In a recent study conducted by Accenture, it was estimated that:
- Organizations are wasting nearly 30% of their cloud spend
- They are overspending on their cloud budgets by 23%
Last year, I spoke to many organizations about streamlining their cloud strategy. While several organizations created a superior business model by going to the cloud, others were driven by overzealous optimism or an urgent push for change that eventually resulted in ever-increasing bills that ended up being questioned by the management.
So, I thought I’d share my three observations (for the greater good of all).
“What exactly do the management teams of these organizations do that result in Cloud Shock” (#cloudshock)
No understanding of the current architecture, its technological complexity, and its business implications: Before you jump into implementing that serverless architecture that Coursera taught you, you need to understand the processing needs of your architecture and ask yourself questions like – “What would my processing needs look like?”, Do I need a dedicated server?”, “Do I have a fallback model?”, “Am I going to test this in parallel with my current system to understand if my systems are working?” Else you will end up slam-dunked by Murphy’s Law. Recently, a company adopted a “no fallback required” strategy against repeated advice from the architecture team, only to have systems to go offline – meaning no coming back for 5 days!
Not understanding the costs associated with the cloud: Several organizations have come up to me claiming that the cloud would save them 70-80% of their costs, but they often were not sure how those numbers would be met. Often, organizations get swept up by the idea of cost-cutting and they refuse to consider downstream costs that could cost them more. Not all implementations are lift and shift. One needs to consider future services and maintenance costs that are a function of the business that the organization operates in. So, as a rule of thumb, if a consultant gives you a number that is upward of 40%, ask questions and understand their premise, because it is important to look at your migration holistically.
Incoherent strategy: Moving to cloud is much more than just lift and shift. It needs to align with the larger vision for your organization. So, if you are not sure of how your data pipelines, data models and AI will reside and use the cloud, be ready for the big bill surprise. AI, especially around image analytics and OCR, are two areas where I have seen customers struggle due to a lack of understanding – the use of out-of-the-box solutions coupled with a general lack of data science – that right there is one of the key drivers of skyrocketing cloud cost.