A successful business gives its owner plenty to worry about. Yoda Yee has a particular concern: the part that is going rather well. A dependable customer, a lucrative service, a marketplace that sends the orders. Success can quietly become a dependency. By the time it feels dangerous, it may already be difficult to replace.
Yee runs Threecolts, the commerce software company he founded in 2021. Its products help sellers find inventory, manage listings, adjust prices and keep track of money. His public account of building the company keeps returning to a practical question: how much of the business rests on something it cannot control?
Five years inside the marketplace
Yee worked at Amazon from 2014 to 2019. Years later, explaining a surge in restrictions affecting sellers, he recalled an operating principle from his time there. Any category, segment or customer contributing more than ten percent of the business needed to be examined for concentration risk. His phrase for the prescribed response was “derisked aggressively.”
He described carrying that thinking into Threecolts. In 2021, subscription businesses diversified its reimbursement revenue. In 2023, enterprise customers broadened its exposure beyond sellers. In 2024, three multichannel software acquisitions and Walmart functionality widened the routes beyond Amazon.
Read together, those decisions suggest a founder using acquisitions to change what the company depends on. The transaction has a second job beyond adding revenue: it gives the business another way to keep operating when circumstances change.
A concentration threshold Yee recalls from his Amazon years.
Buying software, buying another way forward
Threecolts assembled a portfolio of applications with distinct jobs. InventoryLab, FeedbackWhiz, ChannelReply and DimeTyd were among the acquired businesses. In March 2024, the portfolio acquired something different: Marketplace Pulse, the ecommerce research publication founded by Juozas Kaziukėnas in 2015.
Software can tell a merchant what happened inside an account. Research can help explain what is happening around it. Bringing the publication into the group placed a business studying marketplaces alongside businesses helping customers operate on them. The acquisition announcement said the publication would expand its audience and coverage while continuing to look for stories in data.
The purchasing required capital. By June 2025, Threecolts reported approximately $200 million raised, 23 acquired businesses and more than 500 employees worldwide. Those were the company’s figures at that date. They describe the scale of the organization Yee was managing, rather than a personal fortune.
That June, announcing Tanya Andreev Kaspin’s appointment as finance chief, Yee discussed strategic possibilities that included entering public markets. The language was exploratory. His ambition had reached the point where public-market experience mattered to a senior appointment; a stock-market debut still belonged to the future tense.
Yee’s comments about his investors give a glimpse of the working relationships behind that expansion. He praised Crossbeam’s Ali and Sakib for being available and helpful while leaving the team space to execute. It is a fairly specific description of what he wanted from backing: assistance when a financing question needed attention, then room to get on with operating.
That relationship began in 2021, the year Threecolts was founded. The company that followed needed both kinds of work. Someone had to arrange the resources for acquisitions. Someone also had to make the acquired businesses useful to customers once the transaction paperwork was finished.
The uncomfortable arithmetic of a software bill
There is a pleasing awkwardness in a software executive complaining that merchants pay too much for software. Yee did exactly that in February 2025. He set out an illustrative seller’s monthly stack: research, inventory, accounting, repricing, shipping, analytics and customer service.
His example came to $444 a month. It was his constructed comparison, rather than a survey of every seller’s costs. The point was the accumulation. Each subscription could be justified separately. The seller had to pay the total.
A merchant makes room for one more useful application, then another. Soon the business has acquired a small committee of monthly direct debits. None needs to be extravagant for the combined bill to become unpleasant.
Yee’s accompanying product announcement put Seller 365 at $69 a month for ten applications and more than 100 tools. The launch covered sourcing, listing and shipping, pricing, and financial management. He tied it to conversations with customers facing rising shipping costs and platform fees.
Two other offerings addressed different operations: Margin Pro focused on costs and recovery for larger businesses; Multichannel Pro brought together work across sales platforms. The company was reorganizing a collection of capabilities around the jobs customers needed done.
The distance between a purchase and a habit
Owning several applications does not ensure that a customer will use several applications. Yee acknowledged this in a public account of Threecolts’ change of direction. During 2023 and 2024, the company had concentrated on selling individual products to customers already using another.
“It was moderately successful but not explosive.”Yoda Yee, on the earlier cross-selling approach
The bundle changed the offer. Later in 2025, he reported that cross-product usage had increased 28-fold, customer lifetime value had nearly doubled over four months, and the platform had attracted more than 4,000 paying customers. These were founder-reported results, with no independent methodology attached.
The distinction matters to the story. Yee was describing a change in customer behavior, alongside a change in price. A portfolio can grow through purchases. Making its products part of the same working routine requires customers to do something differently.
His April 2025 announcement of the Tactical Arbitrage 2 beta addressed that daily routine more directly. Users had found the existing product powerful but difficult to learn. The redesign simplified search setup, placed filtering on the results page and reduced movement between screens.
The development account named sourcing specialist Chris Grant as a contributor. It presented the redesign as work shaped by sellers, including experienced operators and people still learning the application. The useful detail is modest: fewer clicks, a clearer screen, less explanation before someone can begin.
That is a different kind of progress from an acquisition announcement. It happens where a person is trying to finish a task. The extra application in a subscription becomes valuable when somebody can use it without turning the afternoon into a training session.
A meeting with the marketplace
Yee’s public appearances also put him in conversations where the software company has to answer questions. In February 2025, he joined Clear the Shelf for a discussion about InventoryLab, Tactical Arbitrage, automation and Amazon’s API changes. The episode’s description explicitly promised questions about Threecolts’ past missteps.
In October that year, Silent Sales Machine Radio released his conversation with Jim Cockrum about Amazon resellers. The subjects were concrete: restrictions on selling, maximum-price transparency, counterfeit enforcement and used books. This was the marketplace as encountered through an interrupted working day.
Threecolts describes its advocacy process as gathering seller experiences, meeting Amazon teams and reporting outcomes to the community. Its published agenda includes category access and the pricing information sellers need before an offer loses visibility. Those are company descriptions of the work, rather than guarantees about what Amazon will do next.
Yee’s own March 2025 writing on reimbursement changes offers a smaller example of that role. He urged sellers to keep invoices, track actual costs and prepare inventory records while details were moving. A policy debate became a list of documents someone needed to retrieve.
Freedom for the team, structure for the work
There is another side to Yee’s operating interests. In a 2023 leadership essay, he described Threecolts as “remote first and global since the beginning.” He argued for employee autonomy and against command-and-control management.
His case included asynchronous communication: work need not stop until everyone can assemble at the same moment. For a company bringing together teams and products, that is a practical concern as well as a view about leadership. Decisions still need to travel; people need room to do their jobs.
Read alongside his product writing, the essay suggests a consistent preference for removing avoidable friction. Employees waiting on a meeting and sellers moving between screens face different problems. Both lose time before reaching the task they came to perform.
The next dependency is a convincing answer
By September 2026, Yee was applying an operational test to AI. Threecolts had built six MCP servers, he wrote, then tested them against real seller catalogs. The experience changed its approach.
His argument distinguished reading data from completing bulk work. Showing a price is a tidy demonstration. Updating a large catalog introduces questions about completeness, arithmetic and errors. He described AI Operator as a set of specific roles working within the structures of Threecolts products. These were his claims about the design, rather than independent proof of its performance.
“Bulk action is the job. Therefore, that’s the test.”Yoda Yee, September 2026
The essay brings the profile back to its opening concern. A convenient system can become another dependency. Yee’s proposed test asks what happens when a seller entrusts it with consequential work.
The customers at the other end still have products to list, boxes to ship and money to account for. They can judge the software in those terms. For the founder who learned to examine what was working, that is a suitably demanding audience.