Nearly every major digital enterprise is in the process of adopting cloud computing as a core component of its technology strategy. But despite this overwhelming cloud momentum, the financial models illustrating the business impact of cloud adoption share a critical risk factor: the time and costs to migrate to the cloud are largely unknown and often grossly understated.
The main reason? Many companies fail to hire financial planning experts who understand the technologies, processes, and complexities required to complete a migration budget. Unfortunately, there is no set blueprint, and many companies still insist on shooting from the hip when planning a cloud migration. Worse, others still believe every cloud migration is a simple “lift and shift” effort. After 23 years of financial planning experience in one capacity or another, I can say with a high degree of confidence that a cloud migration effort is seldom that simple.
Case in point, a 2017 Forrester Consulting study found rampant “cost complexity” in the cloud, which sows confusion and leads to misaligned expectations. “Most firms today underestimate costs associated with cloud usage and migration,” the company said in a press release.
So let’s deal with the macro-economics first. There are three general categories of costs that need to be addressed:
1. Current, or business-as-usual, application and infrastructure costs, including the direct and indirect costs of using and maintaining your on-premise IT investment over time, and not just what you pay for infrastructure. Direct costs cover hardware and software (including physical servers), software licenses, maintenance contracts, warranties, supplies, material, spare parts, network bandwidth, storage and database capacity, all labor, and facilities. Indirect costs include loss of revenue and productivity resulting from any outages or downtime. This step is essential to fully understand the business case for cloud adoption and make an apples-to-apples cost comparison.
2. Qualitative Financial Estimate of your cloud infrastructure costs. The top cloud providers all supply pricing calculators to help with this:
- Amazon Web Services (AWS) Total Cost of Ownership (TCO) calculator
- AWS monthly cost calculator
- Google Cloud Platform pricing calculator
- Microsoft Azure’s pricing calculator
- Rackspace’s calculator
- IBM Bluemix’s calculator
These calculators typically address your needs for (virtual) servers, storage, and other services, and sometimes allow you to include advanced details in your calculations, such as labor and networking costs. Obviously, the information from your current environment will be essential to estimate your cloud needs. Note that this is where many businesses make most of their cloud-costing mistakes. One big reason: they fail to include internal headcount required to run the cloud environment and usually under-subscribe to meet their cloud capacity requirements.
3. Cloud migration costs. The third key factor is the cost of the migration itself. That includes the cost of moving your system(s) and related data to the cloud, either via the internet or by using a physical device such as AWS Snowball or Snowmobile. You also have to budget for the time it takes to port your apps to the cloud and account for whether you’re doing a simple “lift and shift” or more complex refactoring to take advantage of advanced cloud capabilities. Don’t forget to consider the cost of creating this inventory—along with a full dependency mapping.
In fact, a 2016 Forrester report estimated that more than half of cloud migration costs are due to labor, not infrastructure and platform services costs. In addition, many organizations use cloud migration automation software, and employ systems integrators and other consultants to help with the move, and those costs must also be accounted for. Finally, security procedures are different in the cloud, and organizations must update their security efforts to fit their new environment.
The microeconomics of cloud migration
A financial plan for a cloud migration requires you to build a timeline for migrating and cutting over each of the target systems, understand your labor resources, and know the weekly operating costs for both the original and cloud environments. You might break things down by hour for smaller migrations that require four weeks or less.
The length of time to complete the migration should be known, measurable, and tracked, because someone will need to explain any variances to plan (and there are always variances). Tracking these from the project outset will significantly reduce the risk to the entire migration as you quickly identify common or recurring issues. Often overlooked, this step is arguably the single greatest cause of cloud migration budgets that go off the rails. Simply put, “We don’t manage what we don’t measure!”
Post-migration costs and the importance of benchmarking
On that note, here’s a key piece of advice to de-risk your cloud migration plan: benchmark everything in the target environment. Without solid performance benchmarks you are virtually guaranteed to miss the target environment retirement date.
That’s because you can’t complete an effective cloud migration plan unless you know when your cloud environment has reached operational parity with your target benchmarks and it’s OK to retire the original systems. So it’s not enough to understand how much you’re spending on your current environment; you also need to understand its performance. Otherwise, you can easily end up running parallel environments—on-premise and in the cloud—well beyond the expected timeline.
Put it all together, and the simple math for calculating the cost of a migration equals the sum total of the parallel environments divided by two, plus software, network, and labor costs. The result of that number is then divided by the number of weeks (or hours) required until “lights out”—the moment the original target environment is finally retired. To reiterate, it’s critical to track not just the total migration cost but also that weekly cost.