The availability of public cloud services has created enthusiasm bordering on hysteria, but early market successes...
for the public cloud make up only a small fraction of total enterprise IT spending. To successfully grow the public cloud services segment, cloud provider business models must be considered carefully, with entrants navigating between the competition created by early leaders like Amazon and the questions surrounding longer-term opportunity.
Research suggests that the PaaS model offers the largest total addressable public cloud services market, with the potential to address nearly 60% of total opportunity.
Broadly speaking, cloud computing services substitute service cost for customer’s capital and support costs. Cloud services are valuable to the buyer because they reduce overall IT cost or provide some other tangible benefit like the ability to create a separate development configuration. Of course they’re valuable to the seller if they earn a suitable profit. The cost and benefit of the services depend on the specific type of cloud services offered, so this is the most fundamental choice a cloud provider must make.
While there are already multiple cloud provider business model choices, each model has its benefits and risks. Below we’ll take a look at the main possibilities: Software as a Service, Platform as a Service, Infrastructure as a Service, combination approaches, and finally, the need for effective and secure data hosting, which can result in Database as a Service offerings
Assessing your optimal cloud business model
Software as a Service (SaaS). SaaS has the most radical tradeoff between service cost and internal cost. The user buys access to an application or set of application components hosted by the cloud provider. In the SaaS model, the user bears no technology costs for servers, software or support, so SaaS displaces the largest amount of internal IT costs for the business user. It’s also relatively easy to manage SaaS services, so using them adds little technical burden on businesses.
It makes sense that SaaS offers potentially the most compelling benefit for enterprises, particularly for firms whose technology support costs are higher than average. That includes most small and medium-size businesses (SMBs), whose ability to attract and retain skilled technology professionals is most likely limited. These firms also include companies in geographic areas where the local labor pool lacks suitable candidates for tech support positions.
Platform as a Service (PaaS). The intermediary service model is Platform as a Service, where the buyer is purchasing a software platform and hardware on which an application or application components are run. Because PaaS services do not include the application software but do include all other technology elements, their software and support costs for the buyer are higher than with SaaS, and the total savings available to businesses may be smaller because PaaS doesn’t displace as many cost elements.
The benefit of PaaS is that it allows users to run their own applications, even customized or self-developed applications. Since the user is contributing some software, the company must also support that software in the cloud. This means that for some users and some applications, PaaS may create an additional management burden large enough to compromise its overall savings for the business. Still, research suggests that the PaaS model offers the largest total addressable public cloud services market, with the potential to address nearly 60% of total opportunity.
Infrastructure as a Service (IaaS). The most general and versatile of all cloud service models is Infrastructure as a Service. IaaS has been the early market leader, but the business case for IaaS services is the most problematic. In effect, IaaS requires that users develop a software or machine image of their application and associated operating system and middleware in the same way they would support a virtual machine. The cloud provider simply hosts a virtual machine where the machine image runs. As a result, the cloud user must provide and sustain all of the software, as well as support and manage the cloud host at the “logical” or “virtual” level, though not the physical hardware. This means that IaaS displaces the smallest cost to the buyer, which means the cloud provider’s services have very tight profit margins to assure adoption. In the SMB market, the ability of businesses to manage IaaS services is often reported to be so complex that the option simply isn’t practical because of the lack of available skills and the thin cost/benefit gains IaaS could offer.
The value of IaaS to users, which has driven its near-term success, is its versatility. Users can run almost anything on an IaaS service, so IaaS has taken off in application development, testing and pilot projects, where buying dedicated hardware is hard to justify. IaaS also offers a flexible and powerful alternative to traditional Web hosting for users who need more than mass-market Web services. Finally, IaaS is easily used to extend data center virtualization, making it a logical cloud companion for enterprises with large virtual machine commitments.
Some cloud providers consider a two-tier cloud service model
The flexibility of IaaS and the limitations in its benefit case have induced some cloud providers to consider a two-tier service model, where the cloud provider offers IaaS services at a wholesale level to software partners that can then build SaaS or even PaaS services on top. This model is particularly valuable to telecom providers and other companies accustomed to making long-lived capital investments with limited return on investment (ROI). They can partner with established software players that have good relationships with buyers and offer software companies a way to offer their software in the cloud without building their own data centers. Verizon’s partnership with SAP is a good example of a combination model. Buyers of this kind of service benefit from having a trusted software element as a service at a lower capital and support cost.
Using your target markets to help define your cloud business model
Another issue important to cloud business model selection is the target market, as my comments on the SMB space show. Cloud services that address applications as simple as a company’s Internet presence can be sold across the full range of vertical markets and company sizes. Cloud services like IaaS that require more technical support are more difficult to sell to SMBs because they often lack access to the resources.
Cloud services that target mission-critical enterprise applications—the largest potential target market—must address enterprise goals of using the cloud for workload offload and backup. These applications are more easily supported in a PaaS model where the cloud’s software framework matches the enterprise’s commitments.
In terms of SaaS, some progress is being made in using principles of application componentization from the service-oriented architecture (SOA) evolution to offer SaaS-based backup and offload to enterprises.
Creating a cost-effective cloud database model for data hosting
The final issue in the cloud business model is the question of cloud data hosting. Applications need data access, and it is not only impractical to replicate most companies’ full data repository in a public cloud, but the companies wouldn’t trust the cloud provider’s data security. Even IaaS services like Amazon’s EC2 offer a form of database hosting, making the service effectively a limited PaaS offering, but there is still considerable technical and marketing work to be done to create a cloud database model that is cost-effective for the buyer, profitable to the seller, and secure and stable enough for businesses to depend on.
Data hosting may be the most significant single factor limiting the growth of public cloud services—so this may be the most significant issue for potential cloud providers to consider as they plan their offerings.
About the author: Tom Nolle is president of CIMI Corporation, a strategic consulting firm specializing in telecommunications and data communications since 1982. He is the publisher of Netwatcher, a journal addressing advanced telecommunications strategy issues. Check out his blog, Uncommon Wisdom