Skyhigh Cloud Costs: Is It The Cloud Itself Or How We Use It?

Skye, the new CFO, is staring at cloud costs as a material line item on this year’s balance sheet. Her eyes are watering as she watches the company’s spending on cloud services skyrocket month after month. 

Skye knows the cloud can save them money, but these numbers are making a material dent in the budget. What gives? Is the cloud inherently expensive? Or are they using it wrong? This month, we’re tackling unexpected cloud costs, exploring public, private, hybrid, and multi-cloud models, and revealing actionable strategies to turn your balance sheets into a more pleasant read.

Is Skye using the cloud wrong?

Skye isn’t the only CFO asking this question. According to a Flexera study, 76% of organisations admitted their cloud spending was higher than expected. At the same time, Raconteur’s survey shows that 30% of cloud spend is wasted on underutilised resources, overprovisioned systems and orphaned workloads. (Read more)

Why is this happening?

  • Without clear visibility, it’s easy to lose track of the big picture as to what’s being used and what’s not 
  • Overprovisioning resources means paying for more than you need, while underprovisioning can cause performance issues
  • Not leveraging reserved instances, spot instances, and other cost-saving strategies can result in unnecessary expenses​
  • Departments that manage cloud resources in silos lead to duplication and inefficiency

Understanding cloud consumption models

Skye is considering four main cloud deployment models to take control of the situation. Of course, the optimal model depends on the needs and priorities of Skye’s organisation, so let’s examine each closely.

The Public Cloud

Going all-in on the public cloud makes the most sense for some organisations (at least for specific workloads). Major players like AWS, Azure, and Google Cloud provide scalable solutions, robust services, and increasingly sophisticated cost management tools within the public cloud landscape.

Pros:

  • Only pay for what you use
  • No hardware overhead
  • Leverage cloud-native architectures and pricing models like spot instances and reserved capacity to optimise spending.

A real-world example shared on Reddit highlights how a Kubernetes environment on AWS cost $20k per month before optimisations like leveraging spot instances brought those costs all the way down to $5k.

The catch? You lose some of the control and predictable costs of private infrastructure. Plus, managing public cloud spend requires diligence in areas like right-sizing resources, eliminating idle waste, accessing reserved pricing, and monitoring usage across accounts and teams.

The Hybrid Cloud

Many businesses embrace a hybrid cloud approach to rein in runaway cloud expenditures. This involves combining on-prem private cloud resources with public cloud services to create a best-of-both-worlds solution.

In reality, Skye’s organisation can leverage the private cloud for sensitive or high-stakes workloads that require elevated security and performance. While the public cloud can be used for less critical workloads and to take advantage of cloud bursting during periods of high demand.

Pros:

  • Optimise workloads by choosing the best environment for each task
  • Reduce upfront costs with the public cloud and leverage existing infrastructure with the private cloud
  • Quickly scale resources up or down to meet fluctuating demand
  • Maintain sensitive data on a private cloud while enjoying the public cloud’s agility for less critical workloads

But hybrid clouds aren’t perfect. Managing and integrating the different environments can be complex. Skye’s teams will need robust tools for visibility across your hybrid landscape. And while a private cloud avoids some public cloud costs like data transfer fees, Skye will still need to factor in hardware/software costs and overhead.

The Multi-Cloud

In this approach, Skye’s organisation can distribute workloads and data across multiple public cloud providers to avoid vendor lock-in, leverage best-of-breed services, and mitigate risk. The goal is to optimise performance, costs, and business needs.

Pros:

  • Quickly adapt to business needs by leveraging the strengths of different cloud providers
  • Avoid single points of failure and vendor lock-in by distributing workloads across multiple clouds
  • Choose the ideal cloud provider for each workload based on performance, latency, and cost
  • Tap into the latest features and services of providers to drive innovation

For example, Skye’s enterprise may run containerised applications on AWS, use Azure for Office productivity workloads, and leverage AI/ML services from Google Cloud to take capitalise on each provider’s unique strengths.

But again, no cloud consumption model is perfect. While Skye gains flexibility and redundancy, the multi-cloud also adds overhead in managing disparate environments, dealing with pricing complexities across providers, and handling data transfer costs. Skye will also need to rely on robust skills across multiple vendor platforms and services.

Public, Hybrid, or Multi: Which cloud is for you?

The cloud itself is not inherently cost-prohibitive. Yes, the pay-as-you-go model can lead to unexpected sprawl, but runaway spending results from how businesses consume and manage (or mismanage) cloud resources.

By selecting the optimal deployment model based on your application portfolio, workload requirements and skills, implementing ongoing governance through FinOps and leveraging the latest optimization techniques – the cloud can be a cost-effective enabler of business transformation.

Kia Retail Case Study

Kia Retail, a leading e-commerce site, faced rising cloud costs and sluggish website performance. Despite internal efforts, no solution was found. They partnered with +OneX’s Cloud Cost Optimisation Service, which thoroughly analysed Kia Retail’s cloud usage. 

+OneX identified areas of overspending and recommended a serverless architecture to boost performance and reduce costs. Then, the team implemented a comprehensive strategy, shutting down unused resources, migrating to serverless, and right-sizing instances. 

The results included:

  • Consolidated Kubernetes clusters on GCP for 30% cost savings
  • Removed orphaned cloud resources
  • Implemented scheduling and start/stop policies
  • Leveraged spot instances and committed use discounts
  • Gained deeper visibility into cloud costs across environments

How +Onex can help CFOs

At +Onex, we understand the complexities of cloud cost management. Our team of experts can help you:

  • Assess your current cloud environment: We’ll analyse your cloud resource consumption and identify areas for optimisation
  • Develop a cost optimisation strategy: We’ll create a tailored plan to reduce your cloud spending without compromising performance
  • Implement cost-saving measures: Our team will help you implement the chosen strategies for sustainable cost reductions
  • Monitor and optimise continuously: We’ll monitor your cloud usage and adjust as needed to ensure long-term cost efficiency

Book a free consultation to gain control over your cloud spending and unlock the full potential of the cloud for your business.