Determining the Best BI Deployment Strategy

With the rapid advance of Software as a Service offerings, data warehouse or business intelligence appliances, and BI capabilities embedded within enterprise applications (enterprise resource planning, customer relationship management, business process management, etc.), companies are now faced with a variety of new deployment options for BI. Traditional on-premise software installations still prevail as the dominant method for BI deployment, but alternative approaches are rapidly taking root and gaining attention within the maturing BI marketplace.

IT and business management executives are asking critical questions to determine the best method for BI deployment: Which approach is the best for my organization? Does any one deployment method stand out as a clear leader for solving the business pressures driving BI deployment costs today? What are the advantages and disadvantages of each approach?

As data volume and complexity grow, and as the number and variety of users increase, the challenges to effectively deploy business intelligence capabilities to the organization become increasingly acute. Companies are beginning to investigate emerging deployment strategies to increase BI implementation effectiveness and improve access to BI capabilities across all levels of the organization. In addition to traditional on-premise client/server implementations, companies are investigating new deployment options, such as:

• On-premise Web-based deployment;
• Multi-tenant hosted BI — referred to as SaaS or on-demand BI;
• Dedicated hosted BI or application service provider BI;
• BI appliance — packaged hardware and software, or virtual machine software; and
• BI embedded within enterprise applications.

New Research

A new research study that will be published at the end of April will uncover the strategies, actions, technology investments and services that Best-in-Class companies are utilizing to effectively deploy BI capabilities to their organizations.

This study will present detailed data, conclusions and recommendations based on an extensive survey of end users to determine how organizations are currently deploying as well as planning to deploy BI capabilities. Respondents were divided into Best-in Class (top 20 percent), Laggards (bottom 30 percent) and Industry Average (the 50 percent in between) based on the key performance indicators presented immediately below.

Key Performance Criteria

Here are some of the highlights of this new study:

• Self-service capability for nontechnical business users of BI applications. Thirty-seven percent of Best-in-Class companies have achieved self-service BI capabilities for nontechnical end users, compared with 23 percent of Industry Average companies and only 10 percent of Laggard companies.
• Average time-to-completion of BI projects. Best-in-Class companies achieve completion of BI projects in an average of 106 days, versus 120 days for Industry Average companies and 206 days for Laggards.
• End-user satisfaction with current BI capabilities. All (100 percent) of Best-in-Class companies report that end users are at least “mostly satisfied” with current BI capabilities, compared with only 12 percent of users at Industry Average companies, and only 6 percent of users at Laggard organizations.

Best-in-Class Performance Characteristics

Here are some more findings from the new study:

• Eighty percent of Best-in-Class companies have experienced an increase in enterprise-wide use of current BI deployments, versus 50 percent of Industry Average and 31 percent of Laggard companies.
• Best-in-Class companies have experienced an increase of 172 percent in the overall number of users accessing BI applications, compared with 116 percent increase among Industry Average companies, and a 15 percent decrease among Laggards.
• Best-in-Class companies have seen their cost-per-user of BI applications increase by 26 percent during the past 12 months, compared to an increase of 38 percent among Industry Average companies, and an increase of 74 percent among Laggards.

To achieve Best-in-Class performance, companies must focus their BI deployment procedures at an enterprise level. Best-in-Class companies are far more likely to have adopted BI across the entire organization than other respondents.

Best-in-Class Strategies

Here are some Best-in-Class strategies for providing BI capabilities to more end users:

• Automation. The ability to remove manual tasks from IT administrators and automate the collection, assembly and delivery of BI capabilities to end-users is the top strategy that Best-in-Class companies are prioritizing. New deployment approaches offer varying degrees of automation capability that early adopters are experiencing today.
• User support and training. The creation of a BI center of excellence signifies a formal approach to developing enterprise standards for BI deployment. This includes a regular process for identifying and defining end-user requirements for information, support and training.
• Culture and adoption. While many BI deployments start at the department or project level, successful companies realize that a culture must be established around BI capabilities if they are to be successfully adopted enterprise-wide.

Best-in-Class and all other respondents view proof-of-concepts from vendors to be equally important. The ability to see how a particular deployment approach performs under real-world conditions — using existing data sets and involving actual end users — can be critical to determining if a specific deployment approach is more or less effective in providing end users with access to BI capabilities.

Meanwhile, Best-in-Class companies are much less likely than other respondents to increase their investment in IT infrastructure to improve BI deployment performance. “Throwing hardware at it” is not a solution that Best-in-Class companies see as being a valid approach and may be a pitfall to avoid based on Industry Average and Laggard performance results.

There is a lot to be gained from adopting a best-practices approach to information delivery for the purposes of decreasing costs and project time lines, increasing on-budget project completion, and especially addressing previously underserved employees.

The selection of a deployment strategy for BI solutions plays a crucial role in the ability to manage the total cost of ownership of BI, and realize return on investment from BI initiatives.

Key Benefits

Selecting the right BI deployment method from the end user organization’s perspective creates the opportunity to do the following:

• Achieve improvement in user information access and support.
• Lower the cost of implementations and on-going maintenance and support.
• Improve BI application deployment and time-to-completion of BI implementation projects.
• Extend the reach of BI applications to all users when, how, and where they need it.

The ability to deploy BI applications that deliver the right information to the right people at the right time is still an illusory goal. Aberdeen research shows that a majority of companies are still trying to achieve this for at least business and IT management, and are still struggling to provide access to BI applications for line-level knowledge workers.

New deployment methods such as SaaS, data warehouse and BI appliances, as well as embedded BI within existing enterprise applications promise varying degrees of deployment enablement, but the adoption of these approaches has not yet reached a critical mass.

Currently, Best-in-Class companies are achieving deployment success by focusing on a combination of capabilities, including data management, automation of report generation, monitoring system usage and implementing methods for measuring project metrics in order to determine what it takes to improve over time. Overall, the building of a corporate culture around BI — particularly one that involves the inclusion of both business and IT management and end-users — is one of the most important ways to build relevant skill sets and achieve deployment success.

Information on this report and any other Aberdeen publications can be found on Aberdeen’s Information Technology site.