The $100K Mistake You Could Be Making in the Cloud
Written by: Maryann Werner
Between January 1, 2018, and December 31, 2019, 196 separate data breaches across the globe were caused primarily by cloud misconfigurations. (TechRepublic, 2020)
Cloud misconfigurations can cause harmful effects on the organization and are often due to avoidable mistakes. As cloud, data, and analytics professionals, we’ve seen the good, the bad, and the ugly when it comes to cloud configurations, primarily when data and analytics teams perform these tasks.
Below are the top 5 mistakes we have found that data and analytics teams make in the cloud.
Lack of Planning
Data-based decisions are going to be as good or bad as the data itself. It’s essential to leverage data governance processes to ensure the hygiene of the data you’re loading can lead to quality decision-making. Also, it can be damaging to undo a data workload migration as there are dozens of variables that come into play. Starting with a cloud strategy, having data governance in place, and understanding the criticality of workloads in the Cloud is critical.
Costs are Misaligned
Cost estimations are a critical part of the planning portion of data migration.
For example, we had a company come to CCG after overspending a steep $100,000 in cloud costs. The customer was performing a sizeable global deployment, anticipated the monthly spend to be around $50,000. Month one was $30,000. Month two was $75,000. Then month three was $130,000.
After coming to CCG to correct the cloud instance, we found that the incorrect cloud purchasing model, in addition to other simple mistakes were the cause of the $100,000 error.
Skillsets are Not Addressed
Teams will deploy based on what they know, not the optimal way to perform the task. Some necessary skill sets may include:
Data migration – Sourcing and transforming data from multiple platforms.
DevOps – Acknowledging the entire software lifecycle in end-to-end implementation.
Database querying languages – Understanding of languages like SQL.
Advanced analytics – systems for data science and ai model development.
Cloud security – Including cloud application and architecture security.
Generally, knowing what the data is being used for, the process to get to that target end-state, and correction control in case of an error are three skills needed for cloud migration.
Security is Secondary
Cloud security keeps the architecture in check for vulnerabilities and risks. A single security breach can be costly, and the team needs to understand all the potential ways private data, customer data, and intellectual properties can potentially be accessed.
Often, data and analytics teams think the cloud environment will just “work.” It’s necessary to understand the ongoing operations like spend, system updates, data access processes, and data ingestion, for example.
These five mistakes from a data and analytics team deploying a cloud can be costly. Learn more about the fundamentals of a cloud workload with this blog and connect with a CCGer on more tips to deploy a cloud workload at firstname.lastname@example.org.
Written by CCG, an organization in Tampa, Florida, that helps companies become more insights-driven, solve complex challenges and accelerate growth through industry-specific data and analytics solutions.