Wherever you look, the importance of data is growing within businesses. Reports by the Economist Intelligence and CompTIA released this year have found that companies are looking to their data to provide more sources of revenue, while Glassdoor has found that the role of Data Scientist is the most in-demand position for 2016, attracting the most attention from both recruiters and businesses.

virtualization_BI3Yet, even as companies are seeking to make their products or services more profitable through smarter use of data, the mechanics behind this process are more difficult than many expect. For enterprises, the job of joining up data from different sources, and ensuring that the results are consistent, is a very challenging one. However, it’s essential to get this right so that business users can get the insights they need.

Part of the problem is that existing BI projects were built with very definite goals in mind. Reports for financial close processes or company revenue statements may have been set up years before, and these reporting solutions will often go untouched. However, for those who want to use different sets of data from those central applications, this can be very difficult to get. Either the company’s business users will have to send requests to data analysts and wait on receiving the report back, or they will muddle along themselves by using existing spreadsheets for looking at the data.

To solve this problem, more teams are looking at cloud-based solutions. Rather than relying on the business analyst or IT team to source this reporting for them, users can serve their own data requests and conduct the analytics queries that suit their specific needs. However, getting to this position requires a new approach.

Moving BI to the cloud

Traditionally, companies have used an approach that depends on centralized IT infrastructures to gather, manage and replicate data. This approach can be more time consuming when it comes to enabling reporting and use of analytics by different areas of the company making use of analytics within their activities. Because all analytics requests have to go through an analyst team, this puts a barrier in place to self-service.

Moving BI to the cloud can enable a different approach by virtualizing the infrastructure and data, and letting these data sources be networked together. This virtualization of BI strategy enables centralized and decentralized BI applications to be transparently connected via a shared analytical data fabric. The result is local self-service and execution of analytic requests while still keeping global governance in place. This approach virtualizes the entire BI ecosystem, which can transform an organisation’s approach to analytics, from development to end-user content generated from data.

As an example, let’s consider how data is used within a business process such as sales. For the marketing team, current analytics implementations can show how investment in campaigns delivers leads to the sales team. The sales function should be able to track its sales results. Operations should be able to see which invoices have been sent out and which ones have been paid. However, each of these departments has its own, siloed view of the customer without seeing the bigger picture.

Achieving a unified view

BI shared across networks aims to deliver a consistent, unified view across the whole process. For the customer journey, each team can see how its activities lead to positive results. In our example, marketing can align with sales and assign more concrete return-on- investment figures to its campaigns, while sales and operations teams can identify where the business can find its “best” customers and concentrate on selling to other, similar organizations.

The result should be greater collaboration across the business. From a technical perspective, the applications that host this data can remain in place while the relationships between data sources can be generated automatically. In this example, the record of a customer in marketing applications such as web analytics and CRM can be networked with records in the ERP and finance systems. As all these records are brought together for each customer, groups or cohorts can be created for reporting purposes within the different departments.

The process for networking this data together involves the creation of a shared fabric of virtual instances that can then be made available for analysis. Each individual can then get access to this data for their own work and reporting. This reporting takes place separately from the original data, so multiple individuals can look at the same central data. This enables the IT team to ensure that central data is governed. At the same time, shared BI and using the Cloud for data enables organizations to increase the agility of using BI across different regions, departments and customers in a more responsive way.

This allows these decentralized groups to expand the the use of their own local data across the organization’s platforms and applications so it’s ready to use. For example, the Sales, Marketing and Finance teams may have their own virtual instances that correspond to how they are managing their functional areas. Each group has its own space and can bring its own data in for analysis. However, all the teams are virtually connected to the central data sets as well. The end result is that each team can get a unified, holistic view of customer activity, orders and the like, as well as seeing how these results compare to their own aims and objectives.

Data holds a huge amount of promise for businesses in supporting decisions and providing more insight. However, without the flexibility to use data in new ways, that promise will go unfulfilled. With a networked BI system, IT teams and business users can both benefit.

The preceding article was originally published on May 11, 2016 in The Stack.