Essential steps to reach effective cloud cost management

Here’s how IT leaders can control and reduce cloud costs without sacrificing value.

In brief

• Cloud costs consisting of compute and data, monitoring and operational expenditures can fluctuate over time if left ungoverned.
• Companies that take proactive steps to gain visibility, set budgets, optimize architecture and measure success will dramatically reduce cloud costs.
• Keeping cloud costs low helps CIOs accomplish more with less IT budget and pass savings on to the business.

Often, companies migrate to the cloud expecting to reduce overall IT costs, but public cloud spend continues to rise. Many enterprises and their technology leaders can be caught off-guard when cloud costs spike. Without a detailed plan in place, these same cloud costs can be very challenging to keep under control. So, if cloud services are designed to help companies save money, why does it appear to be costing more?

Common costs of cloud services

An ungoverned cloud can rain money. When an organization sets up its cloud programs, much of the budget focus is predicated on what fixed costs it will take to run over time, but cloud environments are fluid, and there are many factors that cause costs to fluctuate, like data transference, redundancies, organizational misalignment, talent and lack of automated controls. Additionally, EY teams often notice that some expenses, like shared services, information security and test environments, were left out of initial business cases and budgets. These miscalculations eventually lead to cost overruns. IT leaders need to ask, “How can we proactively prevent cloud costs from spiralling out of control?”

While everyone knows that there will be hard costs associated with cloud management, these costs can rise and fall depending on maturity.

1. Gain visibility

The first step is to aggregate, socialize and normalize all cloud usage and cost data for the required teams. IT teams can gain visibility by focusing on a few key tasks:

• Define and enforce a tagging strategy – A cloud tagging strategy formulates precise rules and practices that an organization agrees to follow and implement. This strategy helps users understand how to create and use tags, and how to make tagging decisions. These policies then need to be enforced through automation to maintain compliance.
• Organizational alignment – It’s critical that CIOs speak with each department that touches the cloud, particularly, procurement, finance and IT to align and define cloud governance models. Comprehensive cost reduction strategies are similar to change management programs, wherein IT teams need to communicate clearly and often with other business units to maintain alignment and work toward a shared goal.

2. Formulate budget

The next step is to define usage and budgets:

• Create consumption models – How much does a company expect an application or platform to consume? How will that affect demand and capacity? Creating consumption models will help set benchmarks and control overspending. Deciding which tasks can be automated will avoid overlap and repetition. Also, auto-scaling features may be available to dynamically adjust resources based on demand.
• Communicate budget – All team members should understand the budget and expected consumption models to avoid costly pitfalls and should elicit cross-department feedback.
• Create alerts – Alerts provide timely awareness of incurred cloud application costs so that they can be addressed or resolved quickly.
• Discounts – Appropriately utilize reserved instances and spot instances. Reserved instances are a specified level of cloud usage from a vendor that has been committed at a discounted rate compared to on-demand resources. Spot instances are unused computing space offered by suppliers at largely discounted rates. Low-performance storage options will discount costs as well. Any of these discount methods can help optimize a cloud budget.

3. Focus on cost implications of architecture decisions

The third step is to account for design and run time when selecting or changing cloud architectures:

• Perform an in-depth analysis of application services – This process will provide a clearer picture of usage and resource peaks to understand whether each application is cost-effective within the cloud. Based on the results, consider eliminating or resizing unused or underutilized resources.
• Cloud environment – Another way of optimizing the environment is to use the type of cloud or clouds that work best for the overall IT strategy. Distributed and industry clouds are becoming more widely used models:
  • Distributed cloud: A distributed cloud helps organizations connect applications and data from various geographical locations. It optimizes performance by taking the best of public cloud, hybrid cloud and edge computing. 
  • Industry cloud: Many large business conglomerates are building sector-specific cloud solutions, or industry clouds. Because of their deep sectorial knowledge, these solutions are customized and address hard-to-tackle vertical challenges. Many companies are building these solutions not only for in-house usage, but also as potential future income sources.
    
• Contracts – Avoid vendor lock-ins by using a hybrid approach or cloud management tool that allows for data and application mobility within contracts. It’s important to maintain flexibility as business and data needs evolve.