Cloud analytics, in a nutshell, is using cloud-based systems and services to perform analytics on business data. Processing and data storage is primarily done in the cloud rather than on-premise.
Cloud analytics systems and services, with the correct configuration, strive to enable self-service for different business users - not just data specialists and IT.
According to Flexera’s State of the Cloud 2021 report, 97% of all respondents reported using at least one public cloud. As businesses increasingly rely on cloud infrastructure and services, it makes sense to move business data and analytics loads to the cloud.
There are two components of cloud analytics:
- Data storage with cloud-native structures like Amazon RedShift or MySQL Server hosted on Azure
- Data analytics achieved by processing data loads on elastic cloud resources
Cloud services like data storage and batch processing have been built for almost infinite scalability. Select the right package from your cloud provider for the scalability you need for the services in your analytics stack. Reassessing packages based on your workloads over time ensures you get the most for your money.
With the scalability of processing power, you can obtain analysis results as quickly as you’re willing to pay.
On-site analytics services can be challenging to access, configure, run, and obtain results from remotely. Using cloud analytics, you’re able to log in and perform these activities on the go, provided your cloud security and access permissions are correctly configured.
The move towards modular component-based systems and leveraging software APIs enables you to swap out components in the future based on business needs and new efficiencies. This saves considerable time and effort than implementing a completely new system.
Even when others are running intensive analysis on data, with the correct configuration (such as concurrent database reads), you can perform your own data analysis. No more waiting for batch processing jobs to run or resolving scheduling conflicts.
Cloud-based systems have security baked into their service offerings. In fact, according to the Oracle and KPMG Cloud Threat Report 2020, only 12% of professionals believed their on-premise environments were somewhat more secure than public cloud, and only 2% thought they were much more secure. 40% of respondents thought the cloud was much more secure, up from 27% in 2019.
Most organizations are already performing various types of analytics on their business data. When choosing a cloud analytics stack, you will want to assess your current systems, determine the goals you want to achieve, what transformation of existing systems will look like, and then look at service providers.
For a start, where is your data currently stored and how? Is it structured or unstructured? Onsite or offsite? In MongoDB, MySQL databases, or Excel? Rolling these into data warehouses in the cloud is one step towards unlocking siloed data. It also makes sense for the computation (i.e., analytics to be pushed to these database layers instead of pulling data into extracts or cubes for BI workloads.
Choosing highly interoperable cloud data storage and analytics services ensures you don’t get locked into the one vendor/one solution problem that causes issues down the road.
With analytics products, in particular, it can be easy to become locked into a proprietary ecosystem. Migration from one service to another can be difficult, if not impossible, to achieve without custom code and many months of work. That’s just one of the reasons we decided to do in-database analytics with bipp. With much of the heavy lifting done on the database side, rather than in our BI product, it means your analytics solution is highly portable - for reduced lock-in should you wish to change products or combine with other analytics solutions down the road.
To see cloud analytics in action…
IN THIS PAGE