top of page
  • Writer's pictureHaydn Shaughnessy

Lessons from the Amazon ecosystem

What we can learn from one of technology's biggest players

The study of business ecosystems has begun to find favour with the larger consultancies, who have also begun to fixate on business platforms. That means the issue has arrived with their clients too, or soon will. It's a case of too little too late.

My view is that ecosystems are the most disruptive feature of economics we have seen but that we don’t understand them well enough to grasp just what they are doing to the economy. They are disrupting it and have already prepared the way for peak US influence on management thinking.

In this article I am going to look at the post-US business structure and illustrate it with one slice of the Amazon ecosystem. It will also illustrate our general state of ignorance about ecosystems.

A brief history of the business ecosystem

I’ve been writing about business ecosystems for a decade now and that leads me to think the reader with less familiarity would benefit from reading some background.

We have previously covered the Airbnb ecosystem here on Disruption Hub and I wrote about Chinese ecosystems on the GARP website as well as the status of fintech as an ecosystem.

In this LinkedIn piece I describe the background to platform and ecosystem thinking - worthwhile reading for the historical perspective. In there you will also find references to books written in the 1990s which already began to explore a particular kind of interlocked ecosystem, for the pursuit of complex innovation. This includes, for example, ARMS Holdings in chip designs, and Microsoft’s Value Added Reseller (VAR) networks, to scale markets quickly.

The Amazon ecosystem in books

What I am now going to argue is that our perception of ecosystems in the US and Europe is too hidebound. Even our best examples of business ecosystems are weak.

So below you see an overview of the Amazon ecosystem in books, and only books. I have called it Amazon Ecosystem Level 2 because it sits below the top line ecosystem that has Amazon connected into Cloud, electronics, and so on.

The colour codes are Yellow for entities owned by Amazon and Blue for entities that are owned by other parties. What you’ll notice is that the blue outnumbers the yellow.

No surprise, you might say, since Apple’s developer ecosystem is vastly more numerous than Apple and its divisions. But that is the wrong comparison. The right one would be Amazon and writers. What we are talking about in this diagram are companies and categories of companies that have been created because of Amazon or that are largely dependent on it.

I’ll explains some of the actors below. But first, why is this different from a supply chain or a developer ecosystem?

A supply chain is always in a direct economic relationship with the company that trades the final product. (I use the term direct even if they sell components to a sub-system.)

The relationships in the Amazon ecosystem are indirect or not at all transacted with Amazon as the endpoint. They are totally separate.

Developer ecosystems depend on an OS or other significant asset of a platform owner.

The companies here just need Amazon to be successful. They need no permissions whatsoever. They are like the SEO community around Google, sometimes antagonistic, but basically expanding the platform’s business without reward from the platform.

So why all this background? Terms like ecosystem are used very loosely but there are clearly different types of ecosystems.

The ARM Holdings design ecosystem is a set of relationships between major mobile device vendors, silicon manufacturers and chip designers, usually organised by Arm. It’s a highly specialised and effective ecosystem that can push chip design in novel directions and move designs to manufacture quickly because of the inclusion of silicon factories and smartphone makers.

Very few ecosystems work this way. In fact the Amazon ecosystem in books is trying to optimise the Amazon platform rather than optimise or maximise market opportunity and customer success. Take the role of book arbitrage (mid-right in blue in the diagram above). In essence this means finding books deep in the Amazon catalogue, buying them cheaply, and then using more effective descriptions to sell them, also on Amazon, at a higher price. It makes up for Amazon's indiscriminate search engine and the poor product descriptions of most booksellers.

It pays the Amazon ecosystem to get any good enough product to market to tap into the long tail, at a very low price (99 cent novels). That is output rather than outcome; it is a product that does not necessarily please customers as much as might be possible at a lower volume of publishing.

The term Amazon VAs stands for Amazon Virtual Assistants - people who can often be hired on Upwork or Fivrr and who specialise in product descriptions and keyword identification. Often they are based on SouthEast Asia and come fairly cheap, which is good labour arbitrage.

Netgalley is a platform of book lovers many of whom can be relied upon to post reviews, which is significant in terms of Amazon search returns and, hence, sales.

I hope you get the picture. Most of these third parties are there to help improve the functioning of the platform.

Looking to the East

I said earlier that the Arm ecosystem is much better. It is built around everybody moving the industry on, innovating designs in order to provide customers with new functionality on smart mobile devices.

Take a look at the picture below. Alibaba and its co-founder Jack Ma own or control companies that equate in purpose and role to all those below. So, for example they own a bank, are involved in operating system development (for cars), have an Amazon-like platform, are involved in insurance, finance, deliveries, travel booking, ticket sales, ride hailing, video streaming, video hosting, data analytics and so on.

That is an ecosystem more akin to the one controlled by Arm Holdings. Bigger and bolder of course, it outclasses anything in Europe or the US in terms of its ability to deliver user experiences.

I think that the new Data Transfer Project (DTP) launched by Google, Facebook and Microsoft is an attempt to counter this incredible power. The DTP is touted as a way to reduce data lock-in at the major platforms we all know and to some degree love.

Looked at through another lens it is in fact a way towards matching the integrated data capabilities of China’s new cartels. Anybody kidding themselves that Apple or Amazon will be the first $trillion business has their eyes closed.


Ron Kersic
Ron Kersic
Feb 19, 2021

Thank you for sharing Bezonomics (and what a great term); will for sure give it a read.

I suggestion I'd like to share: The Keystone Advantage by Lansiti and Levien. Together with their companying HBR article Strategy as Ecology, a bit of a landmark regarding ecosystem-first thinking.


Roger Pena
Roger Pena
Feb 18, 2021

Great book to follow-up with this post. Bezonomicss, by Brian Domaine. Very, very insightful book to Amazon's Ecosystem!


Haydn Shaughnessy
Haydn Shaughnessy
Feb 16, 2021

Hi Ron, I was thinking that through and I can see why one might want to.So, for example, rather than the Airbnb ecosystem one could say, the home rental ecosystem. However, Airbnb is not just about home rentals or sharing. It also sells experiences. What is the essence of this ecosystem? Perhaps it is: the personal asset marketing ecosystem? If so, then I am glad we have had a chance to think more deeply about it. How does that sit with you?


Ron Kersic
Ron Kersic
Feb 15, 2021

Other than referring to the Keystone, other ways to name an ecosystem could be by capturing the (core) population/community or the (core) process/function. For example (and my creativity is obviously running low here): "(local) hospitality", "low-power processor design", "online book publishing"..(?)

And if this approach does not capture the dynamics in full, maybe that's a tell of a Keystone that optimizes for the platform, and not the ecosystem?


Haydn Shaughnessy
Haydn Shaughnessy
Feb 13, 2021

That should read you are right but your observation raises

bottom of page