Why Morgan Stanley Is Wrong About Amazon Web Services
In their research report published the other week, Morgan Stanley came out projecting that Amazon Web Services (AWS) may be on track to reach $22 billion in revenues by 2022. Although it seems that the financiers understand the magnitude of the paradigm shift and the impact it will have on existing IT incumbents, they seem to have missed some key dynamics taking place in the IaaS cloud market, leading to conclusions that could end up being way off the mark.
Originally launched in 2006, AWS is the undisputed cloud pioneer. With revenues expected to hit $3.8 billion U.S. in 2013, they dominate the space. While there are certainly blue skies ahead for the clouds over the coming years, there have already been some early signs of market dynamics that will make it harder for AWS to maintain its relative strength in this market.
In the software business, the money has traditionally been in the enterprise. The same principle is likely to apply to cloud computing too. Granted, the market is still at its infancy as most enterprises are yet to make serious moves towards cloud computing, largely due to security and availability concerns. In February 2013 alone, AWS experienced cloud outages that no serious enterprise would withstand. So it is quite clear that in order for the enterprises to fully embrace the cloud computing paradigm, something has to change.
Looking back at the evolution of enterprise computing, the IBM mainframes dominated the markets for a long time. In many ways, the cloud computing model is not unlike the mainframe-computing model, as the capacity is being consumed from a central location and paid based on use. But while the IT industry is closing a full circle on the centralized computing model, the world we live in has come a long way over the last 50 years. Although a computing outage may not have been that big of an ordeal in the 1960s, in today’s information society it would likely have disastrous consequences. To address the availability concerns, decentralization is the key, as relying on just a few major cloud providers is tantamount to putting all information society’s “eggs” in very few baskets.
AWS’s success in the cloud space has largely been based on the economies of scale they have enjoyed given their status as the largest online retailer back in 2006. Coupled with the fact that during the early days there was not much competition from companies of equal stature, AWS was able to leverage its partnerships with small technology pioneers and go to frontiers no man had gone before. Over the last year or so, however, the success of AWS has attracted the attention of the large IT houses, and the technology advantage it enjoyed is eroding fast.
During 2012, industry bellwethers like Microsoft, Hewlett-Packard and IBM along with somewhat smaller players have introduced new solutions designed to equip Managed Service Providers (MSP) and carriers with the technology required to introduce competing IaaS cloud solutions to AWS. In fact, already in 2012, there were a handful of cloud providers out there that beat AWS to their service automation launches that took place during February 2013. While most of these players do not enjoy AWS’s brand recognition, it has done very little to slow them down since their cloud offerings are targeted at an existing customer base.
To throw more gasoline into the flames, the large IT companies are now offering turnkey cloud platforms to service providers on a pay-as-you-go basis. By leveraging these solutions, hundreds of MSPs and carriers can become cloud providers almost overnight and start competing with AWS without significant investments in the cloud infrastructure. And perhaps more importantly, these emerging cloud providers are able to offer their enterprise customers with IaaS cloud services that offer a number of advantages over AWS, reducing the uncertainty and accelerating the adoption of cloud computing models among the enterprise.
Based on discussions with a number of prospective enterprise end-users, the attributes they value in their decision-making around the cloud involves the following considerations:
Analyst firms such as Gartner have declared the AWS service level agreements (SLA) useless. There have also been complaints that buying cloud services from AWS involves “take it or leave it” terms and conditions that do not fall down well with larger enterprises. The MSPs turned into IaaS cloud providers are used to negotiating SLAs with their enterprise customers, eliminating this concern.
AWS has introduced a number of cloud security innovations such as Virtual Private Cloud (VPC) and robust perimeter security allowing enterprises to securely tap into the computing capacity in the IaaS cloud. Also, many MSPs already have Virtual Private Network (VPN) connectivity between their data centers and their customers’ private networks. This allows them to offer IaaS cloud services as an extension of the enterprise computing environments, used for additional computing capacity when and where needed.
Since AWS consists only of a handful of data centers operated around the world, many enterprise customers would have to connect to the AWS cloud over wide geographical distances. For enterprises running time-critical business applications, the latency associated with geographically remote data centers responsible for running the cloud can easily become a deal breaker. At the same time, an emerging cloud service provider with data centers nearby – with a state of the art cloud application deployment solution – is often better positioned to address this concern.
Although AWS has state-of-the-art security measures in place, there are privacy issues that come with storing sensitive data outside an organization. The privacy issue is amplified with organizations having their domicile outside the United States, as storing critical data in US-owned data centers may be somewhat problematic due to the Patriot Act. Therefore, there is a clear need for regional and national players around the world that will be able to address the local privacy laws and considerations of enterprise and governmental customers.
Having its origins in the business to consumer markets, AWS’s differentiation is largely based on price advantage derived from volume and self-service. While strong brand and aggressive pricing may have been enough to drive AWS’s early cloud initiatives leveraged by SMBs and development teams, the IT departments in the enterprise and government are used to receiving more attention from their IT providers. As many of the cloud newcomers have a solid background in serving these accounts with personal attention, they are better equipped to tailor cloud service packages to their customers.
In summary, although there is ample body of evidence that cloud computing will change the ways in which applications and services are being consumed, the handwriting on the wall says that the revolution may well end up eating its children, at least in relative terms. Enterprises are likely to drive the second wave of cloud computing but will introduce a new set of requirements that the cloud incumbent is not positioned to address entirely. Therefore, while I am confident that the cloud market will enjoy tremendous growth by 2022, I would not bank on AWS being able to hold on to its relative market share.
By Juha Holkkola,
Juha is managing director of Nixu Software Oy Ltd, the cloud application deployment company, an affiliate of Nixu. He joined Nixu in early 2000 and has since held various business and sales management positions. Before Nixu, Juha worked for Nokia Networks and financial services company Danske Bank in marketing and treasury positions.