion-optimized themes built for single product stores and fills in the blanks.
Two he named in the episode:
- Solo Drop, around $200 lifetime, built specifically for one product stores. His hummingbird store runs on it.
- Elixir, about $175 a license. The one he used for the divorce book build.
Both come with the conversion furniture pre-installed: multi-tier discounts, "selling fast" callouts, emoji benefit lists, review blocks.
Three rules from his build process stuck with me:
→ Ignore the homepage. 90% of his ad traffic goes straight to the product page, so that's the only page he polishes.
→ Design for phones only. He says 99% of his traffic is mobile. He doesn't even look at the desktop preview.
→ Let AI write the first pass of everything. During the episode he pasted his old landing page into ChatGPT and asked it to generate the new store copy, then asked for the five main product benefits at 75 characters each, then 15 static ad headlines. He treats the output as an 80% head start, not a final draft, and rewrites the rest by hand.
The ad creatives follow the same pattern. Canva templates, photos of the physical book, HeyGen avatars holding the product (he clones his own voice so the delivery doesn't sound robotic), and headlines pulled from the ChatGPT batch. His best performer on the divorce offer: ads implying your lawyer is holding out on you. "The book your lawyer won't hand you." He says those always win, because that's exactly how the buyer already feels.
For testing, he tosses everything into one CBO campaign on Facebook at a minimum of $50 a day. If the product costs more than that, he spends at least one product's price per day so a single sale breaks even.
One thing I'm not endorsing: Sam admits he seeds new stores with fake reviews and swaps them for real ones later. He said it on camera like it's standard practice. It might be common in dropshipping, it's still a bad idea, and if you're building anything with your name on it, don't.
The proof it compounds: $30M of sugar water
The divorce book is a fun live demo. If you want receipts, look at the hummingbird business.
Sam found a guy on Amazon who had done $30 million over 10 years selling hummingbird nectar. Which is sugar. Literally sugar water, with atrocious branding, and the guy had never run a Facebook ad in his life.
So Sam launched his own, direct to consumer, on a Solo Drop store. He told Chris Koerner on a separate podcast that it came out of a Twitter reply and did over $5,000 in its first 21 days.
Now? He says it takes him about 90 minutes a week. Make a few new ads, check which ones are working, adjust. The ads do the heavy lifting.
And his honest ceiling for this model, stated plainly in the episode: he's never trying to convince anyone these are billion dollar ideas, but he has very tangible proof that businesses like this can make you $100,000 in profit a year.
A $100K/yr profit stream that costs 90 minutes a week is a better business than most funded startups I've seen.
What this means if you build software
Greg's reaction mid-episode was the same as mine: "I'm a software guy. This whole world you're sharing, I had no idea existed."
I build SaaS. You probably do too. So why am I writing about ebooks and sugar water?
Because Sam's stack transfers directly:
- One offer, one page, one traffic source. Most SaaS founders run five channels badly instead of one profitably.
- Know your slot machine math. CAC against first-purchase value, not against a three-year LTV fantasy.
- Validate customer acquisition before you build more product. Sam ships a 70 page version first and only invests after sales.
- Use AI for the 80% head start on copy, creatives, and research, then edit like your money depends on it. Because it does.
If the ecom angle itself pulls you, I've covered adjacent plays before: how a founder rebuilt the AI ecommerce business playbook with agents doing the operating work, and why digital dropshipping is the zero-inventory version of exactly what Sam runs.
Both run on the same engine as Sam's stores: a boring offer and math you actually check.
FAQ
Can a one product Shopify store actually make money?
Yes, if the unit math works. Sam Thompson runs his stores to a simple formula: cost of traffic times conversion rate times average order value has to come out positive. His divorce book funnel ran at break even ($42 average order against a sub-$45 CAC), and his hummingbird nectar store is his proof that a single-product store can clear $100K a year in profit on about 90 minutes of work a week.
How much does it cost to start a one product Shopify store?
Less than you'd think. Sam's stack from the episode: a conversion-optimized theme (Solo Drop at roughly $200 lifetime or Elixir at about $175), Envato Elements at around $15 a month for mockups, Canva for covers and ad creatives, and a Meta test budget of $50 a day minimum. The episode title puts his all-in tooling spend at $289 before ad spend.
Do AI-written ebooks really sell?
Sam's did, at $29 a copy with a $17 bump offer. But note what he actually did: shipped a rough 70 page version first, confirmed people would pay, then invested in a 197 page rewrite with case studies and deep research. The AI wrote the words. The offer, the market insight (huge wedding industry, 50% divorce rate, only lawyers advertising), and the ad angles were his.
What conversion rate should I aim for?
Sam's range: 3% to 5% on a product page from Meta ads is a reasonably good space. Under 3%, rework your copy and offer. Over 5%, either pile on more traffic or raise your price and watch whether the math improves.
Steal playbooks like this every week
This is what I do: find founders with real numbers and pull their playbooks apart so you can run them yourself.
On the Profitable Founder Podcast I interview bootstrapped founders making $100K to $10M a year and get them to show the actual math, the way Sam did here.
