Cost Engineering
The biggest factor in cloud infrastructure costs is not your choice of cloud provider, what region you select, how many users your service supports, or how much data it processes.
The single biggest factor is how effectively your solutions are engineered to make efficient use of the cloud platform.
Nearly all savings opportunities require engineering expertise. Exceptions to this are contract discount negotiations and support plan structure. The majority of cloud cost savings will come through investment in engineering efforts.
Endless Optimization Opportunities
For any solution design there are always opportunities to decrease infrastructure costs.
Services designed, built, and implemented without thought given to cost will likely have ‘low-hanging fruit’, or easily optimized components, that can lead to cost savings with relatively little effort.
For thoroughly optimized services there always still optimizations that could be made, however, the effort to make those optimizations may be greater than the savings. Spending more time on the cost optimization than is returned to the company in savings is not a good investment, especially when you account for opportunity cost where those efforts could have been spent enhancing the product.
Examples of Cost Optimizations
Some of the typical engineered optimizations include right-sizing instances, turning them off when not needed, leveraging auto-scaling mechanisms, or using standard disks instead of SSD when performance requirements allow. These may be low-hanging fruit, or they might require a substantial engineering effort of performance monitoring and load testing. Some legacy systems may not take well to servers dynamically joining/leaving clusters.
In greenfield design, there may be substantial cost savings in not using traditional compute instances at all. Instead, you may use managed platform services to build your solution. Examples of this may be the use of file storage services, serverless functions, or content delivery networks (CDN).
For systems that utilize these platform services, optimization of each service will require becoming an expert on that service; both its technical and billing attributes. This understanding, along with context of your business and the service your company is building on these platform services, are necessary to make most effective use of your cloud platform.
Mastery of your platform’s services is crucial in uncovering opportunities to save money.
Conclusion
Platform engineering expertise with a strong understanding of your chosen cloud platform and its service offerings will play a key role in your cost reduction efforts.