The economics of cloud adoption
The potential value of cloud services is tremendous. However, the cloud is more than just a new way to do things; it's also an economic model that requires careful consideration.
Cloud adoption can be facilitated by a robust strategy that considers that merely transferring information from on-site servers will not ensure seamless or any functionality. The economics of running applications in the cloud differs from more traditional models, something many businesses fail to consider when planning their cloud migration strategies.
By following some of the most important lessons learned by McKinsey, businesses can ensure a successful transition to the cloud.
Сhoose long-term benefits over short-term gains
When some companies move to the cloud, they often simply migrate existing applications with minimal or no redesign. Such approach means that most of the technical and operational inefficiencies of the old applications are retained, preventing a business from taking advantage of the cloud's dynamic infrastructure.
The benefits that can be reaped in the short term pale in comparison to those that companies could capture in one year, such as time to market, access to advanced capabilities, and innovation. Investing more time in foundation development, app remediation, and automation could lead to a 15% – 25% increase over short-term gains.
Move towards an operating expenditure model
Traditional IT infrastructure for most enterprises is based on a capital expenditure model with long-term demand planning. Costs for consuming additional capacity are minimal, and companies use average costs and utilisation rates to evaluate their cost-effectiveness.
On the contrary, cloud service providers have shifted the paradigm to an operational cost model in which businesses pay for what they use. Therefore, companies must adapt and learn to develop a dynamic approach that optimises their marginal costs by selecting cloud services that best meet their current workload requirements.
Develop a priority-based budgeting approach
When companies move to an operational expenditure approach, they often budget for future spending based on past patterns. It can result in more than a 20% discrepancy between forecast and actual expenditures, resulting in poor allocation decisions and the need for difficult rebudgeting.
Organisations should tie it to business priorities for better forecasting and budget planning for the cloud. Effective forecasting requires organisations to establish unit economics for their core applications, such as calculating customer cost per customer. This approach requires a shift in mindset towards a consumption-based model and a competent finance operations team capable of understanding the business drivers of cloud spending and its impact on unit economics.
Make sure cloud services are worth the investment
Successful cloud adoption can be especially useful for workloads with variable consumption to allocate resources more effectively. For example, a video-streaming company that analysed the relationship between the cost of cloud services and business demand drivers accurately predicted future cloud consumption 95% of the time.
Companies that fail to differentiate between workloads that would benefit from on-demand scaling and those that wouldn't end up missing out on potential savings. Thus, companies need to evaluate their workloads individually to determine if elasticity patterns could help them save money on the cloud.
Separating the economics of cloud computing from its architecture
Businesses should be aware that they often overestimate their cloud utilisation. It can lead to increased expenditure because most companies end up with lower use rates than expected. While some enterprises with advanced cloud architectures have utilisation rates of over 60%, most companies fall below 30% and sometimes below 10%.
An infrastructure supporting higher use rates makes it possible to achieve them. However, all too often, the business's cloud economics and architecture are developed independently, resulting in utilisation rates that the underlying infrastructure cannot support. To avoid this problem, companies need to establish a close relationship between the cloud business case and the transformation of their architecture.
Define and prioritise workloads for cloud migration
It is not recommended that every workload be moved to the cloud. Some workloads, such as storage services on custom-designed on-premises infrastructure, can be more valuable if left on-site. Companies with a small number of massively scaled workloads need to be careful when deciding which ones to migrate to the cloud.
Sure, the economics of cloud adoption involve a complex interplay of factors ranging from cost analysis and operational efficiency to strategic planning and architectural considerations. Here's a breakdown of the concepts mentioned in the article:
-
Long-term Benefits vs. Short-term Gains: This concept emphasizes the importance of investing time in foundational development, application remediation, and automation during cloud migration. Short-term gains from simply transferring existing applications to the cloud might not match the long-term advantages that a redesigned, optimized approach could offer.
-
Operating Expenditure Model: Contrasting traditional capital expenditure models of IT infrastructure, cloud services operate on an operational cost model. Companies pay for what they use, necessitating a shift in budgeting and cost evaluation strategies.
-
Priority-based Budgeting: Moving towards an operational expenditure model requires a shift in budgeting practices. Forecasting based on past patterns might lead to discrepancies; instead, aligning cloud spending with business priorities and understanding unit economics becomes essential.
-
Worth of Cloud Services: Evaluating whether cloud adoption is worth the investment involves assessing workloads and their consumption patterns. Workloads with variable consumption patterns can benefit significantly from on-demand scaling, leading to effective resource allocation and potential cost savings.
-
Separating Economics from Architecture: Many businesses overestimate cloud utilization rates, leading to increased expenditure. Aligning cloud economics with the architectural design can optimize utilization rates and prevent overspending.
-
Workload Prioritization: Not every workload should be migrated to the cloud. Certain workloads, especially those with specialized on-premises infrastructure, might be more valuable if kept on-site. Prioritizing workloads for migration ensures a strategic approach to cloud adoption.
These concepts collectively highlight the strategic, financial, and operational considerations that businesses need to address while planning and executing their transition to the cloud. Successful cloud adoption requires a holistic approach that encompasses technical optimization, financial planning, and strategic decision-making.
