We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.

Incorporating Business Intelligence in the Cloud

Originally published August 19, 2009


In the wake of the economic slowdown, organizations are increasingly looking for ways to do more with the same resources and articulate differently – to make every penny, input and contribution count. In such situations, technologies like cloud computing and business intelligence (BI) are becoming increasingly important in gaining and maintaining a competitive edge. When combined, these technologies enable a variety of new analytic data management projects and business possibilities.

Cloud computing will change the economics of business intelligence by making available the hardware, networking, security and software needed to create data marts and data warehouses on demand with a pay-as-you-go approach to usage and licensing.

More businesses are turning to analytic applications to provide critical business insights. Whether focused on achieving higher ROI, better understanding of the competitive landscape or improving product and service quality, business intelligence is one of the few technologies that can equip organizations to more effectively prepare for tomorrow today. Consequently, the BI platform market is expected to grow by 7.9% through 2012 (according to Gartner).

Concept of Cloud Computing and Business Intelligence

Cloud computing is characterized by its ability to utilize resources and scale consumption as required. The advent of infrastructure as a service implies that computational power is available on demand and on a pay-as-you-go basis with similar characteristics applying to storage of data as well. This enables a layer of services sitting on top of this infrastructure to decouple the delivery aspect of the services from the core business oriented aspects involved in these services. Related to this is the fact that storage of the underlying data is also decoupled and segregated from the services.

Business intelligence involves intelligent reporting on top of existing data which helps in prompt and actionable decision making. These decisions might involve “geography based investment decision for a multinational company” or even a “buy decision for a product by the consumer.” Business intelligence has evolved over time, but the key components still continue to hold true. It is still necessary to be able to aggregate the factual data from various data sources and perform involved transformations. This data then either needs to be stored in a data mart or warehouse to enable reporting and analysis on top, or it could then be further aggregated into metrics that are reported later. Nevertheless, the ability to perform business intelligence involves key aspects related to data management and computationally expensive analytics or reporting.

Cloud computing provides key enablers for business intelligence in these specific areas:

  • Computational resources can be consumed for heavy calculations that could be involved, for example, in predictive analytics.

  • Extremely heavy data loads could be stored for cheaper prices in storage resources in the cloud.

  • Reporting and visualization are naturally fit for a software as a service (SaaS) model, enabling newer consumption behavior for these specific BI components. In addition, given that the web is the most common delivery method for reports, the addition of new users is seamless.

  • As more and more applications and data sets move to the cloud, BI services need to adapt to look at cloud as the data source.

  • ROI proposition for BI investments is better as a result of the pay as you go model.

Approach for Using Business Intelligence on the Cloud

Typically, business intelligence consists of three sub-systems that are commonly referred to as stage, store and report. First, you would stage data from different sources, process it using data integration tools, store it in a data store/warehouse and then conduct analysis on it. These three sub-systems will stay the same as enterprises try and move to cloud. However, the operational and cost constraints on the cloud are different from those hosted in the enterprise; hence, the BI architecture will adapt in the following ways on the cloud.

Using the Cloud for Specific BI Solutions
Given that the challenges of migrating data onto the cloud are enormous, people would prefer implementing BI solutions where the data migration issues are minimal (i.e., the amount of data that needs to be migrated/integrated is small and the data is not very sensitive). This would typically involve applications where the user base for those could be both in and outside the organizations and/or the data required for the system comes from outside the organization. These applications are very easy to set up and it is very easy to leverage the advantages of the cloud environment to the maximum, without any shortcomings. Such BI applications could be marketing automation, vendor management, campaign management and so forth.

Using Non-Standard Technology for the BI Stack
For enterprises, the BI stack consists of tools like extract, transform and load (ETL) tools, followed by a relational data warehouse and a graphical user interface (GUI) reporting tool. When BI applications move onto the cloud – with the exception of reporting – the underlying tech-stack “can” and “will” change. Most of the BI tools today borrow ideas from the relational database world and provide compatibility with SQL interface; however, going forward, the integration between different BI tools may not be a requirement as long as data integration is available. Also, the constraints on the cloud are different, along with the costs of hardware, maintenance and scaling. Run time costs are equally important, given that some of the implementations of data storage/data integration that have been used in very large systems (like search engines and web portals) are likely to be used in implementing BI systems.

Implementing Specialized Analytics on the Cloud
Of the three sub-systems, analytics is where cloud offers maximum potential. The ability to tap potentially large resources for a short period of time will allow implementation of sophisticated analytics, which can have a huge impact on the overall organization.

Cloud Strategy for BI Tool Vendors

Business intelligence on the cloud allows BI tool vendors to get over their most vexing problem, which they face whenever they make a sale in an enterprise, thereby justifying the ROI of a BI tool. A cloud BI implementation with a pay-as-you go model becomes very attractive for enterprises especially in this environment. Another persistent request of having pervasive/collaborative business intelligence also becomes straightforward with the cloud offering.

Almost all BI tool vendors either have announced a strategy or are in the process of announcing the same for cloud – the products using “on demand” to label their cloud offerings. These offerings vary in complexity and technology. Some of the smaller vendors are offering their tools on Amazon EC2, which makes it easy for vendors to use them if they already have data on the EC2 cloud. Some ETL vendors are allowing people to seamlessly upload their data on different SaaS platforms like SF.com, Google Apps, etc. Alternatively, some of the established vendors are offering to completely host/manage the BI stack for their customers, which involves data integration, warehousing and reporting outside the enterprise and on the cloud. Most of these are in the early stages and it is not clear which of these approaches will gain traction.

Issues with Cloud Computing

While business intelligence can benefit from cloud computing, it is not a silver bullet, and there are several potential challenges, such as:

  • Moving data to the cloud – Large data sets in silos sitting on premises need to get to the cloud before they can be completely used. This is an expensive proposition due to the network costs. The cheapest option is to ship disks, and this is often recommended by cloud providers like Amazon. Although this introduces latencies, it is often acceptable for most BI options.

  • Storing data in the cloud – the majority of data is core and proprietary to enterprises. Secure storage of this data is essential and, in quite a few cases, it might be mandatory to keep this data on premise. There are several options to secure data in the cloud including during storage and at rest, which include standard security measures like encryption keys, SSL and certificates. Compliance and regulatory reasons also require data to be stored securely.

  • BI components as a service – So far, only a limited set of services are available from established BI vendors. This includes some reporting capabilities and ability to do visualizations. Most established vendors are yet to introduce complete product features over the cloud.

  • Integration with on premise data – It is challenging to integrate on-premise data with cloud components, as it continues to exist in silos and requires access to data behind the firewall.

What the Future Holds

The following are likely to be key trends with respect to business intelligence and cloud computing:

  • BI services like ETL, analytics, reporting and visualization will start to be embedded within platform as a service (PaaS) offerings.

  • Increasing public data sets will be available on the cloud, which will trigger a range of associated BI services to be available.

  • As more and more data moves from on premise to on the cloud – which can be a result of SaaS applications increasing in volume – it will become more feasible to do business intelligence in the cloud.

  • An à la carte combination of various BI tools to create your own BI framework will be possible as a pay as you use model.

  • A further extension for this could be pay for performance as payment gets linked to the BI framework delivering on key operational metrics or delivering measurable value.
  • Mukund DeshpandeMukund Deshpande
    Heading the business intelligence competency center at Persistent Systems, Dr. Mukund Deshpande is responsible for building business intelligence (BI) expertise and providing consultancy to projects within the company on data warehouse design, ETL design and analytics. Mukund holds a Ph.D. in Computer Science from the University of Minnesota, Minneapolis and wrote his thesis on “Data Mining Techniques for Sequences and Graphs” that proposed new algorithms for solving problems in variety of domains, including web and e-commerce. He earned a bachelors degree in Mechanical Engineering from the Government College of Engineering Pune, India and a Master’s degree in Engineering in System Science & Automation from the Indian Institute of Science, Bangalore, India.
  • Shreekanth JoshiShreekanth Joshi
    Shreekanth has more than 12 years of experience in the software industry and is mainly responsible for establishing the SaaS practice at Persistent. As the Associate Vice President at Persistent, he is responsible for driving the software as a service (SaaS) program and executing projects in the SaaS area, providing consulting to all existing and new ISV customers of Persistent in their efforts to migrate to SaaS and cloud computing. Shreekanth earned his Bachelor’s in Engineering from Pune University and Masters from Michigan University.


Want to post a comment? Login or become a member today!

Posted February 11, 2010 by Tushar Kale

Hi Mukund and Shreekanth,

Excellent article on Cloud BI. Great insights.

Thank you,

Tushar Kale

Is this comment inappropriate? Click here to flag this comment.