Puzzle Ready: 99% Amazon FBA, $3M/Year — a Single-Product Puzzle-Table Brand Shaped Into What Buyers Want, Sold at 6x Profit

Dollar amounts throughout are converted at roughly ¥150/$1.

Conventional Amazon-seller wisdom says “get out from under Amazon dependence”, build your own storefront, grow a social following, expand into wholesale. But Puzzle Ready, a puzzle-table e-commerce brand Neil Jervis started in 2018, deliberately rejected that playbook. It kept 99% of sales on Amazon FBA (Fulfillment by Amazon), stuck to essentially a single product line, reached $3M/year in revenue (~¥450 million), and sold in December 2021 to the Finland-based brand aggregator eBrands for seven figures (over ¥150 million). The interesting part of this story is that refusing to diversify, far from being a weakness, was part of the deal terms.

The numbers through the sale

What was decided before day one

Jervis didn’t start on a whim. He landed on the puzzle-table category only after years of continuous product research on Amazon. At the time, this market had effectively just one other seller, thin competition and, by his read, stable demand. Niche but open. He waited until that condition was met, and that was his starting point.

What he built was a wooden board for assembling, storing, and transporting puzzles. The surface has an anti-slip finish, and it includes color-coded drawers for sorting pieces. Offered in two sizes (1,000 and 1,500 pieces) priced above competitors in the $70–140 range. He took the position on quality, not on discounting.

Manufacturing was outsourced to a factory in China, found through a sourcing agent. Finished goods ship directly from the factory to Amazon’s warehouses. Jervis himself never holds inventory. Operations involved just two people, based in the Philippines and China, three people total. Running $3M a year with three people is an extremely light structure, inventory risk aside.

The moment things changed, and a second decision

The external factor that pushed the numbers up is clear: COVID drove a spike in jigsaw-puzzle demand, and puzzle tables rode that wave. 31,000 units and $3M over the 12 months through December 2021. That growth can’t be explained by his own effort alone.

But the real turning point in this story sits elsewhere. Early on, Jervis decided “the buyer is going to be an aggregator” and built the business backward from that. He deliberately avoided the diversification other Amazon sellers pursue as a matter of course, building an owned store, spreading across channels, expanding the product line. His own reasoning is blunt: “I knew aggregators were my target buyer. They didn’t really care about that [diversification].”

Here the sequence is reversed from the norm. Rather than growing the business first and then hunting for a buyer, he set the buyer’s evaluation criteria first and shaped the business to score as high as possible against that specific criteria. A 99% Amazon FBA concentration is a vulnerability from a founder’s perspective, but to an Amazon aggregator, it means “this drops straight into our existing operations,” i.e., low integration cost after acquisition. No warehouse, no employees, one product line, a design with almost zero friction in the handoff.

Why a 6x multiple

The sale was at 6x profit, 3.5x upfront, 2.5x tied to one year of performance (an earnout). This level breaks down into three factors.

Take how he found buyers. Jervis used a broker who specialized in aggregators. That broker’s fee, notably, was paid by the buyer, not the seller, keeping the seller’s real cost low. He’s noted that this arrangement has become rarer now that the aggregator market has cooled.

He also created competition. He talked to 4–5 prospective buyers, and two, including eBrands, emerged as serious finalists. He’d been using eBrands as a “stalking horse” to drive up the price with the other bidder, but eBrands itself ended up offering the best terms, and he says he “had to accept it.” Having the stalking horse turn out to be the real buyer is an irony, but there’s no question that running multiple candidates in parallel pushed the final price up.

Then timing: 2021 was the peak year for valuation multiples on Amazon FBA businesses. Jervis himself has acknowledged that part of the strong valuation came down to market timing.

He sold not because it was declining, but because it was thriving

At the time of sale, the business was at its all-time best. He still let it go, because a structural fear outweighed the profit.

The biggest pressure was cash flow. Annual sales peak concentrates around the year-end shopping season, which means placing and paying for large inventory orders months in advance. “The forecasting was terrifying. We were looking at needing to sell about $1M [~¥150M] worth of boards in December alone.” If it sold, that’s profit. If the forecast missed, he’s stuck holding inventory, and ocean freight rates were spiking at the time.

Add to that single-product dependence. With only one product line, any one of several risks (copycats entering, an Amazon algorithm change, post-COVID demand reverting) could wipe out most of the revenue in one stroke. His summary is candid: “It felt very fragile. The risk was starting to outweigh the reward.”

This is less a story of “growing it to the max and then selling” than one of “stepping away, at the moment it looked its best, because he could no longer keep absorbing the risk.” After the sale, he launched Datarova, an analytics SaaS for Amazon sellers. Moving toward a business that doesn’t hold inventory reads as the mirror image of this experience.

What’s reproducible, and what isn’t

What’s most reproducible is the design sequence: decide the type of buyer you’re aiming for as an exit first, then shape the business to match the metrics that buyer values. This way of thinking applies regardless of industry. An aggregator values “ease of handoff”, and things individuals tend to assume are valuable, like a founder’s personal network or a diversified sales channel mix, can actually count against you as integration cost, depending on the buyer. It’s worth remembering that the evaluation axis flips depending on who’s buying.

Running multiple prospective buyers in parallel to create competition is also reproducible. Even talking to just 4–5 companies and narrowing to 2 can move the price. Counting, before you even start a sale process, how many buyers exist who’d want your specific category of business is preparation that pays off before price negotiation even begins.

All the same, some conditions clearly aren’t reproducible. Start with 2021 as the peak year for FBA valuation multiples, the same business sold after 2023 wouldn’t likely command 6x. Add finding a category with only one competitor at entry. That takes years of research and no small amount of luck. And the buyer-paid broker fee arrangement is, as he himself notes, a feature of that specific period.

Finally, a single-product, 99%-FBA structure only works “on the assumption there’s a buyer waiting.” If no buyer shows up, that same structure becomes the worst possible form of risk concentration. Jervis understood this, ran with it for three years, and got out while the exit door was still open. What’s worth taking from this case isn’t a product-selection trick. It’s a pattern of judgment: decide the ceiling on the risk you’re willing to carry, and step away once you’ve crossed it, even while you’re winning.

Related reading

• ScrapingBee — buyer selection and multiple negotiation, seen from the SaaS side
• Microns — how pricing gets set in the small-M&A market