Recently, a Google employee called Steve Yegge posted a very long message about a few things his former employer, Amazon, does better than Google. It was intended to be an internal message, not something public, and he therefore made statements he probably would have avoided otherwise. But he published it to the whole world by mistake, and the quality of his observations ensured instant popularity for his post.
Yegge’s key point is that Amazon approaches its business with the ambition of building a platform, while Google is focused on the development of individual (and discrete) products. And the problem with this approach is that very few people, in computing, have ever been very good at reliably predicting what products consumers want. Many companies create a couple of very popular products, but very few can reproduce this outside of their initial business, or maintain that business in the face of changing conditions.
Working for a manufacturer of physical products, I could not help but wonder how we stack up, however distant software development may seem.
The concept of platform essentially enables companies to attract external developers who will create products that consumers want. According to Yegge, Amazon became a major platform player because of a top-down management decision, prompted by evaporating margins. It took several years to achieve, but without adding any new functions (i.e. not through traditional external growth), Amazon has expanded its operations into several new business areas, deriving profitability from every existing part of the company.
Anything Amazon used to build for its own online book sales, is now available to third-parties: the warehouses, the warehousing management software, inventory management, logistics including package tracking, the computer systems (in a myriad of individual services), the online merchandizing and sales systems, payment processing, affiliate management, etc.
All of these services are part of the Amazon platform, but are offered individually. The platform powers first and foremost its founding “killer app”, the original Amazon.com, but a very large number of other products also rely on some of its elements:
- My ASICS, which relies on web hosting, and benefits from very affordable prices and high-speed global coverage
- my local bookstore’s web presence, which makes use of the “marketplace” element, whereby stores can offer their stock through Amazon.com
- a university’s physics calculations unit, which relies on raw computing power once in a while to crunch numbers
What’s amazing is that none of these actually represent a stretch for Amazon: the services are all something they would maintain anyway because they are needed internally. But because of the way they have been built, they can very easily be marketed externally as standalone products, without disturbing in the least the company’s core operations. On the contrary, such additional paying users of each service ensure extra income at every level, which probably transformed many cost centers into profit centers!
Unlike Google, Microsoft or Facebook, Amazon is not a pure-play digital company. It owns and operates a vast logistics network, and physically ships hundreds of millions of products the world over each year.
But because its consumer-facing business is digital, they had the opportunity to think in those terms. They also had the necessary engineering resource, and the management culture to make the switch. Other companies have always been in the business of platforms (Microsoft) or have recently made the conversion (Facebook).
Yegge’s point shows that a powerful company such as Google can be immensely successful while not applying such an approach. However, he identifies this as an opportunity to ensure the long-term sustainability of the business.
Could ASICS learn from Yegge’s observations? Is there a way for us to make use of our existing infrastructure to offer new products and services to consumers, or to third parties who would in turn develop products and services made from elements of our platform?
There may be some elements to be found in the collaboration we are currently pursuing with Seiko. A successful platform relies on:
- attractive elements that have intrinsic value
- accessibility for third-parties to those elements
- a “killer app” that has broad consumer appeal
Our footwear products are obviously the current killer app: they are massively popular and represent a very significant mass of contact with consumers. It’s now up to us to “unbundle” the different elements that make up this product, and start imagining how individual parts can be recomposed into different products — and perhaps even allowing third-parties to do so directly.
P.S. Nice follow-up by Yegge on Google and Amazon.
ASICS as a platform
Recently, a Google employee called Steve Yegge posted a very long message about a few things his former employer, Amazon, does better than Google. It was intended to be an internal message, not something public, and he therefore made statements he probably would have avoided otherwise. But he published it to the whole world by mistake, and the quality of his observations ensured instant popularity for his post.
Yegge’s key point is that Amazon approaches its business with the ambition of building a platform, while Google is focused on the development of individual (and discrete) products. And the problem with this approach is that very few people, in computing, have ever been very good at reliably predicting what products consumers want. Many companies create a couple of very popular products, but very few can reproduce this outside of their initial business, or maintain that business in the face of changing conditions.
Working for a manufacturer of physical products, I could not help but wonder how we stack up, however distant software development may seem.
The concept of platform essentially enables companies to attract external developers who will create products that consumers want. According to Yegge, Amazon became a major platform player because of a top-down management decision, prompted by evaporating margins. It took several years to achieve, but without adding any new functions (i.e. not through traditional external growth), Amazon has expanded its operations into several new business areas, deriving profitability from every existing part of the company.
Anything Amazon used to build for its own online book sales, is now available to third-parties: the warehouses, the warehousing management software, inventory management, logistics including package tracking, the computer systems (in a myriad of individual services), the online merchandizing and sales systems, payment processing, affiliate management, etc.
All of these services are part of the Amazon platform, but are offered individually. The platform powers first and foremost its founding “killer app”, the original Amazon.com, but a very large number of other products also rely on some of its elements:
What’s amazing is that none of these actually represent a stretch for Amazon: the services are all something they would maintain anyway because they are needed internally. But because of the way they have been built, they can very easily be marketed externally as standalone products, without disturbing in the least the company’s core operations. On the contrary, such additional paying users of each service ensure extra income at every level, which probably transformed many cost centers into profit centers!
Unlike Google, Microsoft or Facebook, Amazon is not a pure-play digital company. It owns and operates a vast logistics network, and physically ships hundreds of millions of products the world over each year.
But because its consumer-facing business is digital, they had the opportunity to think in those terms. They also had the necessary engineering resource, and the management culture to make the switch. Other companies have always been in the business of platforms (Microsoft) or have recently made the conversion (Facebook).
Yegge’s point shows that a powerful company such as Google can be immensely successful while not applying such an approach. However, he identifies this as an opportunity to ensure the long-term sustainability of the business.
Could ASICS learn from Yegge’s observations? Is there a way for us to make use of our existing infrastructure to offer new products and services to consumers, or to third parties who would in turn develop products and services made from elements of our platform?
There may be some elements to be found in the collaboration we are currently pursuing with Seiko. A successful platform relies on:
Our footwear products are obviously the current killer app: they are massively popular and represent a very significant mass of contact with consumers. It’s now up to us to “unbundle” the different elements that make up this product, and start imagining how individual parts can be recomposed into different products — and perhaps even allowing third-parties to do so directly.
P.S. Nice follow-up by Yegge on Google and Amazon.