Reduce the Total Cost of BI Ownership

IT and business management are increasingly expressing alarm at rising costs associated with Business Intelligence (BI) implementations. The fear of hidden costs has kept many companies from making an investment, and many adopters have found that ongoing support and maintenance costs for ever-changing analytical and reporting requirements inhibits user penetration and the ability to manage total cost of ownership (TCO).

A new report from Aberdeen Group entitled "Managing the Total Cost of Ownership of Business Intelligence" reveals that top-performing companies are doing more with less when it comes to managing the costs associated with implementing, deploying and supporting a BI solution. The report finds that Best-in-Class companies are distinguished from all others in that they achieved significantly better year-over-year results on four key metrics:

This executive summary details key findings of Aberdeen's "Managing the Total Cost of Ownership of Business Intelligence" report (to download the full report at no charge, click here to register). Among the conclusions shared are the findings that To achieve Best-in-Class performance, companies must:

BUSINESS CONTEXT

Aberdeen's research conducted among 194 professionals at 162 companies between March and April 2009 reveals that top-performing companies are "doing more with less" when it comes to one of the major costs associated with implementing, deploying and managing a BI solution (see "Best-in-Class Cost Management" chart at right).

Interestingly, despite requiring fewer full-time equivalent employees (FTEs) to support BI deployments, Best-in-Class companies are capable of completing BI projects, from start-to-finish, both on-budget and within expected timeframes. Additionally, they are delivering BI capabilities to more enterprise users than their counterparts. This report will dive deep into the specific actions and capabilities that Best-in-Class companies are taking to achieve this result.

Research completed in February 2008 for Aberdeen's first annual Managing the TCO of Business Intelligence report revealed that the top two pressures driving the desire to gain improved control over the cost of BI were to move beyond data integration difficulties (42%) and the need to deliver BI capabilities to more end users (29%). The latest research reveals that a change has occurred at the top of the pressures list when it comes to determining what drives companies to focus on lowering their cost of ownership of BI implementations (see "Pressures Change with Better Performance" at below right).

This change between 2008 and 2009 respondent priorities can partially be explained by the advancement in deployment longevity that has occurred over time. Respondents report that they are now more likely to have greater maturity with their BI implementations. In fact, 70% of all respondents to this year's survey have had a BI implementation installed for a year or more, and 45% have been managing their implementation for four or more years.

Until recently, BI was deployed mainly to mangers, business analysts, and C-level executives as a tool used to gain better visibility into company performance, trends and standard reporting. In today's more challenging economy, BI is now being viewed as having a potential impact on operational performance.

Aberdeen Group research has found that as companies begin to mature in their use and deployment of BI technologies, they also begin to realize that there are hidden costs associated with different types of deployments. The resources required to meet the information and analytical needs of a department, group or enterprise can create an unexpected strain, both financially and in terms of time and resource allocation. In order to determine the true TCO of a BI implementation, several cost factors must be defined and understood. BEST-IN-CLASS STRATEGIES

Aberdeen used several key performance criteria to distinguish the Best-in-Class (top performers when it comes to managing the TCO of their BI implementations) from Industry Average and Laggard organizations. The following describes the individual performance metrics used to benchmark respondents for this study:

The strategic approaches that Best-in-Class companies are taking toward aligning their organizations and resources to deal with managing BI costs are numerous. Foremost among the top-five approaches is an initiative to understand end-user requirements. Best-in-Class companies are 28 percent more likely to formally poll end users regarding their specific needs for BI access and use. Gaining an understanding of end-user needs is paramount when it comes to determining what level of access and capability to deploy throughout the organization. Interviews with end-user organizations often reveal that companies purchase one type of BI tool (typically evaluated by IT and departmental power users) and deploy it to the organization only to find that a majority of people have not adopted or refuse to use the new BI tools. The interviews almost always reveal that ease of use is the top complaint expressed by end users. This alone may begin to explain why Best-in-Class companies are able to do more with less.

While all respondents generally are attempting to identify appropriate business areas for obtaining the most rapid return on investment (ROI) from their BI investment, Best-in-Class companies are 25% more likely to define the specific business rules and calculations required to generate reports and analytic views. By drilling down to precise requirements, Best-in-Class companies are paving the way for improved implementation, deployment and overall cost management.

Interviews with respondents revealed that there are no shortcuts to finding true ROI from BI investments. Research respondents from Best-in-Class companies stated emphatically that unless a thorough and formal approach to defining metrics, business rules and reporting/analytical requirements was taken, there would be very little chance of obtaining rapid or even any ROI. These same interviewees also reported that they were more apt to organize a temporary, cross-functional team at the outset of BI projects that would then be disbanded once the project deliverables had been created and deployed. This lends some insight into the reasons why Best-in-Class companies are capable of controlling implementation and deployment costs while staffing projects with fewer FTEs.

BENCHMARKING REQUIREMENTS FOR SUCCESS

From the end-users' perspective, the ability to identify, track and manage all cost factors associated with BI implementations starts with the direct, up-front costs associated with software licensing, implementation consulting, user training, and the potential requirement of additional hardware.

There are several indirect costs that come into play at the time of purchase that must also be considered, such as ease of use, end-user requirements and data-integration challenges. Direct and indirect costs are not the only factors to be considered. Many of the costs associated with BI implementations are realized over time, such as ongoing maintenance/support and upgrades, the effect of changing analytics requirements, the growth of data volumes, the potential effects of mergers and acquisitions, and the increasing footprint of BI in the enterprise as it matures from a departmental to a group or, potentially, an enterprise application.

There are many cost factors that must be considered throughout the life-cycle of a BI project, and also many more that represent "indirect" costs to consider. For example, while user license costs are typically felt at both the time of purchase and on an ongoing basis as more users are added (depending on the deployment methods and licensing options offered), additional hardware costs may not be understood at the time of purchase. In fact, 14 out of 21 companies interviewed as part of the follow-up research for this study reported that they did not recognize the potential for added hardware costs until the implementation was being deployed, and existing hardware (both servers and networking components) exhibited signs of the added stress from the BI deployment.

An even more indirect cost is associated with the ease of use (or lack thereof) of selected tools. Too many organizations leave it to highly trained technical professionals and power users to evaluate BI software products only to find that the non-technical users in the company do not have the skill sets (both analytical and technical) required to fully use the new software. CAPABILITIES AND ENABLERS

Aberdeen's analysis of Best-in-Class respondents finds that a combination of process, organizational, knowledge and performance management capabilities combined with specific technology enablement is likely to lead to improved TCO for BI implementations.

Process Management. Best-in-Class companies exhibit a significantly higher capability when it comes to data integration and cleansing. Aberdeen research conducted over the past 18 months among more than 6,000 companies consistently shows that data integration challenges are the heart of trouble when it comes to getting BI projects off the ground. Additional research conducted for Aberdeen's TCO management reports over the last two years shows that data-integration capabilities are far more prevalent among Best-in-Class companies than among Average and Laggard companies. The data further suggests that data integration has consistently been a challenge for companies that have not achieved Best-in-Class performance. The end-result is that budgets are spent and timeframes are exceeded. This can also result in disenchantment of executives who might have supported a project at the outset.

Organizational Management. End-user training is a critical component of any BI implementation. Like most business applications, users tend to use only a small percentage of features and functionality. Companies that invest in training tend to see a marked improvement in user adoption and, not surprisingly, faster and broader deployments.

Best-in-Class companies are 37% more likely to invest in end-user training for BI than their combined Average and Laggard counterparts. Moreover, Best-in-Class firms are 40% more likely to establish a formal committee dedicated to obtaining end-user and corporate requirements for BI implementations.

Knowledge Management. The path to an improved view into end-user and corporate requirements for BI reporting and analytics applications is traversed through straight-forward formalized user polling. Yet, 71% of all respondents have no mechanism in place for gaining this knowledge. Meanwhile, 43% of Best-in-Class companies poll their end-users and collect knowledge about requirements, usage patterns and information needs.

While gathering requirements may seem to be an obvious step, many companies do not bother to take it. One reason for this might be explained by a lack of resources, yet Best-in-Class companies are putting more FTEs to the task than their counterparts. In fact, Best-in-Class firms are 44% more likely to use vendor-provided end-user training services, and Aberdeen's research into the capabilities of more than 200 BI solution providers shows that almost all of them perform some form of formal end-user polling as part of their training services. CAPABILITIES AND ENABLERS (continued)

Technology Management. Based on the difficulties with data integration discussed above, it's important to analyze the current use of back-end technologies from a Best-in-Class perspective. Research for this study explored the technologies involved in building BI solutions, from initial data collection, integration and cleansing, to metadata management, data model/cube building tools, as well as the services associated with implementation consulting and IT/technical staff training. The research reveals that in each case, Best-in-Class companies are more likely to have made an investment in these back-end tools and services. In fact, in the case of metadata management tools, Best-in-Class companies are more than two-and-a-half times as likely as Average performers to use this technology. Meanwhile, before data can be managed, it must be clean. Best-in-class companies are twice as likely as their industry peers to have some form of data cleansing capability enabled (see "BIC Back-End Technology Enablement" at right).

Meanwhile, respondents also exhibit different levels of investment in front-end technologies designed to deliver information to the right people at the right time. Once again, Best-in-Class companies have exhibited specific technology adoption and usage characteristics that distinguish them from their Average and Laggard counterparts. In fact, with one exception, the Best-in-Class are more likely to have made greater investments than their peers across all of the front-end enabling technologies studied.

The one exception is the use of spreadsheets as a primary BI tool. While use of spreadsheets is still prevalent, particularly within workgroups that are underserved or inappropriately provisioned with access to BI capabilities, Best-in-Class companies are 17% less likely to allow this approach. Moreover, Best-in-Class companies are more than twice as likely to implement operational-BI reporting and analytics solutions that address day-to-day business activities and transactions rather than taking a high-level, trending view over large expanses of time (months/quarters/years) (see "BIC Front-End Technology Enablement" at right). Technology Licensing and Deployment. The approach to licensing BI software does not necessarily become a question of which path leads to Best-in-Class performance. While traditional "per-seat" user licensing is still the most dominant style of licensing in use, other approaches are rapidly beginning to encroach. Foremost among these is Server-based licensing, followed by Concurrent user licensing. Interviews with respondents reveal that concurrent licensing is the most desired, but unfortunately is not offered by all solution providers. Best-in-Class companies are more than twice as likely as all others to license BI in this fashion, which suggests they have likely made this an important item on their buying criteria wish list.

Open source licensing is also seeing increased interest. While it remains at a lower adoption level than other more traditional licensing approaches, Best-in-Class firms are also more than twice as likely to be adopting open-source BI solutions as their counterparts. The most prominent finding here is that Best-in-Class companies are far more likely to be using open source software across the entire range of BI capabilities, and are particularly more likely to be utilizing open source data cleansing tools. Meanwhile, Average performers are investing resources more heavily toward the development of open source reporting tools and are slightly more apt to be developing open source dashboard solutions.

Until recently, open source options have not been abundant across the spectrum of BI capabilities. Best-in-Class companies are far more likely to be early adopters of new open source offerings and are twice as likely to be developing full analytic applications with open source tool sets. Open source has traditionally been considered a "build-it" approach rather than a "buy-it" approach, characterized by tool kits that developers use to produce reporting and analysis applications. Increasingly, open source solution providers are offering capabilities with far more end-user interface controls and wizards, and are also beginning to offer "in the cloud" services or platforms from which BI solutions can be built.

Finally, deployment options are also factoring in heavily when it comes to managing the total cost of ownership of a BI implementation. In this instance "Deployment" refers to the methods that BI technologies are made available to customers and users. For the first time in 22 research studies conducted by Aberdeen's Business Intelligence research practice, the Best-in-Class use of traditional on-premise deployment is now lower than that of all other respondents. Best-in-Class performers are now turning to alternative deployment options, including a tendency to opt for SaaS BI deployment at twice the rate of all others, and higher adoption of both Hybrid and Appliance approaches as well.

Why is this important? As companies have matured with their BI implementations (Best-in-Class companies are far more likely to have been implementing BI for four years or longer), they have found that the skill sets required to manage, maintain and support the solutions are in short supply, and the associated direct and indirect costs are high. The growing trend toward managed services is being driven by these factors and has resulted in a slow retreat from traditional on-premise deployment.

In addition to these deployment styles, there is one other method that deserves to be called-out in relation to the findings of this study. Respondents have clearly indicated that they expect their enterprise application solution provider to deliver BI capabilities, and increasingly, Aberdeen research is finding that this is the case. The delivery of BI capabilities within enterprise applications leads to a greater pervasiveness of BI within the organization. It also supports faster deployment as, typically, the data integration issues are tied only to the source data within the enterprise application itself. Also, users accustomed to the application interface are less likely to be jarred by the introduction of a new BI application with a different look and feel.

Performance Management. Finally, all of the capabilities and technologies discussed above are not going to produce value unless the results are measurable. Top-performing companies are significantly more likely than Average companies to track the costs associated with BI implementations. In order to achieve top performance when it comes to completing BI projects on-time and on-budget, the execution of these activities must be visible to decision makers. Best-in-Class companies are more likely to measure their abilities than are Average and Laggard firms.

REQUIRED ACTIONS

Whether a company is trying to move its total cost of ownership management from Laggard to Industry Average or from Industry Average to Best-in-Class performance, the research points to the following actions to help spur improvement:

Aberdeen's full, 28-page "Managing the Total Cost of Ownership of Business Intelligence" report offers seven detailed "steps to success" for Laggard, Average and Best-in-Class firms. The report also includes more than 15 charts and tables with in-depth findings not included in this executive summary article. To download the full report at no charge, click here (registration required).