According to Gartner, through 2015, cloud services brokerage (CSB) will represent the single largest revenue growth opportunity in cloud computing. But like so many areas of cloud computing, there is confusion about both the terms and the very nature of this technology and concept.
First, there is not just one type of cloud services broker. As ubiquitous or utility computing evolves, there will be multiple types of brokerage services appearing, in the same way there are multiple types of clouds. For example, Amazon, Google, and Rackspace all use different underlying (and basically incompatible) technologies and approaches. There are also multiple “as a service” types of offerings -- such as Infrastructure as a Service, Platform as a Service, and Software as a Service -- and all of those providers are diverse and incompatible. And that’s just the tip of the iceberg: This is a technology that is varied, evolving, and diverging as services become more nuanced and specialized.
The brokerage market will be just as diverse as the cloud is. There will be brokers that do nothing but perform a management and integration service across multiple clouds and cloud type vendors for infrastructure and platform needs. The cloud computing landscape is evolving rapidly, with more and more players introducing cloud products and services of all kinds. Most recently we’ve seen the announcements by VMware partners, including Terremark, BlueLock, and others, as well as the introduction of Rackspace’s Cloud Servers. EMC is planning to offer a compute cloud, in addition to its existing Atmos storage cloud. And Dell is taking a broad approach through partnerships with Microsoft (Windows Azure Platform appliance), VMware, and by helping to develop Open Stack.
As the proliferation of offerings continues to accelerate, IT managers have questions about how to proceed: How can you evaluate the range of potential cloud offerings to find the right match? How do you route an application or workload to a target cloud and make sure that it works? How do you integrate it with other applications running back in the datacenter? Even within a single cloud, deploying an application requires learning the provider’s operating environment, management tools, and business terms and conditions. Doing this for every cloud provider you may wish to utilize is likely to prove daunting and not cost-effective. In a cloud environment characterized by multiple providers, each with its own service terms, operating platforms, management systems, security levels, and disaster recovery approaches, the specialized expertise and value-add of a cloud services broker will help IT managers find the right cloud offering, deploy their application in the cloud, and manage it properly.
Brokers will also form around domain areas (pharmaceuticals, IT, soft goods retail, process manufacturing, etc.) or applications areas (supply chain management, enterprise resource planning, human resources, payroll, etc.) -- perhaps even combinations of the two -- so that they become shopping centers from which companies can assemble the necessary services and functionalities for their businesses.
In a similar vein, we can expect the rise of cloud services brokers that specialize in data -- a consequence of the evolving connectivity of things and the concurrent big data deluge. These brokers will specialize in the collection and storage of domain, industry, or applications area data and the associated analytics. An example of this could be a clearinghouse for point of sale (POS) data collected across retailers and made available to manufacturers.
It is likely that the real evolution of cloud services brokers will be the logical extension of current business process outsourcing services (BPO) to include more end-to-end and higher-level knowledge skills incorporated with applications (or application components) and data. One can easily imagine everything that's not core to a business, or not directly creating value to a business's customers, eventually being outsourced. For example, the simple process of hiring someone could be done completely virtually by using multiple applications, databases, services, and external businesses via the cloud.
I even suspect there will be a transitional market opportunity for companies to sell excess capacity on their internal systems via cloud brokers, especially as they move their hygienic and housekeeping systems to the cloud. You already see this with Toronto-based Enomaly, which lets companies buy and sell unused cloud computing capacity in a clearinghouse called SpotCloud. It also gives companies the ability to switch services on demand to get the best price while still receiving a single bill.
And lastly, there is the traditional broker -- companies like Sterling Commerce -- that can route purchases or service requests to the “best” provider. It's likely to arise to deal with the variety of offerings. Think of it as “least-cost routing” for cloud computing. The benefits of this cloud and/or process fitting, whereby cloud services brokers use specialized tools to identify the most appropriate resource, and then map the requirements of an enterprise application or process to it, will soon become apparent. Cloud services brokers will be able to automatically route data, applications, and infrastructure needs based on key criteria such as price, location (including many legislative and regulatory jurisdictional data storage location requirements), latency needs, SLA level, supported operating systems, scalability, backup/disaster recovery capabilities, and regulatory requirements. IT and business process managers will be able to run applications or route workflows where they truly belong, while the broker takes care of the underlying details that make the cloud so compelling. For example, if you are hiring an employee in India, some of the BPO activities might be routed to a different service provider than if you were hiring someone in Russia or China or the United States.
So, given that “cloud services broker” is about as well defined as a marshmallow, what can we say about them?
Well, I agree with Gartner -- the market, however it evolves, will be huge. “Cloud” has taken so much friction out of the economics of IT that a long tail will be in full effect, with countless different service offerings to be evaluated, chosen from, integrated with, managed, etc. -- and companies will gladly pass that pain off to a broker.