Reading Time: 4:45

Apex predators reside at the top of their food chain and have few or no predators of their own. We introduced the notion of Apex Internet Predators (or apex predators for short) and “Amazonization” here.

(Before Amazonization, there were fears of another apex predator, Walmart and the “Walmartization” of America.  Low wages and prices were seen as an existential threat to local retailers and living wages.)

Amazon is not the only apex internet predator today but, to many CEOs, the existential threat of Amazon entering their business’s space is a real nightmare to lose sleep over. (If Amazon is not the peak apex predator you worry about, there are others to consider, like Alibaba.)

This post is for everyone who should worry about apex predators. We’re going to look at Amazon and what they’ve done.

Amazon is not without its own flaws

Amazon is not the perfect apex predator. Its latest US SEC 10-K filing contains 8 pages of material risks that could jeopardize the firm. At the top of the list of risks, Amazon declares “We Face Intense Competition” first. After all, other former apex predators have fallen (as in Sears was the Amazon of its day) and some see everything Amazon is doing as a repetition of the history of firms like Sears (The History of Search Predicts Nearly Everything Amazon is Doing.)  Recall that before firms worried about being “Amazonized” they were worried about being “Walmartized.” 

(Amazon might object to the term “predator” because, as their CEO, Jeff Bezos has declared, they are focused on delighting consumers, not competing with other firms.) Whatever Amazon’s intentions, their capabilities represent an intentional or unintentional threat. 

Amazon is an R&D monster

Amazon has been spending large amounts of money on R&D since it was founded and its R&D spend has been on a clear upward path since its creation. They invested more than any other commercial enterprise in 2017, $22.6 billion.  That was 12.7% of total Amazon revenue for the year. In 2018, Amazon R&D grew to $28.8 billion (a 27.4% increase year over year) but sales grew slightly faster so R&D came in at 12.4% of sales for 2018.

By contrast, across the US economy, R&D spending runs overall about 2.8% of GDP.  And CPG firms typically spend 1 to 2% of revenue on R&D.

Amazon started late in the game in AI (compared, for example, to Microsoft and Google) but it has more than made up for its late start. Wired provides an insightful look at how Amazon succeeded in AI.

Look back to Jeff Bezos letter to shareholders from 2010 (when Amazon’s R&D investment was roughly one-tenth of its 2017 level.) 

We use high-performance transactions systems, complex rendering and object caching, workflow and queuing systems, business intelligence and data analytics, machine learning and pattern recognition, neural networks and probabilistic decision making, and a wide variety of other techniques. And while many of our systems are based on the latest in computer science research, this often hasn’t been sufficient: our architects and engineers have had to advance research in directions that no academic had yet taken. Many of the problems we face have no textbook solutions, and so we — happily — invent new approaches.


Our e-commerce platform is composed of a federation of hundreds of software services that work in concert to deliver functionality ranging from recommendations to order fulfillment to inventory tracking. For example, to construct a product detail page for a customer visiting, our software calls on between 200 and 300 services to present a highly personalized experience for that customer.

All the effort we put into technology might not matter that much if we kept technology off to the side in some sort of R&D department, but we don’t take that approach. Technology infuses all of our teams, all of our processes, our decision-making, and our approach to innovation in each of our businesses. It is deeply integrated into everything we do.

It’s now nine years later, Amazon’s R&D spend is ten times what it was then and they’ve been pouring all this money into their own, unique, homegrown technologies (many of which they acquired by purchasing small startups.) 

And Amazon repurposes whatever it builds for itself for resale to others, generating more profits to plow back into their technology-rich commercial offerings “virtuous cycle” or flywheel.

Commercial off-the-shelf technology? Not much

It’s homegrown (or acquired), not commercial off-the-shelf technology. 

Amazon is the largest Internet company in the world (by revenue) and the second largest employer in the United States. Amazon is many things including:

  • A cloud-based services provider (AWS is the number one Cloud Services Provider)
  • A retail platform 
  • A marketplace (selling goods from other merchants)
  • A web advertising engine (pulling advertising business from Google and Facebook)
  • A purveyor of media (homegrown and from others)
  • Inventor of the most successful premium loyalty program (Prime)
  • A major business process and business technology innovator (as in Amazon Go)
  • The market leader in verbal virtual assistants (aka Echo’s Alexa)

Amazon exploits a broad range of AI technologies in every one of those areas.

To achieve that success, Amazon has built a new class of systems that exploit many different technologies including AI. These systems include:

  • Robotics in fulfillment
  • Transportation 
  • Knowledge graphs
  • Sentiment analysis
  • Predictive analytics
  • Micro-targeting (assembling offers for markets as small as one person)
  • Well-architected APIs
  • Cloud computing

Amazon generates and ingests massive amounts of data from almost everywhere, including:

  • Its supply chain
  • Cloud services
  • Buyers
  • The browsing public
  • Its Prime members
  • Competitors
  • Open data available from various sources

And AI is now virtually everywhere in what Amazon does.

We have a massive conflict of approaches 

In Only 5% of US enterprises are investing in serious AI-based business projects we pointed out US CEOs saw internet related issues as more important, here and now than AI was. (And they perceived that AI would become a priority five years hence, perhaps when the internet threat was over.)

But the existential threat from the internet (and apex internet predators) is at least in part (large part) based on AI! 

Meanwhile, we’ve sensed that CEOs are gun-shy about major AI-related investments to the extent that many of the major “moon shot” projects announced a few years ago have not produced independently verifiable evidence that massive “moon shot” investments pay off. 

That’s contradicted by the success of Amazon with AI.

Then again, it took Amazon, with all its resources, years to make the plan pay off (although they focused on accumulating a series of incremental wins within a vision that was much broader.)


Take another look at the pattern described in the Wired article on the flowering of AI at Amazon. I’ll dive into that and related analysis in the next major post in this series.




Disclosure: I wrote this post. It reflects my current opinions. No one paid for this work and I have no vested interests (long or short) in any of the firms mentioned here.

This post is the third in a series on AI strategies in the enterprise. The first two are