Cloud Analytics Everything You Need to Know

Components of Cloud Analytics

Gartner defined six elements of analytics as:

  • Data sources
    These are the original sources of data which could include ERPs, CRMs, social media data, or website usage data. An example of a cloud-based data source would be Twitter sentiment data.

  • Data models
    Cloud-based data models make sense of and standardize how data points are related to each other. These are typically created with structured data types.

  • Processing applications
    These applications process large volumes of big data, as it’s ingested into a data warehouse. Hadoop is a popular application for data processing.

  • Computing power
    Companies need raw computing power at scale to ingest, structure, clean, analyze, and serve business data.

  • Analytic models
    These mathematical models are closed functions used to predict outcomes and require strong computing power to create.

  • Sharing and/or storage of
    Data warehouses as a service enable organizations to quickly implement a modern analytics architecture and easily scale.

Cloud analytics encompasses any implementation of these elements in the cloud. Popular cloud computing platforms, which include most of these needed components, are Amazon Web Services (AWS) and Microsoft’s Azure.

Types of Cloud Analytics

  • Public Cloud
    Public clouds offer organizations applications-as-a-service, such as virtual machines, storage, and data processing. They are available to the public and sit on a multi-tenant architecture where IT systems are shared, but data is not. This allows for companies to reduce cost and streamline IT management.

  • Private Cloud
    Private clouds are proprietary clouds dedicated to a single organization. They serve as extensions of an organization’s existing IT infrastructure and are accessible only to the company. These are implemented when data privacy and security are a top priority. The downside to this implementation is its high cost.

  • Hybrid Cloud
    Hybrid clouds are a combination of public and private clouds. These implementations enable organizations to reap the benefits of on-demand IT infrastructure for non-sensitive data, but also maintain sensitive data in a private cloud.

Benefits of Cloud Analytics

  • Enterprise data consolidation
    Large enterprises have many disparate data sources, and it’s difficult to see how all the moving parts of an organization are working together if they’re in different places. A cloud implementation can provide a data warehouse that’s accessible to those who need the data. Companies can easily ensure data governance so only those who need the data get it. Another advantage of consolidation is the ability to use online services to perform data mining and advanced analytics to create prediction models updated in real time.

  • Ease of access
    Data in the cloud can be accessed by both employees and external stakeholders, and governance controls can be put in place to control access to the right people. Managing access from disparate data sources requires more resources to manage internally and slows down innovation and insights.

  • Sharing and collaboration
    Increased ease of access and data consolidation lead to more sharing and collaboration between employees, which is why cloud analytics is a good fit for global companies. Employees can easily transfer files and collaborate in real time when they view analytics in the cloud from anywhere in the world. This is also conducive to the growing trend of a telecommuting work culture. Data discovery becomes an everyday part of the culture when cloud analytics is implemented within a BI system.

  • Reduced operating costs
    Setting up an in-house analytics solution can be extremely costly, especially for smaller organizations who may not have the internal skillsets to do so. With cloud analytics, organizations don’t need to purchase hardware and provide continuous support, which can be very demanding and creates vulnerability if not properly executed. There are also ongoing upgrades which need to occur and can create unnecessary downtime. A cloud solution will take this burden off an organization’s hands so they can focus on their core competency.

  • Scalability
    It’s also easier to scale up capacity as the business grows, as the organization can simply increase its number of subscriptions as opposed to purchasing new hardware. It also ensures systems scale up accordingly if there is a sudden increase in demand for the analytics systems.

Challenges of Cloud Analytics

  • Performance
    While many cloud analytics solution providers are very reliable, an organization is still at their mercy when it comes to potential downtimes. Most vendors will promote their uptime rates, which should be considered when a vendor is chosen. For enterprises where 100% uptime is needed, a hybrid approach to cloud analytics may be the best approach.

  • Data security
    According to RightScale, data security is the number one challenge cited by corporate cloud users. They fear data loss and leakage, and cloud implementation can create some vulnerabilities. However, as cloud adoption increases, data security becomes less of a concern as organizations become more familiar with managing risk. Training and certification can help improve security efforts to minimize this risk.

  • Finding the right skillsets for the job
    Companies have a hard time ensuring they have the right skillsets to build and manage a cloud analytics operation. The challenge is training and hiring to keep up with changing technology. There is a known shortage of cloud architects and developers according to a RightScale survey. However, many tasks are becoming more automated, so the value proposition for cloud analytics has improved. For example, tools exist to monitor usage patterns, resources, and automate backups.

  • Managing cost
    The on-demand nature of cloud analytics allows organizations to scale as needed, but companies often underestimate how much they will use their cloud analytics capabilities. They also need to ensure safeguards are in place to shut down instances that are forgotten about. This can lead to unexpected costs. Additionally, hiring cloud experts can be costly because of the shortage of professionals with this skillset.

  • Migration
    Migrating legacy systems such as data warehouses to the cloud can be time-consuming and expensive. There’s also a risk of data loss if the migration is not done properly. It’s important that companies consider all aspects of a platform migration and back up all data in a secure fashion.

Frequently Asked Questions