This research details business intelligence (BI) professionals' perceptions of the relative strengths of their installed BI platform capabilities, based on the results of a survey conducted as part of Gartner's research for "Magic Quadrant for Business Intelligence Platforms." It will help BI leaders identify a complete set of capabilities to meet current and future business needs.

• The survey (see Note 1) found that BI platforms serve information delivery and analysis needs well, with reporting, ad hoc analysis and dashboards providing the most used and most satisfactory capabilities.

• Interactive visualization and dashboard capabilities showed the greatest increase in use over last year's survey, increasingly replacing reporting and ad hoc analysis, which showed the greatest decline.

• The integration capabilities of BI platforms rated lower for usage and satisfaction, including BI infrastructure, metadata management and collaboration. Search-based BI and predictive analytics also achieved weak usage and satisfaction ratings, although extensive use of metadata management and predictive analytics grew over last year.

• Customers of megavendors continued to supplement their deployments with products from data discovery and independent vendors to fill ease of use and functional gaps. (see Appendix 1). This trend was evident in last year's survey and has accelerated.

• The survey shows that some large BI platform vendors (including megavendors) have narrower product use than some smaller independent vendors.

• All the data discovery tools score among the highest for both ease of use and complexity of analysis performed by users. This paradox is core to their value proposition and the fuel behind their tremendous market momentum. This sweet spot also holds true for SaaS vendor Birst and for small independent vendor, Targit. Oracle is the only megavendor with this distinction.

• Look for ways to provide a balanced portfolio of BI capabilities that moves beyond reporting to put analysis into the hands of more users and to optimize business processes.

• Incorporate collaboration capabilities if decisions involve many users who interact as peers and need ways to discuss information about the business in the context of the decisions that they must make.

• Build predictive analytics and data mining into BI applications as you adjust your BI strategy for a return to growth and must seek information patterns, which can help you anticipate customer needs and market developments.

BI leaders implementing BI platforms face evolving business user requirements that indicate a need for a shift from basic reporting to more advanced types of analysis without sacrificing ease of use. However, BI leaders are not always aware of the full breadth of capabilities that may help their enterprise and how these compare across vendors. Those who understand the full range of capabilities available can give users and managers a better idea of what's going on with the business, provide more consistent information to those who serve customers and solve other problems, such as cost and revenue optimization to give the enterprise a competitive advantage.

Gartner has defined a BI platform as software that delivers 13 capabilities in three categories: information delivery, integration and analysis (see Table 1). To help BI leaders identify and compare all the capabilities that their enterprise needs, we conducted a survey of 1,225 peers in late 2010 (see Note 1). We asked respondents how much they use each of these capabilities and we asked how well they matched their enterprises' needs. Information delivery remains at the core of most BI projects, but enterprises increasingly need to focus on analysis to discover new insights and on integration to cost-effectively implement those insights.

Respondents say that their installed BI platforms serve their information delivery and analysis needs well (see Figure 1). The bars in Figure 1 show the average score for each capability while the line shows the percentage of respondents that use the capability extensively across all vendors. The percentage on top of each bar indicates the percentage change in extensive use between this year and last year's survey (conducted late 2010 versus late 2009). None of the capabilities does extremely well — reporting comes closest with respondents rating their satisfaction about 8.5 out of 10 — and none does extremely badly, the worst ratings of about 7.4 going to predictive modeling and metadata management. The areas of relative weakness lie mainly in the integration category; infrastructure, metadata management, development tools and collaboration serve users less well. Search-based BI and scorecards also receive relatively weak ratings. As in previous years, predictive analytics receives lower ratings and less use than other BI platform capabilities, although use is growing.

With the exception of development tools and BI infrastructure, in general, the satisfaction with capabilities rises along with the prevalence of use. Only reporting is used extensively by most respondents (68%) while six capabilities are extensively used by 25% or less. Use of reporting and ad hoc analysis experienced the most significant decline in extensive use while interactive visualization and dashboard use grew the most. These results suggest that many enterprises are moving beyond a focus on measurement of the past. In particular, the recent popularity of interactive visualization tools and vendors' increased focus on delivering easier-to-use advanced analysis, such as predictive analytics suggest that these capabilities will see increased use in 2011.

Figure 2 evaluates major vendors' products based on how respondents rated their satisfaction with the capabilities and those products' ease of use. (The dotted lines show the survey average for both metrics.) The data shows that products with the highest ease of use scores tend to also have the highest composite product functionality rates. This result is consistent with another Magic Quadrant survey finding, which shows that for the first time since we have been conducting the Magic Quadrant customer survey, this year, "ease of use" is now the No. 1 reason why organizations select a BI platform — surpassing "functionality," which has traditionally been the No. 1 reason. Strong functionality is clearly no longer enough. All the data discovery tool vendors including QlikTech, Tibco Spotfire, and Advizor, upwardly mobile niche vendors, such as LogiXML, Bitam, and Board scored above the survey average for both measures. This combination has been a key driver of their market momentum. Enterprises that use a megavendor's BI platform continued the trend to supplement it with products from data discovery and independent vendors to meet their needs for ease of use and functionality (see Appendix 1).

Figure 3 rates satisfaction with the functionality of vendors' products against the products' breadth of use, that is, how many of the 13 capabilities the average customer uses. (The dotted lines show the survey average for both metrics.) The survey data found no correlation between breadth of product use and size of vendor. Some large vendors (including megavendors) have narrower product use than many smaller vendors. For example, customers of IBM Cognos, Information Builders, Microsoft, and SAP use fewer of these vendors' capabilities than average while customers of MicroStrategy, Tableau, Board, Targit, Tibco Spotfire, Bitam, and Salient use more of their capabilities than average and also have above average composite product scores. This result combined with Figure 2 showing the correlation between ease of use and product score suggests that trends of enterprises standardizing on BI vendors and sourcing from the software stack vendors often do not trump functional needs when business users decide which products to buy.

Figure 4 rates each platform's ease of use against its support for complex of the types of analysis (e.g., complex ad hoc analysis, interactive visualization, and predictive analytics) performed by users with the platform. An orange color of the dot indicates an above average breadth of use score. All the data discovery tools score among the highest for both ease of use and complexity of analysis supported. This paradox is core to their value proposition and the fuel behind their tremendous market momentum. This sweet spot also holds true for SaaS vendor Birst and for small independent vendor, Targit. Oracle is the only megavendor with this distinction.

The survey also rated customers' satisfaction with their BI platform's performance as defined by speed of query response time. Figure 5 measures each vendor's average performance score and average deployment size (this is a combined measure of average number of users and average data size). The orange color dot identifies those vendors that also have above average scores for supporting complex types of analysis (complexity of use). Large independent vendors, MicroStrategy and Information Builders score most favorably on performance for very large deployments with MicroStrategy also supporting more complex types of analysis. The data discovery tools also score well on performance for complex use cases, but on smaller deployments than the survey average. The megavendors support the largest deployments, but have lower performance scores. With the exception of Oracle, they also support less complex types of analytic workloads. These results further highlight the bifurcation in the BI market between traditional BI platforms (upper left hand corner) and emerging alternatives (lower right hand corner). Note there is also a high correlation between the two camps with ease of use scores. In general, the vendors in the upper left hand quadrant tend to have lower ease of use scores, while the vendors in the lower right hand quadrant tend score above the survey average for ease of use.

We asked respondents to rate their satisfaction with and use of their vendor's reporting capabilities, which provides powerful formatted and interactive reporting and highly scalable distribution and scheduling (see Figure 6). Reporting remains the core of BI platform usage, with 83% of survey respondents using the reporting capabilities and 68% making extensive use of it. Respondents expressed the highest satisfaction with this one out of the 13 BI capabilities; it achieved a mean rating of 8.37 out of a possible 10.

Web-based dashboards deliver intuitive displays of information, including dials, gauges and traffic lights and can support near-real-time alerts from operational data sources. The hype around dashboards is turning into reality with 5.6 percentage points growth for extensive use. While 65% of survey respondents used this capability last year, with only 35% using it extensively, this year, 74% use the capability with almost 41% using it extensively. The usage varies widely between vendors. Figure 7 shows some vendors' customers (80%) use dashboards, while other vendors' dashboards are used by only just over 12% of customers. Overall, users do seem satisfied with their dashboards, which won a rating of 8.32. This will remain a core and growing BI function, particularly as this capability overlaps with interactive visualization functionality.

Ad hoc analysis leverages a robust semantic layer to enable users to ask questions of the data and get answers quickly, without relying on the IT organization. After reporting, ad hoc query was the most used information delivery capability, with 79% of respondents using it and awarding a satisfaction rating of 8.28 (see Figure 8). However, in addition to reporting, ad hoc query is the only other BI capability to decline in extensive use. These traditional report-centric approaches are being increasing supplanted by use of interactive visualization capabilities that appear to have grown at their expense.

BI platforms should have excellent integration with Microsoft Office tools, including support for document formats, formulas, data refresh and report creation. This area presents a challenge for most enterprises because so many users gather and analyze data in the ubiquitous Microsoft Office applications, particularly Excel outside of an organization's standard BI tool. This capability has not been widely adopted to the point where it can bring these unsupervised efforts under control or at least pull them into a managed BI environment, but use is increasing. This year, 60% of respondents use this capability with 23% of respondents saying they use the functionality extensively compared to only 19% last year (see Figure 9). Respondents listed their satisfaction at an average of 7.99, just above the average rating for all 13 capabilities of 7.98.

Search-based BI applies a search index to structured and unstructured data sources and maps them into dimensions and measures that users can easily explore using a Google-like search interface. Users want fast, easy access to information and they have got used to search for navigating the Web. Therefore, vendors have rolled out this function but with limited adoption to date. Most vendors in fact provide little more than key word search of reports and metadata rather than delivering the full breadth of functionality and potential of search-based BI. This could in part explain the lower satisfaction ratings. Only 29% of respondents claim to use the capability (two percentage points above last year) — but still the lowest of all 13 capabilities in the survey — with only 9.5% using it extensively, a slight increase of 0.2% above last year and lowest in the survey behind predictive analytics (see Figure 10). This result agrees with anecdotal evidence from client inquiries. Moreover, the satisfaction score of 7.78 for search-based BI rated it higher only than predictive analytics and collaboration. While the value proposition of search-based data discovery is in theory compelling, vendors must show more production deployments of this functionality before it achieves widespread adoption. Information Builders, SAP BusinessObjects Explorer and new entrant, Endeca show promise.

Ideally, enterprises want all their BI tools to use the same infrastructure (security, metadata, and administration, for example) and have the same look and feel. The fourth most used capability overall, BI platform infrastructure achieved a usage of 71% with 35.5% of respondents using the capability extensively, but with a below-average satisfaction rating of 7.86 (see Figure 11). Extensive use of the BI platform infrastructure varies widely by vendor, but in general customers of the megavendors and the large independent vendors (MicroStrategy, Information Builders, and SAS) reporting the highest percentage of extensive use.

One of the benefits of traditional BI tools is a common library of metadata objects, such as dimensions, measures and report layouts. Alongside infrastructure and development tools, metadata management is a core capability of BI platform integration. Use of this capability is up from 40% of survey respondents using this capability from their BI platform vendor last year to over 50% this year. Extensive use is also up from only 14.8% last year to over 20% this year (see Figure 12), although this level of usage still falls dramatically below the 71% of respondents in our 2008 survey. The relative lack of usage may be because an increasing number of deployments are departmental or for data discovery tools, where those BI platforms do not require an enterprise semantic layer — users can access and combine data sources directly. Respondents also gave the function a below-average 7.70 for satisfaction, a result which also helps to explain the low usage.

BI platforms should provide a visual interface for custom development with virtually no need to write code. Survey respondents (73%) said their enterprise uses its BI platform's development environment with 39% using it extensively, fourth, after reporting, dashboards and ad hoc analysis (see Figure 13). This function achieved a satisfaction rating of 7.92, just about average for all 13 capabilities.

BI platforms should enable users to share and discuss information and decisions and manage hierarchies and metrics via discussion threads, chat, annotations (either embedded in the application or through integration with other collaboration applications), analytical master data management and social software. Respondents (37%) use collaboration, with 13% using it extensively, the least of all 13 capabilities. They gave it the third lowest rating for satisfaction, 7.77 (see Figure 14). It should come as no surprise that Microsoft customers report the most extensive use since SharePoint is an integral part of delivering Microsoft's BI platform capabilities. However, the use of this capability is likely to continue to grow as collaborative decision making becomes more important. Collaborative decisions often involve a large number of users who interact as peers, rather than through an organizational hierarchy and they will need access to information about the business, along with modes of communication to discuss this information in the context of the decisions that they must make.

OLAP enables users to analyze data with fast query and calculation performance. This staple of BI platforms received an above average satisfaction score of 8.05, and 59% of survey respondents use OLAP (31%) extensively (see Figure 15). Thus, OLAP ranks fifth in satisfaction again this year as it did last, compared with the 2008 survey, where it received the highest rating. This drop may reflect a shift toward data discovery tools that enable advanced analysis through interactive visualization and alternative in-memory architectures.

BI platforms may provide sophisticated visualization of multidimensional data using the color, size and shape of objects. Users can easily explore data by interacting with the visuals (for example, lassoing dots on a scatter plot) without predefining drill paths in advance. Users increasingly accept this technology for analyzing complex data. Survey participants (66%) now use the capability, up from 52% this year, which ranks fourth after dashboards, and 31% now use it extensively, up from 25% last year. This represents the largest increase in extensive use of any of the 13 capabilities (see Figure 16). Respondents also gave interactive visualization a high satisfaction rating of 8.11.

This function enables users to classify variables and to estimate their values by using advanced mathematical techniques. Predictive modeling and data mining is the least used and lowest rated of any capability surveyed (see Figure 17). Just 30% of respondents use it at all (up from 27% in 2009), with only 7.6% using it extensively. It received a satisfaction rating of 7.4. Because the predictive, optimization and simulation applications built with this capability tend to deliver high business impact, it is likely to play a larger role, particularly as enterprises acquire or hire the necessary skills and as vendors deliver tools and packaged analytic applications that are usable by a broader range of users.

Scorecarding improves the user's ability to align key performance indicators with strategic objectives by using strategy maps and performance management methods. Only 50% of respondents used their BI platform's scorecarding capability (see Figure 18). This capability is used extensively by 18.4% of respondents, less than half of the respondents using dashboarding extensively (40.7%). Respondents rated their satisfaction with this capability at 7.96, just above the survey average for all 13 capabilities. These results suggest that some enterprises are missing an opportunity to use this capability to better align BI efforts with business objectives.

We also refer to traditional versus data discovery platforms in this report. The table below shows a high-level distinction between the two types.