Cloud Business Intelligence's Perfectly Stormy Future
The cloud will reduce the cost of delivering simple, powerful business intelligence so organizations can finally enjoy pervasive BI.
by Brian Gentile
Much has been said and written about cloud computing. The cost and scale advantages it can offer public and private computing infrastructures are astounding - providing genuine hope for more efficient and far-reaching, next-generation computing. In fact, cloud BI holds great potential to substantially disrupt the business intelligence market because of its perfect storm of low-cost, scalability, and flexibility.
Consider that, perhaps surprisingly, this year's CODiE Award for Best Cloud Infrastructure Platform was given to a business intelligence platform: Jaspersoft's business intelligence solution "stack," which also includes RightScale's cloud management tools, the Vertica analytic database, and Talend's data integration software. This award, and the interest in this integrated, cloud-based BI solution, demonstrates an age-old computing principle that can be easily forgotten: solving real business problems with a well-integrated solution is necessary, not optional.
The future will be very bright for the use of BI in the cloud, both because of the advantages that underpin this new computing paradigm as well as the explosion of digital data that grows each day. In a recent IDC survey [i], 50 percent of respondents said it was highly likely they would pursue the public cloud for BI/analytics and nearly 70 percent said it was likely they would pursue a private cloud deployment. To best understand what's next for business intelligence in the cloud, first an understanding of where it is (and what it is) is vital.
Business Intelligence in the Cloud Today
Cloud BI represents a way for reporting and analysis solutions to be developed, installed, and consumed more easily due to its lower cost and easier deployment. Ideally, a cloud-based business intelligence platform makes use of infrastructure-as-a-service (IaaS), complements and extends today's platform-as-a-service (PAAS), utilizes an on-demand, virtualized, elastic software and hardware environment, and delivers application-level functionality as a service (commonly referred to as software-as-a-service). Additionally, a business intelligence tool should easily deploy and even migrate from on-premise to the cloud (and back), providing a new kind of Web-based flexibility that accompanies the most modern platform architecture.
Typically, a cloud-based BI platform is used to solve one of three primary customer needs:
1. As a horizontal BI tool to deliver standalone, internally facing reporting and analysis applications -- probably using a traditional relational database (or data mart) as the primary source data system.
2. As an application framework or pre-built reporting and analysis template for systems integrators to use for assembling customer-specific solutions more quickly. These solutions are probably function- or domain-specific and contain reusable components and application logic (but are assembled uniquely for each customer).
3. As a development platform that enables embeddable, externally-facing applications that solve a function-specific data analysis problem (for example, CRM analytics, financial analytics, or supply chain analytics). In this case, an ISV (or an enterprise IT team with appropriate skills) would probably use the BI platform to deliver reporting and analytics as a well-defined and well-featured layer within its larger application. The result is an analytic application that solves a customer problem with minimal customization and that is ideally delivered using a software-as-a-service architecture on top of a cloud infrastructure.
Further, the early adopters of cloud-based BI tools can be described and categorized by both size (or type) and usage. The three most common are:
1. Small to Midsize Enterprises (SMEs): Most often SMEs would use a cloud BI platform as a horizontal tool (see customer need #1 above) because it offers an instant-on infrastructure and pay-as-you go model. In the right hands, this can enable a powerful, affordable internally facing BI solution. Some SMEs may also be developers (ISVs or VARs) and, therefore, prefer cloud BI as a development platform (see customer need #3 above) for its flexibility and affordability. For outfitting a Web-based application with sophisticated reporting and analytics, a cloud-based BI platform just may offer the fastest time to value.
2. Large Enterprises: Many large enterprises are interested in cloud BI as a horizontal tool to provide a simple, distinct, affordable "IT sandbox" where project experimentation and evaluation can occur far from a production environment. The organization may also want to use cloud BI to develop and deliver a departmental BI project more quickly and inexpensively than other options. In both cases, avoiding the time and expense of buying and configuring server hardware, operating software, and database software holds a strong appeal and greatly accelerates the evaluation-to-deployment cycle.
3. Systems Integrators (SIs): Systems integrators want to use the cloud BI platform as an application framework (see customer need #2 above) to more rapidly build highly customized BI solutions for clients of all sizes. The SI's cloud attraction is based on architectural advantages (discussed below) as well as the ability to reuse application components (such as frameworks) to build and deliver custom solutions quickly and affordably. More specifically, cloud BI enables a VAR to create highly customized solutions tailored to the specific needs of each customer. This approach also enables very flexible data-source integration as well as the ability to grow into a data warehouse should it be required. The fact that the cloud is also more efficient, scalable, and less costly is an added benefit.
Figure 1 (below) summarizes the most common early-adopter customer categories, the needs being met by these early cloud BI adopters, and the most likely project type or usage of the cloud BI tool.
|Horizontal BI Tool
||Internal BI Solution
||IT "Sandbox" or
Analytics (For ISV
Figure 1: Cloud BI adoption by customer category, customer need, and project type
In total, these early adopters are choosing a cloud-based BI tool because of a well-understood, well-documented set of advantages that spans the cost/operating model, the architectural model, and the business model.
Cost/Operating Model Advantages
- Low-cost software and hardware infrastructure with instant-on capabilities and no capital outlay.
- Predictable, on-going operating expenses on a known pricing algorithm. Although the BI tool usage pattern may be less predictable, the operating expenses should be easily calculable. In this sense, capital expense is eliminated in favor of flexible operating expense.
- Scalable (can effectively handle unpredictable workloads with elasticity and reliability).
- Simplified operations via cloud management platform minimizes the need for on-board expertise and IT skill.
Architectural Model Advantages
- Multi-tenant capability available at least as an option; typically desirable both for ISVs and larger enterprises.
- Re-skinnable and customizable look-and-feel, complete with well-defined APIs.
- Automatic elasticity, because deploying to the cloud is necessary but not sufficient for many uses.
- Data integration from any source, on-premise, SaaS, or cloud; data is everywhere and the cloud should be the unifier whenever possible.
- On-premise migration to allow the entire cloud-based BI "stack" to easily be moved on-premise (should the customer wish to do so).
- Consistent and reliable deployment via public or private cloud, using a cloud management platform such as RightScale
- Private cloud infrastructures provide the path to broad cloud BI adoption by the enterprise thanks to their inherent ability to reduce security and latency obstacles.
Business Model Advantages
- Integrated, ready-to-use "stack" can be purchased via "one-stop-shop" resellers, so the front-end of the cloud experience is as simple as the rest.
- Integrated technical support that has knowledge of the entire BI stack and clear escalation paths to the underlying software vendors, making the back-end of the cloud experience as simple as the rest.
Business Intelligence in the Cloud -- A Look Ahead
As more BI projects utilize cloud services, the advancement curve will steepen. In turn, customer requirements will drive the next generation of BI in the Cloud toward higher service levels. Here's what to expect across the cost and operating model, the architecture and technology, and the business model.
- "Multi-Cloud Clustering" will provide seamless load balancing, scaling, and even fail-over between defined cloud nodes, either public or private.
- An integrated, ready-to-use "BI stack" will be purchased via auto-provisioning and auto-billing and through a single location. SaaS BI provides this today because it is a more constrained solution with fewer configurable options.
"Big Data" will become more easily bridled as cloud BI scales more effectively both vertically and horizontally. Vertical scaling is traditional, with database, application servers, etc. sharing the load. Horizontal scaling is across more data source end nodes, which is required as "big data" is increasingly distributed and in a variety of formats. These varied data formats will more commonly (and necessarily) include unstructured and semi-structured formats (especially text), requiring the BI tool to be adept at helping to find meaning across it all.
- Customers will be able to choose from among a marketplace of pre-built, standalone but complementary components that can accelerate a BI solution for a variety of business functions or domains.
- Rather than monthly pricing, economies of scale will enable ever-increasing options at the BI level just as we today enjoy at the IaaS cloud level. For example, hourly pricing or reserved pricing will likely be available options.
The Last Word
The current interest and advancement of cloud BI helps confirm that reporting and analytics have really become part of the fundamental fabric for cloud platform infrastructure in much the same way as they already have for application development in other deployment models. Therefore, the definition of "infrastructure," as in IaaS, continues to advance -- making a larger percentage of the computing experience more affordable and powerful for a broader audience.
The inevitability of this advancement is mathematical, according to Cloudonomics.com (see Note 2). "A pay-per-use solution obviously makes sense if the unit cost of cloud services is lower than dedicated, owned capacity," says Joe Weinman, the blog's author. You can expect the cloud to continually reduce the unit cost of delivering simple, powerful business intelligence, so that the long-held goal of pervasive BI is finally realized.
Note 1: IDC survey and presentation, The Maturing Cloud: What the Grateful Dead Can Teach Us About Cloud Economics by Frank Gens (senior vice president and chief analyst) and Amy Konary (research director of software licensing and provisioning); April 6, 2010
Note 2: Mathematical Proof of the Inevitability of Cloud Computing, a blog on how cloud computing creates value; Joe Weinman
Brian Gentile is the chief executive officer at Jaspersoft.