Consolidation Reported in Business Intelligence

Driven by the current focus on vendor management costs, IT organizations are trying to reduce the number of vendor relationships. The BI tool vendors, almost without exception, are responding by broadening their product suites.
Written by Aaron Zornes, David Folge , Contributor

The BI tool marketplace has long been an oligarchy of medium-sized vendors in the $200M-$600M annual revenue range. This reflects the reality that most large enterprises must invest in products from at least four or five BI vendors to fully cover their needs. Driven by the current focus on vendor management costs, IT organizations are trying to reduce the number of vendor relationships. The BI tool vendors, almost without exception, are responding by broadening their product suites. By 2006/07, we expect the BI market to consolidate around two to four large vendors with comprehensive product suites, leaving a series of smaller specialists that have niche offerings. This transition from the current best-of-breed market to an integrated suite market will be traumatic for many vendors, and we expect much consolidation and partnering during the next five years.

Two recently announced mega-mergers, plus one related deal, mark the beginning of this market consolidation Business Objects acquiring Crystal Decisions, Hyperion acquiring Brio, and in distant third place, Actuate acquiring Nimble Technologies. All of the top 10 BI vendors agree that an industry shakeout is coming (and is indeed underway), and they all believe that they will be among the final winners the vendors with the most successful and complete suites. Obviously, not all of them will be in that final group of approximately four vendors. The others will be bought out to fill in the suites of the eventual leaders, will have to find defendable niches in which to survive, or will face extinction.

Among the contenders, Cognos and Business Objects have clear leads and are expected to be among the ultimate winners unless, of course, they are purchased by large outside vendors that want to buy their way into this market. The combined Business Objects/Crystal organization expands its product suite in the reporting arena and solidifies its position as an eventual winner. Cognos, on the other hand, remains in a strong position with its newly announced ReportNet product. The Hyperion acquisition of Brio makes Hyperion a contender for full-suite leadership (see Figures 1 and 2).

Business Objects and Crystal
Business Objects has long excelled in ad hoc query and end-user reporting. While still a leading BI company, Business Objects has in recent years lost market share to Cognos. This was in large measure due to long delays in developing the most recent release (2Q03) of its core product, Enterprise 6. Although the new release addresses key problem areas (e.g., with improved thin-client access and better scalability), it does not represent a major breakthrough in functionality. We believe this acquisition is an effort to shore up the product line in the formatted reporting arena, address product development deficiencies by acquiring products, and broaden its customer base to the large number of Crystal Reports users. This acquisition, along with the Acta purchase (giving Business Objects extract/transform/load capability), rounds out most of the requirements for a full-suite player. The primary area needing improvement is OLAP (we believe Crystal Holos will remain in maintenance-only mode). While it is not necessary for a BI suite vendor to have an OLAP engine, given that much of the engine market will go to the database vendors, it is important to have a first-rate OLAP viewer. The combined company must quickly leverage the best feature of Crystal Analysis and the Business Objects OLAP tool to achieve a first-rate OLAP viewer (a la ProClarity).

The considerable overlap between the Business Objects and Crystal product lines (most end users consider the two companies direct competitors) presents considerable integration challenges. While the integration of company operations (e.g., accounting, HR, administration) will be no more difficult than most mergers, the sales and development organizations present problems. The direct sales forces must be integrated and presented with a rational set of products that they can clearly differentiate to prospective buyers. The overlapping distributor networks must be managed to reduce overlap (two organizations addressing the same customers) and to avoid channel conflicts with the direct sales organization. In R&D, it is unlikely that a high degree of integration is possible at least in the core products. To maintain the viability of the key Business Objects products (e.g., Web Intelligence and client version) and Crystal (Crystal Reports and Enterprise), compatibility with old releases, attempts must be made to combine code bases. Some peripheral product duplications (e.g., OLAP clients, data source connectors), however, can be reduced or eliminated. In general, it is our view that Business Objects will struggle with the implications of this merger for at least 18-24 months. Even in a worst-case scenario, where the merger results in underperformance for an extended period, we believe the core products will remain viable, and we encourage end users to deploy them as needed.

Hyperion and Brio
With the Brio acquisition, Hyperion fills out some missing components (e.g., ad hoc SQL query and reporting, enterprise reporting, dashboards) of its business performance management (BPM) platform. The fit between these companies’ products is remarkably free of overlap they are highly complementary. Also, unlike Business Objects, which is adding another development location with Crystal (in Canada), Brio is located within 10 miles of Hyperion’s Sunnyvale, California, headquarters. Brio’s smaller size (about one-fifth of Hyperion in revenue) and similar business model should make the always-difficult job of integration somewhat easier than the Business Objects/Crystal merger. Key integration tasks include replacing the currently resold Crystal reporting product with Brio/SQR, integrating the dashboard/scorecard products from both companies, rationalizing overlap in OLAP views (Brio Intelligence, Hyperion Analyzer), and integrating the Brio product with Hyperion’s application business.

Unlike the Business Objects/Crystal situation, where both companies are in strong market positions premerger, Brio has suffered a sales slowdown in recent years. Although reference checks within the META Group client base indicate Brio still has a loyal following of customers and technically viable products, most organizations no longer consider Brio for new projects. While most prospective Brio buyers cite company viability as a key concern, this will no longer be a problem after the acquisition. Hyperion must address the differentiation and market-positioning problems inherited from Brio to restart sales.

The Battle Is Not Yet Over
These vendors face potential challenges from several other organizations. Microsoft appears to be moving into the BI market. If it becomes serious about BI, it can be expected to make a strong play for a dominant role using a combination of internal development and the purchase of existing vendors, much as it is doing on the low end of the ERP market. SPSS is the dominant BI vendor in the still-growing AS/400 market and a $230M company. Like SAS, it must make a major push to expand out of its niche if it is to become a dominant player. Also like SAS, it has the resources; the question is whether it has the desire. Information Builders, primarily know as an enterprise reporting vendor, is also a dark horse for a suite leadership position.

Oracle, SAP, and Siebel are all building their own BI toolsets. The main question concerning these vendors is whether they will create truly standalone products that work as well in heterogeneous environments as do the independent vendor products or whether they will be satisfied with focusing on their own installed bases, essentially becoming niche BI players. Siebel and SAP are the most likely to take the general tool route and contend with the BI vendors for a spot in the final four, but it is too early to be sure which path any of these vendors will follow or what success they may find, particularly outside their user base.

User Action
Users should standardize as much as possible on four to six best-of-breed BI vendors that, among them, cover all the organization’s major BI needs outside of data mining, which will remain a specialized, embedded function rather than becoming part of any vendor’s tool suite. Users should attempt to include at least one strong contender for “final four” status in their list of standard BI solutions. Over time, as that vendor expands its suite, the user should adjust its standards to focus more on that tool suite until it becomes the standard for 90% of their needs.

None of this will eliminate alternative tools that already are in use in the organization. Rather, a standard approach should freeze use of non-standard solutions, confining them to niches within the organization, while the standard tools are applied as appropriate to new BI projects.

We believe that current Crystal Reports and Business Objects users can continue with additional deployments as planned. Brio (when integrated into the Hyperion BPM Platform) should be considered a long-term viable player for query, analysis, and enterprise reporting evaluations. Given the complementary nature of the products, it is unlikely that major Brio products will be canceled.

Business Impact: Reducing the number of software vendors in use within an organization can save considerable training and maintenance costs.

Bottom Line: Within three years, consolidation in the business intelligence market will result in the emergence of a few large, dominant BI suite vendors capable of covering 95% of an organization’s BI needs. IT organizations should include one of the current large suite players in their approved BI vendor list and use it as a base to begin reducing tool proliferation.

META Group originally published this article on 30 July 2003.

Editorial standards