How do you optimize cloud costs with your teams and stakeholders?

Ensure that all your relevant stakeholders are aware of each other, working together, and pulling in the same direction. This includes FinOps, ITAM, Procurement, Finance, and Engineering - and also the executives that those teams report into. Having strong communication, good understanding, and a clear organisation wide strategy is key. Focus on spend efficiency not cost reduction. Cloud spend going up can be a good thing - if it is because you're onboarding new customers or delivering additional revenue generating services etc. Understanding the business goals will help you create a realistic picture of what your cloud spend should be doing.

At the start of the project, it's crucial to have a clear understanding of the business goals. The business partner should articulate these goals to the Cloud Team explicitly. It's essential to gain clarity on why we are opting for a cloud solution for the specific issue at hand. Based on these considerations, the Cloud Team can develop a matrix comparing investment to the value of return.

Optimizing cloud costs is critical for any organization using cloud services, as it helps ensure the organization is not overspending on resources. I would suggest the following steps: 1) Understand the cloud cost by provider billing tool(s), cost analytics services, and detailed report to breakdown the expenses by services consumed. 2) Set the objective by setting up optimization goals and Key Performance indicators (KPIs). 3) Continuous monitoring and analyzing the cost by tagging and labeling the resource/services accordingly. 4) Adjust the resources by rightsizing. and 5) Implement cost alerts and educate the team by involving representative from IT, finance, engineering, and other stakeholders. In end measure the progress regularly!

It has to start with true collaboration and a clear understanding of why your looking at your cloud costs. It also requires not only a mindset change from the relevant individuals, but it's a culture change as well. But overall it's about making people accountable for what, why and where they are spending.

We need to analyze cloud costs at the resource level. For instance, consider how many VMs we're using and the purpose behind their use. It's important to examine how much CPU or RAM these machines utilize and determine if they are underutilized. Similarly, we should evaluate other resources to understand their actual usage and ascertain whether they are being utilized optimally. Questions to ponder include: Do we have the appropriate scale-up and scale-down mechanisms in place? Is our shutdown policy effectively implemented? Are we thoroughly testing resources before deployment? Have we equipped the development team with the knowledge to write cost-effective code?

To effectively monitor and analyze cloud spend, organizations should implement a solution that provides real-time visibility into their cloud usage and costs. This solution should also be able to identify and flag any anomalies or areas where costs can be optimized. There are a number of solutions available that offer these capabilities, including those that are bundled with security features.

Resource hibernation aims to shut down unused or underutilized cloud resources when demand is lower. For example, R&D environments can be stopped outside working hours and automatically resumed when needed. Build automation to schedule resource stop and start times based on usage data and metrics. Cloud management platforms like AWS CloudFormation can assist. Mandate that eligible resources implement hibernation policies except in documented exception cases. Make sure the process is enforced. Monitor resource utilization continuously and tweak(Long weekends) schedules to maximize hibernation while not affecting uptime requirements. Show cost savings from hibernation policies in cloud governance reports. Demonstrate impact and get buy-in.

Set Clear Goals: Define precise cost optimization goals aligned with your business objectives. This provides a clear direction for your optimization efforts. Usage Monitoring: Implement robust cloud monitoring tools to track resource utilization. Identify underutilized or idle resources. Cost Allocation: Allocate cloud costs to specific departments or projects. This accountability encourages teams to optimize their usage. Resource Sizing: Right-size resources based on actual usage patterns. Avoid over-provisioning and scale resources as needed. Reserved Instances: Explore options for reserved instances or savings plans offered by cloud providers. Tagging Strategy: Develop a consistent tagging strategy for resources.

The Cloud Team should conduct monthly reviews of cloud costs with the relevant stakeholders. Such engagement facilitates a closer understanding of cloud bills and allows for the identification of areas for improvement. It's essential to communicate clearly with stakeholders and involve them in optimization efforts. This not only helps the organization manage cloud expenses better but also promotes transparency.

Make daily cloud spending visible to everybody in the organization, especially to engineering teams. Not to slap them on the wrist when cost goes through the roof. But to help them understand the financial impact of architectural decisions and find new ways to optimize the use of cloud resources.

Periodically review the architecture of systems to adapt the components to new cloud solutions. Every year new out of the box components are coming to availability, a continuous transformation strategy will lead to cost reduction and architectural simplicity.

Cost and business model should align. Find the dimension you’re going to make money over, then make sure the architecture follows the money. For example you can architect your e-commerce application so that when the number of orders go up, infrastructure and operation costs rise. You shouldn’t aim to minimize your spend but aim to maximize the value. Cost optimization is a continuous journey and incremental. Unobserved systems enable wasteful habits and lead to unknown costs. The return on investment in observability typically far outweighs the expense. Make critical metrics (utilization, spending, errors) visible to engineers. This enables ongoing inspection to spot excess spend and tune operations to trim expenses.