April 24, 2012
Summary – it’s a long post so here’s what you need to know. Amazon Web Services is moving up the stack. Existing ecosystem partners will benefit from that in the short term but it’s a risk for them long term. Other infrastructure vendors will surely follow suit. It’s game on! Now if you have the time, read on for the rationale.
Last week Amazon Web Services launched its Marketplace, a service that provides a one-stop-shop where customers can find applications, purchase them on a subscription basis and have Amazon take care of all the underlying infrastructure that makes them work. The Marketplace offers a vast array of applications – from CRM to Business Intelligence, from collaboration to core Operating Systems. It’s a clear indication that, at least in part, AWS realizes that the premise on which it has built its business, the provision of dumb infrastructure, is becoming commoditized and that increasingly the value will be gained further up the stack. While it is obviously still attractive to leverage a physical server to sell dozens of virtual servers to customers, it is far more attractive to leverage that server to sell applications on top of virtual servers.
In covering the announcement, Alex WIlliams rightly pointed to this fact saying that it;
makes the company a full-fledged app seller with the undoubtable potential to move further up the stack and become the next Microsoft.
While infrastructure vendors are spending attention cycles discussing the (admittedly important) issue of open versus closed, of different flavors of open, and of which infrastructure components are best to offer customers, AWS is quietly building an exceptionally compelling vertical proposition for customers. After all, if the core value of cloud computing is to allow people to forget servers, than surely the extension of that value is to allow people to forget servers, to forget deploying hardware, in fact to forget everything bar the specific customizations they need to map the software to their own business process.
I’m actually surprised that more people haven’t covered this move and identified it for what it is – a fairly seismic shift for the cloud industry. As I said in a recent post when discussion the big opportunities in the cloud, the areas t watch are ones which straddle a bunch of different working environments – multi vendor control software, multi software integration plays, broad API enablement platforms, things of this sort. The AWS marketplace straddles both vertical and horizontal areas – from a vertical perspective it ties together infrastructure and software, while from a horizontal perspective it offers up a vast array of software for customers needs.
The question is who does this move threaten. Head of AWS Werner Vogels has famously said that they don’t want to be a PaaS, I suspect what they do want to be is far more dangerous for far more players than simply a PaaS. So it is telling to look at those who have become part of the marketplace by virtue of the services they provide. Case in point BitNami, a company that packages different applications and deploys them on infrastructure platforms. BitNami is a partner of the marketplace and customers acquiring WordPress, Drupal and many other apps on the marketplace do so by using BitNami’s pre-configuration services. While that is beneficial for BitNami short term, I don’t doubt that they took a long hard look at the situation and are under no illusions that the dragon that is AWS could turn right around and eat their lunch. I put this to BitNami CEO Erica Brescia who told me that;
There is certainly the risk that Amazon moves higher up in the stack and starts packaging applications themselves (the same way they “repackage” RHEL with Amazon Linux or MySQL with Amazon RDS). The risk will grow bigger the more successful we are. However, packaging apps is far enough removed from their core expertise and cannot benefit from the economies of scale of their IaaS and PaaS that we really think we could have a chance at competing with the platform vendor if it came down to that. We have competitive assets in the fact that we already have relationships in place with hundreds of ISVs through our pre-existing packaging tools offering and also on the ability to support other clouds/platforms, such as VMWare, native installs, etc. which are important since few organizations and developers live entirely in the cloud. But as we have expressed in the past, building on top of a platform involves having to be acutely aware of where the platform vendor is moving. The biggest risk for us at this point would be not to build on top of Amazon.
Brescia raises a good point, the money to be made out of what BitNami does is negligible from AWS’ grander perspective. It is worth their while to encourage these sort of ecosystem partners to do the low lying tasks. Still, it needs to be said that there are a number of ecosystem partners who currently make their money deploying applications onto AWS, there’s obviously still money on the table in terms of consulting and software customization. But for pure vanilla installs, customers will bypass partners and go directly to the marketplace. At what point does the business case become compelling for AWS to further lubricate the process of software deployment and include services that were once considered the realm of consulting firms? Or, more correctly, at what point does AWS see that BitNami has a successful formula and simply acquires the company to continue creating these packages?
I asked Andy Pattinson, Commercial Director at ProQuest consulting and an advocate for and partner of AWS for his thoughts on what the marketplace means for them and specifically if it threaten some of their potential custom, Pattinson’s answer was positive and spoke to the bigger oppotunitites here;
I think the AWS market place is an excellent step forward for Amazon in the same way the Appexchange was for salesforce & the app store for iOS. The opportunity to broadly market solutions for the entire community is excellent & to consume others solutions for the benefit of clients does nothing but benefit everyone.
Of course the other infrastructure vendors aren’t sitting still. Equinix in particular already has a private marketplace for customers that offers a similar sort of functionality. I’m aware that Equinix is developing this service strongly and is no doubt looking at the ramifications of rolling it out on a public basis. VMware too has been, for the past couple of years, making strong moves up the stack. On a platform level it has it’s CloudFoundry OpenSource PaaS, while at an application level is has built a small suite through the acquisition of SlideRocket, Zimbra and SocialCast – this approach however is very closed and doesn’t lend itself to the building of a strong, broad and varied ecosystem.
Increasingly infrastructure vendors seek to differentiate themselves and do so by moving further up the stack. The ultimate aim is to offer a location where customers acquire their infrastructure needs and software requirements. AWS has introduced something that will become a model for future technology delivery.