Amazon Web Services (AWS) is the undisputed leader in infrastructure-as-a-service (IaaS). Historically, this meant focusing on low-level “building block” services such as object storage, scalable compute and caching.
But recent years have seen them build and launch higher-level products such as databases and content delivery platforms. These services are cheap, scalable and performant, but still narrowly defined. Now Amazon seems to be getting even more aggressive, and more serious, about moving “up the stack.” No longer content with reselling fractionalized hardware, they’ve set to work building complex software products to serve more of their customers’ needs.
Some cloud entrepreneurs are concerned, and rightly so. Among the new products launched this year and on the roadmap are many that bring them into direct competition with startups — startups that never expected to have to go head-to-head with AWS.
For example, Amazon’s new Elasticsearch Service allows a customer to deploy an Elasticsearch cluster with the click of a button, and costs no more than the standard rates for the underlying EC2 nodes on which it will run. At Elastic, the company behind the open-source Elasticsearch project, they have built a substantial business supporting these clusters, many of which run on AWS. Amazon didn’t pull the plug on them, but they made it easier than ever for customers to deploy Elasticsearch without ever picking up the phone to do business with Elastic. Other software vendors, open source or otherwise, who see their offering on Amazon’s roadmap may feel equally threatened.
Before we all give up and go home, I think it’s important to consider why Amazon is moving up the stack, and what it means for the industry as a whole. The dream of the public cloud has always been to achieve levels of scale and cost savings such that private data centers could never compete.
Cheap, instantly-deployable and infinitely-scalable cloud infrastructure is equally good for big companies with enormous demands and evolving needs as it is for the startup ecosystem just trying to innovate and show value to customers as quickly as possible.
Now that the foundation of infrastructure services is in place, Amazon is building the other features their offering needs to make it an attractive home for even the biggest and most demanding customers.
For every customer, the goal is to serve 100% of their infrastructure needs from the public cloud. In order to do this — in order for big companies to feel comfortable shutting down datacenters and re-learning how to manage infrastructure — Amazon must have an offering for every need. But it also means that AWS is maturing overall, both as a platform and as a business.
Beyond launching new, higher “ambition” products, Amazon is also behaving more like a big, growing and innovative company. From their partnership with Accenture, to the roadmaps they published during re:Invent, to the growing business development and customer service teams which make it easier to get someone on the phone — AWS is taking steps to be a supportive partner and provide transparency and comfort that investments today will not be squashed tomorrow.
Unlike Facebook, I wouldn’t be surprised if AWS’s motto was “break nothing and plan ahead.”
So a big, mature and ambitious AWS is good for the whole ecosystem, but many startup founders will still find themselves in direct competition with a much larger company, which is never fun. If that’s the case, there are several things you can do.
Be better at something important.
Amazon, and many big companies, try to make sure they have something for everyone, but would acknowledge that they’re not trying to build the best of absolutely everything. They’re exceptionally good at fundamental infrastructure, and their broad platforms like Redshift are great options when they work for the need, but by applying the 80-20 rule you can hone in on something important where you can beat them.
Amazon has offered CloudWatch since mid-2009, but that hasn’t stopped New Relic (disclaimer: I was an early investor and am on the board) from becoming the dominant player in the application performance management space. In fact, New Relic recently launched a beta version of an AWS-specific product, particularly well-suited to ephemeral EC2 instances that other monitoring products struggle with.
Aggressively move up the stack.
Amazon is talking more openly about its roadmap, and while there are some exceptions, most of what they’re building remains relatively low level. To stay out of their crosshairs, move up the stack, or build cross-platform and cross-cloud services that it would never make sense for AWS to build.
Make friends with their enemies.
Speaking of cross-cloud, you can always work with other cloud providers like Google, Microsoft, Rackspace or IBM. They all feel enormously threatened by AWS and are scrambling to close the gap in their offerings. If you have a solution that can help them I’m betting they’ll be receptive.
You could also try to get acquired. AWS has become more acquisitive as they take a strategic approach to enriching their product portfolio, though they may make you move to Seattle…
Five weeks ago at AWS re:Invent Amazon launched several products that sit at levels above their core IaaS products and attempt to solve big problems for their customers. Presumably these products also generate higher margin, but the more important thing they do is fill out a platform and bring it closer to the public cloud one-stop-shop that AWS wants to be.
While this may feel threatening to startups that see their vision on Amazon’s roadmap, the future is brighter for all of us who build products for the public cloud — change your strategy, build some new relationships and be the best at what you do. We’re still in early innings.