An economic framework for cloud migration (part 1)
More organizations than ever are looking at moving some or all of their technology workloads to the public cloud. But for many, the process behind migration is complicated. Although the operational benefits offered by the cloud are well understood, business barriers remain that prevent adoption.
Following extensive discussions about this challenge with Apptio’s customers and our partners in the cloud provider community, I want to share a framework that offers a pragmatic approach to increasing your usage of the cloud.
»Related content: "Accelerating migration: from 'random acts of cloud' to the new normal," a webinar with AWS and Deloitte
There are lots of decisions that need to get made as you think through what your cloud migration will look like.
- Will you pick a public or private cloud model? Will you straddle both with a hybrid operational model?
- Which workloads get prioritized? Low-risk dev and test workloads, or new applications that will drive innovation for the business?
- Will you move your applications to the cloud as-is, in a “lift and shift” model, or refactor your key applications to take advantage of cloud-native architecture?
- How does your developer strategy evolve with cloud? Do you have plans to adopt agile processes in a DevOps-style culture?
And likely, these questions just scratch the surface. For each of them, you’ll factor in impacts to your architectural patterns and operational practices.
At Apptio, we help our customers understand the operational and architectural impacts within the context of economics. For each of those decisions, and the many more that go along with them, the economic impact of your cloud migration should play a key role in decision making.
As you’d probably expect from us, our framework is focused on the economic aspects of cloud migration. But Apptio works with great partners that can help with the architectural and operational aspects of the migration as well - reach out to me and I’ll be happy to connect you with people that can help.
»Hear Brent Eubanks, VP Technology Optimization at Live Nation, share his cloud optimization experience at the TBM Conference on Thursday, 11/9. Register here.
The economic framework for cloud has four stages:
In this installment of the blog, we’ll cover the first two phases, and then come back in a couple weeks to cover the next two phases. Breaking it up this way makes sense because, while it’s possible you’ll repeat the first two phases for different migration projects, the latter two phases will almost certainly be repeated as an ongoing governance process and act as a ‘guide’ along the path of your migration.
Step 1: Pre-migration
Before you start a cloud migration at any scale, it’s important to have a plan and an objective that will guide your overall “thesis” on the migration. This is effectively where you make sure you know what you already have.
If your plan is in part to shift the balance of spend towards OpEx and away from CapEx, you need to know what your costs are today. If part of the migration is retiring old and end-of-life infrastructure, it’s important to know where that infrastructure is, and the applications it supports.
This process of baselining is critical for two reasons: first, it helps tell you what the opportunity surface is for the migration. Second, it establishes the baseline from which you’ll measure the success of your cloud migration once you start moving workloads. Without it, you’ll never be able to quantify the business benefit of the cloud migration.
The other main outcome at this stage is to understand the key drivers of your migration. Even if you plan to move your entire technology portfolio to the cloud, it won’t happen all at once, and there are shortcuts available to help prioritize which workloads get moved first, particularly from an economic point of view.
These transformation drivers essentially start building the blueprint you can follow to naturally move your business to the cloud. Imagine your data center footprint today—maybe it’s using leased capacity from a colocation firm. If you have a lease ending in the next 24 months, that could be an inflection point that allows you start shifting that co-located infrastructure to the cloud.
As another example, maybe you’ve got a long-term enterprise-style purchase agreement with a legacy software provider. If part of your migration plan includes reducing your dependence on those providers and adopting new and more modern cloud-native services, an upcoming expiration of that contract could enable you to focus on building a migration plan to move away from those services.
As you conclude this stage, you’ll ideally have two key sets of information. First you’ll have data on what you have today (and granular total cost information) to create a pre-migration baseline. You’ll also have a roadmap of sorts that illustrates the inflection points in your organization that can drive an initial prioritization effort for your migration, one that maximizes your ability to take advantage of the innovation brought by cloud but minimizes the impact of long term spend duplication.
Step 2: Migration business case
When most people think of the economics of a cloud migration, this is the phase to which they’re most commonly referring. In fact, many of you have probably already done an exercise like this with a cloud provider (or two).
At this phase, you’ve probably been working with one or more cloud providers for a while, and gotten a sense for their individual strengths and weaknesses. Unlike the first step, where you may have already established a baseline (especially if you’re an Apptio customer), when you begin the business case process, you’ll probably do it in conjunction with a cloud provider and often an additional partner. This step is the first point at which you’ll establish the projected benefit of a cloud migration across all three axes—operational, architectural, and economic.
The most accurate and realistic business cases will begin with an actuals-based baseline total cost of ownership (TCO). The same kind of baseline you (hopefully) established with Apptio in the first step. If you did, you’re way ahead of the game. Without an accurate baseline, you’ll be left to substitute estimates based on recommendations from the cloud provider or their partners.
Apptio has close relationships with a number of migration partners that can leverage Apptio’s actuals-based baseline for the business case process. Companies like TSOLogic, often engaged by AWS, as well as the major SIs like Deloitte and Accenture, all have experience using a customer’s Apptio data to power their business case projects.
These business cases will usually look at the characteristics of your infrastructure as deployed today—this includes both configuration data (how many vCPUs, server location, operating system, etc.) as well as performance data (min / max / average CPU utilization, consumed storage, etc.).
The best tools factor this data into sophisticated models developed in conjunction with the leading cloud providers to create a future-state configuration on the cloud of choice. These models usually take into account an optimized configuration—along with pre-purchase commitments, volume discounts, and more advanced Platform-as-a-Service (PaaS), these configurations are the key inputs that enable the public cloud to save customers 30-35% over their on-premises deployments.
Keep in mind that business cases tend to be a point in time snapshot, and may not include a model to showcase cost of migration over time, the effects of developer transformation, or etc. But working closely with many leading cloud providers and SI partners allows us to add this additional context to cloud business case exercises.
At the close of step 2, you’re probably looking at a business case artifact that demonstrates a massive savings opportunity for your business if you move to the cloud. Hopefully that business case was developed by comparing the provider’s estimates to highly-accurate on-prem costs from a solution like Apptio.
If you’re one of those customers that sees potential cost savings in the 30%+ range, you’re probably pretty excited! And for good reason - those cost savings can then be re-invested to support innovation projects. You’ll get the operating leverage that’s so often associated with the cloud, and still be saving money previously locked up in long-term CapEx commitments.
But in reality, your cloud migration is just beginning - migrations take a lot of time! In our next installment, we’ll explore what happens after you’ve decided to migrate. We’ll look at ways to ensure you’re getting the most value out of your migration, and techniques you can use to capitalize on the fungible nature of cloud-based infrastructure through ongoing optimization.
»Want more on this topic? Try this helpful executive brief from Apptio: Demystifying cloud costs.