Editor's Note: When it's time to make a buying decision, matching product features with their intended use always tops the list. In the final part of our three-part series on building and integrating a cloud infrastructure to enable cloud services, we look at considerations that should seal the deal for cloud providers. Consider these factors before finalizing the purchase -- weighing product functionality against cost, cloud integration costs for adding new products into your existing infrastructure, balancing product flexibility and more hidden factors.
The most important question in selecting cloud infrastructure elements is whether the proposed product fits within the context of the cloud service platform the operator is trying to build. Its suitability for intended use is always the most important buying criteria, and buyers always list cost as the next most important criteria. Cloud infrastructure procurement is complicated because it's often difficult to know what the cost relationships are. Because of that, cloud providers face three main challenges.
Challenge 1: Comparing product functionality and weighing the cost
The first challenge in selecting cloud vendors is the lack of consistency in the scope of comparable functionality. It's easy to compare two products that do the same thing, but with cloud-building, it's not uncommon to look at combinations of products that fit together in a variety of ways. Picking Product X for a given role because it costs less might mean that to complete the functional picture, Product C is required instead of Product D, and C might be more expensive. In order to get a picture of true cost, it may be necessary to create ad hoc suites of products and compare the prices overall.
Challenge 2: Pricing integration costs
The second challenge in cloud vendor selection is pricing realistic integration costs. A few vendors can literally provide a "cloud in a box," but most offer components that require integration with others to develop full functionality. Businesses have consistently reported that integration costs are the most likely to have reported over-runs. One reason is that it's often difficult for cloud operators and vendors to understand exactly how the pieces are expected to fit together. That means that the "glue" needed to create a unified cloud infrastructure may depart from the plan, often significantly.
Challenge 3: Balancing services cost control and tactical flexibility
In terms of the services a cloud provider can offer, an offering that is too narrow in service scope may be difficult to sell, creating a loss of sales force momentum and limiting customer reference successes that help build service credibility. It may also delay time-to-revenue and hurt early financial milestones. On the other hand, flexibility costs money, and too broad a service scope can raise a cloud provider's costs and even dilute sales efforts by spreading the sales organization too thin. As a result, it may be important to look at the speed with which an early cloud offering can be expanded to a broader scope, and that's almost certainly a function of the flexibility of the cloud components and the ease of integrating them.
Read the rest of this series:
Table of contents: Enabling cloud services
Cloud infrastructure determines services flexibility
How to buy cloud platforms and components
10 questions for potential cloud computing vendors
To limit their service risks, most providers decide on a phased evolution from lab trial to field trial to a targeted offering and finally to a complete service offering. The goal of a lab trial is to validate the key technical/integration assumptions and ensure that the functionality of the cloud infrastructure matches the business goals. The field trial is an opportunity to test operating practices and efficiencies, and the targeted offering allows the provider to align its sales and marketing, deployment and customer support functions and gain insight into how quickly a revenue ramp-up can be expected.
The big question about deploying cloud services for providers is how to balance the need to build experience with a set of vendors and components with needing to be able to change out some components if they fail to perform. Getting the balance right depends on making sure that the early phases of your trials provide some exposure to the later issues (operations, deployment and support). Every phase must test the suitability of the choices made for the next phase or there is a risk of a major problem and a major delay.
All of this must be considered before a set of components/vendors has been selected, not after, which means you should consider these points in the final negotiations with vendors. To prepare for that, narrow the spectrum of possible players to two or three "suites" or collections of components. The goal then is to pick the combination that will work best.
Sealing the deal: Final factors to consider
Here are some points to consider in the final selection phase:
- Understand where each vendor fits in the spectrum of necessary cloud infrastructure components. Generally, players with broader scope will offer you a lower price than those with a narrow scope. At a minimum, the integration difficulties will be higher for the latter.
- Do not always assume that more flexibility or greater scope of possible services justifies a higher cost. Apply realistic values to both of these benefits, then compare choices on cost.
- Start your selection of products with the components of your infrastructure that are the most fundamental to your cloud service plans and build out from there. Otherwise you may reduce your range of choices before you get to your key areas of need.
- Providers that offer integration are generally preferred, but only if they will integrate your full range of elements and stand behind the results.
- Be sure you understand how each potential provider can integrate platform and software components into their infrastructure if you plan on evolving to higher-layer cloud services -- and you should be.
- Your greatest leverage will come by selecting the smallest number of suppliers possible; each then has a higher stake in the project. This will normally reduce your integration costs and difficulties as well.
- Be sure to write your full project schedule, from lab trials to full deployment, into your requirements and get vendor buy-in for participation in each phase under terms that are acceptable to your management. You don't want your trials to strand assets if you change your mind.
- Look for providers that can offer their components as an integrated piece of a public cloud service—it may allow you to trial their solution at little no cost.
- Engage every potential supplier in a dialog on your business plans and technical strategy. You need to understand how well they understand your approach and your needs. Every supplier will be a part of a whole, and only that whole can make you successful.
- Don't try to lay off too much risk by writing in detailed milestones and stringent penalties because it may make you commit to more than you currently know about your target market and your initial service decisions. Finally, understand your own service buyer's business case and make sure that your providers understand and accept their place in meeting it.
The early cloud service market focus was the opportunity for providers to host IT applications and infrastructure that has traditionally been private IT. Even now, with less than 2% of cloud opportunity being realized, cloud providers are looking beyond current IT applications and spending to deploy new consumer and business applications that redefine the relationships among computers, networks and people. This may be the most significant development of the decade, and the most significant opportunity for providers. Flexibility in following the opportunities that emerge in the cloud space may therefore be the most critical thing to look for in your own cloud infrastructure.
Learn more about Tom Nolle, president of CIMI Corp.