The three phases of a successful cloud migration

By Allen Bernard, Business Technology Journalist

Moving to cloud is not a decision to take lightly. Having the right plan can make all the difference.

Even though moving to the cloud may seem like an easy decision these days, it isn’t an activity to be undertaken lightly. Off-loading infrastructure and applications to a third-party provider have many such as improved business agility and a more flexible IT budget, but it’s important to first explore and understand the impact moving to the cloud will have on your business processes and operations at a very granular level.

While something will likely go such as the unintended disruption of an essential business service or the loss of critical data, with a little forethought and planning, many problems can be avoided.

Having helped many clients transition to the cloud over the years, we’ve learned a few things about what works and what doesn’t. While not exhaustive, the list of considerations we present here should give you a good starting point to begin your deliberations.

Once the decision has been made that cloud is the best option to solve the business challenges you face (and it may not be), there are three phases of moving to the cloud: planning, mid-shift, and go-live. Of the three, planning is probably the most important yet often overlooked part of the process.

Planning

During the planning phase, it is important to first define which applications and infrastructure will be moved and to map the application and database dependencies associated with each application and piece of hardware.

You’ll need to identify why these things are important from an operational perspective, which, surprisingly, many organizations do not take into consideration (or at least as much as they should). The operational pieces make up the baseline requirements that support a successful cloud migration. If you fail to think of these things from the start, you can still move to cloud but you may not realize the benefits and you could create more work for your IT team while negatively impacting the user experience.

If you are adopting software-as-a-service to replace an existing application, then this exercise is less important. But you still need to map out all the application dependencies that may break as you migrate off the old application and on to the new one.

It is also a good idea to undertake an asset management discovery process to get a high-level view of where all the hardware and software in your organization resides. That way, when a server or application is shut down and a service no longer works, you will have a roadmap to help to find the root cause quickly.

From a connectivity standpoint, moving from an on-premise solution to a cloud-based solution creates higher demand on customer connectivity and will potentially require increased bandwidth. It’s a good idea to check your application guidelines for minimum data speed recommendations and make sure they will be available when the application is hosted in the cloud.

Because cloud is not some magical place in the sky, but simply hardware hosted in somebody else’s data center, you will want to know physically where your application will be running and the networking that is available to move data between it and the outside world. With new privacy regulations

With ransomware and other cyber-crime on the rise, cybersecurity also is an issue that must be explored in-depth. Moving to the cloud will expand your network edge exponentially as it grows to incorporate new cloud providers and all of their employees and partners while simultaneously reducing your direct control over those assets and data.

Another important consideration is the user experience. If the application is not performing up to expectations and users are unhappy, this will negatively impact your operations and, potentially, sales.

Basically, during the planning phase, you want to run through all of the things that could go wrong and how you will respond to them. This means you will also have to take into account your business continuity and disaster recovery plans and how moving data to and from your backups will be impacted.

Mid-shift

The mid-shift phase is where things can get a little messy so you will want to prioritize your cloud migration over other projects until it is complete. As applications and functionality are being moved physically from one place to another, first to see what the impact will be on a limited set of users in real-time. That way when go-live happens you avoid any unforeseen problems.

You also will want to make sure that you have over-scoped your cloud resources so that you will not hit contracted capacity limits too quickly. We are working with one client, for example, to remedy a situation where the enterprise architect they hired did not scope enough capacity, which is leading to big-budget overages.

Another thing to think about during the mid-shift is identifying any friction that has worked its way into your processes from the culture shift that cloud brings with it. This could be something as seemingly simple as introducing multi-factor authentication for employee logins. The key here is making sure your employees get as much training on the new systems as early as possible in the process.

Go-live

Mid-shift and go-live share similar concerns. Once the application is in production you want to make sure that it is operating correctly and that people have access to the functionality and data that they need. From a customer success point of view (which is how we like to look at things) you’ll want to make sure that your move to cloud is achieving the business results you expected such as reduced IT costs, faster time to market for new products, and services, or a better user experience for your employees and customers.

Of course, everyone knows measuring KPIs after the fact is not often done. But it is still a good idea to do things such as reviewing trouble tickets to see if the volume has increased. Also, look at the types of complaints people have to understand how the application is performing. You also can engage your employees with surveys to see if they are getting the benefits you planned for.

You’ll also want to follow the money to make sure that you’re not being over-billed for instances of compute, storage, networking, or services that you don’t need, want, or don’t know about. This can be exceptionally difficult to discern given that cloud providers often send invoices with millions of line items. Finding a good expense management solution or service provider is something to consider to help short-circuit the process.

Perhaps most importantly, it’s a good idea to remember that cloud is not a panacea. It is simply a way to ease the management burden of technology so your IT organization can spend time providing you with the right technology you need to thrive instead of just keeping the lights on.