Exploiting the three business advantages of multi-cloud for analytics

According to Gartner, by 2025 80 percent of all enterprises will have a multi-cloud and/or hybrid cloud deployment. The vast majority of enterprises are already using the cloud today. They just need to use one more application on a different cloud provider to be multi-cloud. But that alone does not a cloud strategy make.

This is where CIOs are spending a lot of their time today—on figuring out how best to take strategic advantage of both multi-cloud and hybrid cloud deployments. To be clear, since these terms are often used interchangeably, multi-cloud is when you deploy different, unrelated applications or workloads on different cloud platforms. You may use Google Cloud to run business applications such as Workday, or Salesforce.com; Azure for disaster recovery and collaboration applications; and AWS for micro-services and storage. Hybrid cloud is when you're using two different cloud infrastructures in an interoperable fashion to optimize a single workload. This allows your applications to be portable.

Analytics are perhaps better suited to multi and hybrid cloud deployments than any other type of workload given the elastic nature of these applications. In fact, analytic workloads are some of the fastest growing workloads in the cloud precisely due to the technological and business advantages that cloud computing offers. 

The technology advantages are pretty apparent: improved reliability, almost unlimited scalability, and tremendous application performance. It’s when you consider the business advantages that it becomes clear that analytics should be a key piece of any multi-cloud strategy:

1. Accelerated time to insight. When it comes to data analytics, time to insight is paramount. Most organizations today are striving to use data insights for competitive advantage. Those that can do it fastest have the most advantage.
   
   Working in a multi-cloud or hybrid cloud environment, you have accelerated time to insight. You can develop, test, and deploy your applications faster because you're not installing, configuring, and provisioning infrastructure. You have immediate access to the latest compute technology, with better and faster performance than you may have in your proprietary data center. And you can use the right compute infrastructure for the right job.
   
   Without the cloud, you would have to make incredible investments in infrastructure in order to have a comparable range of choices, or else make some very difficult choices about which projects and workloads get prioritized. With cloud infrastructure, you don’t even have to decide. You can set an SLA or a response time you want from, let's say, a query, and software can determine which computer is available to meet your SLA at the lowest cost. This access to tremendous speed and compute on demand translates to accelerated time to insight.
   
   
2. Cost savings. Unlike a core business application which is continually running and uses a predictable amount of compute, analytics workloads are very dynamic because when you're doing analytics you are often reacting to specific requests. Your business may need to respond to macroeconomic changes; regulatory changes; weather; disasters; political events, or changes to the competitive landscape. You don't know what is going to happen, or when you're going to need insight. It would be extremely expensive to try to accommodate all the potential use cases in your own data center. For example, when you're doing things like machine learning, you're often looking for patterns or anomalies in large data sets. It’s very compute intensive; in fact, a lot of machine learning use cases were out of reach of most enterprises until cloud computing became available.
   
   The other cost factor is that a lot of the most important data that we want to use in analytics today is not within the four walls of an enterprise. If you think about all the smart devices in the world; or all of the online commerce that's happening, or all the transactions that happen in the B2B world, much of this computing is happening in the cloud. Therefore, that's where the data resides. Rather than moving data into your data center, if you have applications running in Amazon, Google, or Microsoft, and you want to do analytics against those data those applications create, it's a lot easier—and cheaper--to do it right there in the cloud.
   
   
3. Avoiding vendor lock in for maximum flexibility Even in the cloud, compute, storage, and networking are still the most significant line items in an IT budget. Avoiding cloud vendor lock in gives you flexibility and negotiating power. It also allows you to take advantage of the intense competition between cloud vendors—the big three, Google, Amazon, and Microsoft—and also Oracle, Dell, IBM, and VMware.
   
   Secondly, multinational corporations doing analytics on a global basis are often constrained by restrictions on moving sensitive data across borders. You want to be able to do your compute on a regional basis, so you want to have the flexibility of working with the vendor that can best address your needs in a particular country or region. Let’s say you have a machine learning algorithm that determines your most advantageous pricing, and you want to run that in both Europe and the Americas. Well, you can run the same application on different cloud providers in each region, using the data from their region.

It’s imperative for every company to have a multi-cloud and hybrid cloud strategy if they want their business to remain competitive. The advantages are just too great to overlook. We’re all familiar with the technology advantages--improved reliability, scalability, and performance, but more importantly, there are incredible business advantages.

If you leverage containers, and resource management technologies like Kubernetes, you can architect your applications to be portable across all cloud providers, including private cloud. And that portability and interoperability is how multi-cloud delivers tremendous business advantages—flexibility, negotiating power, cost savings, the ability to leverage innovation, and most importantly for analytics, accelerated time to insight. Whether from your data, customer data, partner data or third-party data, you get insights much quicker than if you're trying to do this either with a single cloud or in your own on-prem deployments. Every CIO wants to take advantage of that.