In the early two-thousands, a small San Francisco company called Rain Design had invented a graceful aluminum laptop stand. It became a surprise best-seller on Amazon.
Then, in 2016, a similar product from Williams-Sonoma began to sell on Amazon as well. This time, the company called it the Orb.
Like the Orb, this chair was also made from aluminum, with a graceful curve. It was a success, and so were its Amazon competitors, who copied it with almost the same design.
When a customer placed an order, Amazon would send a notification to the seller. The seller would then ship the order to the customer.
As the retailer grew, so did its customer base. It had more and more people using its website, which required a growing variety of products. This led to an increasingly complex process for merchandising and inventory control.
To help with this, Amazon rolled out a technology platform that allowed its software engineers to provision on-demand resources like servers and storage instantly.
This gave Amazon a powerful advantage over a traditional retailer. Where a traditional retailer had to weigh tradeoffs within finite shelf space, Amazon could display page after page of products, with near-zero marginal cost for each additional item that it added to its catalog.
Once Amazon had solved the vendor bottleneck, it began to discover other constraints on growth. In this new world of infinite shelves, it discovered that it simply did not have the computing power or data storage capacity to fulfill orders quickly.
After all, its software engineers were often waiting weeks to have their hardware and data storage provisioned by Amazon’s internal IT department.
Then, in 2002, a six-page memo advocated the creation of a product that would remove all of these constraints. The solution was Amazon Marketplace, a system that would allow third-party sellers to display their products on Amazon’s site immediately.
As a result, third-party sellers could sell their products at scale, without having to worry about inventory and the painful process of negotiating with Amazon’s technology staff. It was an elegant solution to three problems at once: a massive source of revenue, a powerful marketing tool, and a quick way to introduce new products to the market.
While it wasn’t the first time a platform had been used to solve a growing problem, Amazon’s approach was particularly clever and powerful. It was built around a core concept: that if the broader market could solve some problem, Amazon could do the same thing.
Amazon’s emergence as a technology company was driven by a fundamentally different way of thinking from most other companies in the business: that its goal was to make things more efficient for customers, not for itself.
This mindset is what helped Amazon overcome its many growth obstacles and become a force to be reckoned with. It also enabled the company to create a platform-driven culture that thrived on constant improvement and innovation.
But the algorithm that powered Amazon’s rapid growth was a flawed one. Its platforms solved the problems that the company encountered, but they weren’t designed to be the final answer. They were a means to an end, and each one of them led the company to a catastrophic mistake.