Business Intelligence Network - The Vision for BI and Beyond

What Does Software as a Service Mean for Business Intelligence?

Published: May 22, 2006

As companies move toward the use of software as a service, they must carefully evaluate the impact of this on the overall enterprise business intelligence system.

Software as a service, or SaaS, is a confusing term that means different things to different people. To the consumer, it may mean the ability to buy and download desktop software using the Internet, or to pay an annual fee to vendors like McAfee or Symantec for keeping desktop virus definitions up to date via the Internet. At the other extreme, for a business organization, it may mean purchasing Internet-based application services from a third-party vendor.

Broadly speaking, SaaS can be broken down into three main types of service:

• On-demand software purchasing where individual users or organizations try, buy and download personal, workgroup, or enterprise software across the Internet. For vendors, software-on-demand provides a cost-effective sales channel and software delivery mechanism. For users, it provides a fast and easy way to obtain software.
  
  
• On-demand IT service-oriented architecture (SOA) where IT system and application processing is defined and developed as a set of services that can interoperate and exchange information with each other. A SOA offers a flexible approach to application development that encourages service re-use and reduces the need to build point-to-point connections for data and application integration.
  
  
• On-demand application services where individual users or organizations pay external third-party providers for the use of their application services. The objective of this type of SaaS is to trim software and hardware costs, and to reduce IT staffing and skill requirements. This approach may also be used as an interim step before bringing application software in house.

Some industry analysts describe the on-demand application services model as SaaS 2.0, because it extends the capabilities of earlier SaaS initiatives. An on-demand application services vendor may, for example, support a SOA for providing easier access to the services it offers.

Two key characteristics of SaaS 2.0 are:

• Network and Web-based access to commercial software computing services where the processing is done on a third-party server, rather than at each customer’s location.
  
  
• A tenant-based pricing model for hardware, software, administration and consulting services.

Services for printing photographs, providing search capabilities or checking e-mail for spam and viruses are examples of SaaS 2.0 for the consumer. These services may be paid for directly by the consumer, or indirectly through advertising revenue (as is the case with Google, for example).

For commercial organizations, SaaS 2.0 could involve the offloading of front-office and back-office business processing to a third-party. Salesforce.com is a good example of a very successful SaaS provider in this category.

A survey (“SaaS 2.0: Software-as-a-Service as Next-Gen Business Platform,” Saugatuck Technology research report SSR-239, April 2006) from Saugatuck Technology indicates that 12 percent of U.S. companies have at least one major SaaS 2.0 application in use, with an additional 13 percent currently designing, prototyping or implementing their first SaaS application. Another 14 percent are planning to do so later in 2006 or in 2007. Saugatuck expects continued strong provider growth over the next 18 months, especially among application providers such as Employease, NetSuite, PerfectCommerce, Right Now Technologies and Salesforce.com. The Saugatuck survey also indicates that small and medium-sized businesses (SMBs) are a key driving force in SaaS adoption. SMBs are embracing SaaS at twice the rate of large enterprises. (A summary of this research can be found on http://www.sandhill.com/opinion/editorial.php?id)

The Business Case for Saas
At first sight, SaaS 2.0 appears to be another version of the application service provider (ASP) model that failed miserably after the dot-com bust. There are, however, some important differences between the two models. The first is that the ASP model focused primarily on giving an organization the ability to move certain application processing workloads to a third-party managed server. Unlike SaaS, ASPs were not necessarily concerned about providing shared services to multiple tenants. Also, most ASPs did not have a significant amount of application and business domain knowledge about the applications they were running. SaaS providers, on the other hand, usually have a large amount of domain expertise.

Another difference between the ASP and SaaS approaches is that most ASP-supported applications were monolithic client-server programs with simple HTML Web interfaces. Today’s modern SaaS solutions, however, are designed for the Web environment, which improves usability and manageability.

One reason why the SaaS 2.0 model is finding more traction than the ASAP approach is because ASP vendors rushed their offerings to market before performance, security, customization and integration issues were solved, and before many IT organizations were ready to adopt the ASP model. Today, many of these problems have been solved. Both IT and business users are better equipped to take advantage of SaaS. IT experience with Web-based and services-oriented technologies is improving, and business needs such as compliance legislation are providing the momentum for companies to adopt an SaaS approach.

How Will SaaS 2.0 Affect BI?
Business intelligence (BI) vendors are beginning to recognize the importance of SaaS, and the need for BI capabilities in the SaaS environment. Several BI vendors, for example, support the Salesforce.com AppExchange program, which allows customers to develop applications for use with the Salesforce.com CRM service. SalesForce.com also provides interfaces for moving data into and out of the CRM environment. These interfaces can be used to capture Salesforce.com data for an organization’s enterprise data warehouse. Informatica has a close working relationship with Salesforce.com for providing data integration services in this area.

Another direction of the business intelligence industry is for vendors to offer their own SaaS offerings. Business Objects recently announced Crystalreports.com, which enables customers to post reports to the site for viewing. Automated e-mail alerts are generated by the service to inform interested parties when new reports are available. The Business Objects service is aimed at companies that need to distribute reports to third-parties, but who don’t have the IT resources to develop the infrastructure for managing and securing report distribution. This approach provides a more secure and more efficient alternative to using e-mails containing Adobe PDF report files, for example.

SAS also recently joined the rush to join the SaaS movement with the release of several on-demand business intelligence services. These include access to its complete BI platform, along with supplier relationship management, marketing automation, Veridiem marketing relationship management and anti-money laundering.

Several new business intelligence vendors specializing specifically in SaaS have appeared in the marketplace. Oco Inc., for example, offers its Oco Majik service for producing product, sales and customer analytics. Another example is Host Analytics that provides a Web-based service for business performance management in budgeting, planning, forecasting and financial consolidation.

In summary, there is a clear market direction toward the use of SaaS for providing on-demand application services. This is especially true in the SMB marketplace. This direction is encouraging BI vendors to join the SaaS movement by either supplementing application service vendor offerings, or by producing their own SaaS BI solutions. Regardless, as companies move toward the use of SaaS, they must carefully evaluate the impact of this on the overall enterprise business intelligence system.