One of the bigger debates in the IT world the previous five years had been about whether to go public or private cloud. For the most part, the majority in the tech world believe that it is a "hybrid" world, one spread between on-prem and public clouds. Amazon AWS’s announcement of Outposts at its last re:Invent I attended (read coverage here) and Azure Stack sealed the direction of the industry.
The latest cloud industry discussion is around "multi-cloud," in which there are two schools of thought. One definition is where enterprises using many different private and clouds in IaaS, PaaS or SaaS service models. As simpler definition are multiple public cloud environments. Multi-cloud may not be the most efficient, and security models are still being figured out, but the reality is most are leveraging a SaaS solution like Salesforce or Office 365.
Regardless of the conversation, it's important to come up with some methodology to figure out the direction of which strategic cloud provider to pick. Based on the various infrastructure conferences I have attended the past year and talking to real customers, the following are a list of areas I think enterprises should be looking at to determine which cloud to use. This is a simple “starter” list and for the sake of brevity will keep it this way. There are thousands of lines of spreadsheets behind this which isn’t pretty which we have done enterprises looking for comparisons.
Hard cost savings
The first, and perhaps most obvious area an organization should be looking at when weighing cloud options is potential infrastructure cost savings. Unpacking and measuring your organization’s infrastructure costs is essential to determining the right cloud solution to suit your needs. I see this as broken down into three obvious areas: server, storage, networking, and facilities.
Server cost tracking should include the necessary hardware (including the server, rack, chassis, power supplies and power conditioners, top of rack switches) plus maintenance as well as the software side of the equation (operating system and virtualization licenses). Storage cost tracking should include hardware, such as internal chassis storage, NAS and SAN/Fibre Channel switches and cables, and storage software costs. Networking cost tracking should include hardware such as LAN switches, load balancers, and the recurring ISP and bandwidth costs. Facilities costs are gigantic, but sometimes get ignored in the calculations and should account for the power, cooling and the space itself. Oh, and don’t forget about the facilities taxes, insurance, and security. Of course, there’s more to be accounted for—many organizations must consider database, management, and other software costs.
It’s easy to run up a major cloud tab (many have), and the key is to have a transparent cost-management and governance approach. Anecdotally, I have heard from customers that the more they use, the more money they save as their “cloud IQ” rises. It takes some time to “right-size” compute instances, match with the right storage options, and get into the right upshifting and downshifting modes. Time on the till matters. The other thing to consider is that you will save more costs as you transform apps from simple “lift and shift” to “cloudifying” them, let’s say to containers and serverless.
Look for strategic cloud providers who are actively trying to help you reduce costs to you through education, intelligent cost portals and auto downshifting of services you don’t need. Avoid those vendors with minimal services and one-size-fits-all offerings.
Finally, be sure to factor any incremental costs incurred with a multi-cloud scenario. Multi-cloud is fine to get competing services bidding, but realize it also means that you will likely miss out of volume purchase discounts, may need to dumb down the apps to support multi-cloud, and will need extra staff to manage, train and code to the multiple clouds.
An organization should consider the cloud’s potential to improve staff productivity. Think closely about which cloud option is best suited for your organization to save your employees the most time and effort if that’s your strategic center-point.
There’s a myriad of traditional IT tasks that are either greatly simplified or no longer needed with the right cloud fit. Try and measure all of the time and effort that traditionally went into server budgeting, planning, negotiating and purchasing. Tasks such as longer-term capacity planning can be automated, as can the installation, upgrading, and removing of software. With the public cloud, IT no longer has to deal with hardware repair or failure. Consider the costs of security updates and patching. Security doesn’t go away with your strategic cloud provider, as you still need to manage security “in” the cloud, but your provider will provide security “of” the cloud.
Time is money—if IT staff isn’t having to focus on all of these issues, it frees them up to work on more value-added projects for your organization closer to the business units and end customers.
Another factor to consider is the uptime or operational resiliency a public cloud option can provide to an organization. The cloud can provide many benefits in terms of reliability, scalability, and security, and can save organizations big bucks in terms of the costs associated with outages and downtime. I can’t help but notice that the most high-profile breaches like Equifax and even Facebook have on-prem deployments. Five years ago, IT was worried about cloud security, but now they’re wondering how they can afford on-prem security as good as the public clouds.
Let’s just tick off some of those costs of downtime:
- third party contractors and consultants brought in to deal with an outage
- new equipment purchases and maintenance
- lost time and costs associated with incident detection, investigation, and recovery
- hits to both end-user and IT personnel productivity
- revenue loss due to business disruption (including damage to an organization’s reputation).
The cloud isn’t infallible, but its fault-resilient design and proper use of AZs are designed to avoid most of these issues.
Flexibility and agility
The winners and losers in business are no longer just defined by size as they were in the ’80s and ’90s, but their ability to move to where the revenue and customer puck is going. It’s not your imagination- change is increasing at a rapidly increasing pace, and only those with the ability to shift directions and speeds will win. Organizations should be cognizant of the business agility benefits the public cloud brings with it.
There’s a lot of different metrics organizations can monitor to get a handle on their “business agility and flexibility,” and many of them have the potential for improvement with a public cloud. These include, but are by no means limited to:
- number of new applications launched per year
- time-to-market for new applications
- response time to product and service defects
- time to provision new environments (compute, storage, networking)
- number of incidents and defects
- access to game-changing bleeding-edge capabilities
Business agility can also pay dividends in terms of employee and customer satisfaction and retention and reducing absenteeism in the office. Generally speaking, the quicker an organization can innovate and respond, the happier everyone is, inside and outside the company. While I don’t underestimate the comfort of sticking with what you know with older systems, this is now an innovation trade-off.
While there are so many factors to consider, there’s potential for a lot of positive business impact in choosing a strategic cloud partner…. if you ask the right questions and track everything that matters. My simple advice to organizations is to look for a strategic public cloud partner that can maximize all of these benefits for your organization—hard cost savings, IT productivity, uptime, and business flexibility and agility. Determine where your organization is at in these areas now, and where you’d ideally like for it to be. The right public cloud can likely help you achieve that.