ra2 studio - Fotolia
While many companies have headed to the cloud for IT infrastructure and applications, there are some obstacles to going there. One hindrance has been the absence of a one-stop shop among cloud services. While there are many cloud providers, no single cloud vendor delivers everything the market now demands. And companies are partial to having one throat to choke with IT vendors -- or only a handful to manage.
While there are a variety of cloud providers, there is no single standard for features and performance. Each of the major cloud providers offers a cross-section of cost, features and performance tradeoffs that attract only a subset of the market.
The result, especially in the high-end enterprise, is a multicloud implementation where several cloud providers are brought in to provide the full breadth of functionality and performance that the enterprise requires. Many such implementations exist; among enterprise-level organizations (with 1,000-plus employees), more than half have invested in a multicloud strategy. And that can be a hefty investment for even the largest companies.
Picking and choosing
Every organization has its own priorities, and it's not unusual that organizations seeking to migrate might have to shop around to find the platform that best matches their needs. What stands out among major cloud platforms, however, is that the distinguishing characteristics aren't a matter of nuance; for the most part, cloud providers that excel in one area are dismal in others.
Enterprises often have to go to extra trouble and expense to implement a cloud that includes several clouds, creating more work and cost, when the idea was to have less.
Viewed from the standpoint of user needs, rather than touted features, the ideal single cloud platform might look something like this:
Increasingly, big data rules in the planning of the modern enterprise. Data amassed over many years, even decades, can be unmined treasure or provide a mission-critical baseline for analytics of all kinds, any one of which can represent a strategic advantage in the marketplace or yield operational improvements.
Mass migration of data to a cloud platform, while a gargantuan chore that should not be underestimated, can offer a huge savings on storage and maintenance costs in the long run. If analytical functionality is available with the platform or easily deployed, all the better.
You would think that this would be a design imperative for every cloud provider, but many of the major providers come up short here. Salesforce, a long-time leader in data storage for analytics consumption, has the built-in software with Wave Analytics, but the platform is designed for field use, not for big data. It isn't built for both.
Microsoft Dynamics CRM has similar limitations, and the Azure cloud has big data capacity. But the indigenous storage model -- SQL Server -- is historically only moderately friendly to big data, due to the in-house speed paradigm of its indexing functionality. Microsoft is implementing a new database paradigm in Azure (and is very hush-hush about it), but we've yet to see how it will measure up. That said, Microsoft offers premium storage for higher rates.
With its Office 365 platform (which includes SharePoint) and Exchange hosting capability, Microsoft might seem the runaway winner in cloud utility, with available apps that are not only fingertip-accessible, but well-known to almost every computer user in the known universe. Indeed, an enterprise can license an Office 365 tenant for its users and migrate all email, Office, and SharePoint functionality into the cloud at once -- most in-house computing needs in a single move. And users could then easily access all their apps and data from anywhere.
But that move would come with a potentially serious compromise: Interfaces with other in-house apps (or, for that matter, with other companies) often entail custom code, sometimes of considerable complexity, and are often bottom-up -- extract, transform and load (ETL) occurring behind the scenes in SharePoint. Azure isn't friendly to those customizations, and Microsoft offers no guarantees that they will continue to function as intended or offer support if they don't. Enterprises are left with a trade-off between retooling their Microsoft customizations and rebuilding from scratch on the Salesforce Force.com platform or Amazon Web Services.
Public clouds are more common than private ones, as a cloud deployment offers a convenient and relatively inexpensive platform for interfacing with customers. Amazon leads in this area, with many agencies at all levels of government, as well as many public utilities, making information available to citizens and consumers via Amazon Web Services.
Microsoft, well-positioned to move forward here, has taken curious steps backward -- deprecating public-facing SharePoint, for instance, to the point where third-party software will soon be required to make it happen.
Google is strongly positioned here and, with its Force.com platform, Salesforce can potentially take the lead.
In general, availability isn't a high priority for cloud providers, regardless of its importance to the enterprise. Microsoft, Amazon and Google SLAs offer 99.9%, which is good but not great. Salesforce can deliver almost four nines of availability, but only with reduced performance. Google is the innovator here, offering reduced availability for lower fees.
Interface with other clouds
A major emerging issue in cloud platform selection is the ease with which one cloud platform can share data with another. This weakness is d to Salesforce, because its high-end analytics are friendly only to data actually originating in Salesforce: bringing in data from other sources is difficult at best.
Microsoft is, and has provided for a decade now, friendly interfaces with other data sources -- as long as the interface is Microsoft's.
In supply chain analytics in particular, this weakness can be detrimental: rapid and clean ETL between partner data sources is essential for effective, real-time decision making. Lack of a robust cloud interface standard is hurting the industry overall.
The next generation
These are only a few of the user-side concerns in crafting a one-size-fits-all single cloud solution. Championed by IBM and Amazon, infrastructure as a service is another layer to be considered as the industry moves forward.
What's the best short-term fix? Mergers between existing players hold promise: While there were rumors of Microsoft's bid earlier this year to acquire Salesforce, which should have benefited both companies by patching the other's weakness, a quibbling $15 billion came between them. Among companies this large and powerful, there are bound to be barriers: but, having taken business computing to new heights with the advent of the cloud, the industry now stands under the spotlight of user demand, and one or more of the major providers needs to move closer to that demand.
Multiple cloud strategies on the rise
Factors to consider when choosing a cloud vendor
A guide for comparing cloud vendors