C. Middleware
3. Competitive assessment
3.1. Unilateral effects
3.1.2. Middleware sub-segments
775. As regards the different middleware sub-segments, Oracle and Sun's combined market shares exceeds 15% only in the offering of (i) application server middleware475, (ii) enterprise portals476, and (iii) integration and process automation middleware477 (and more specifically ESB478 and BPMS479).
3.1.2.1. Application server software
776. Table 5 below sets out IDC's worldwide market share estimates for Oracle, Sun and their main competitors in relation to application server middleware.
Table 5: Application Server Middleware shares 2006-2007480
Market shares % Vendors 2006 2007 IBM [40-50]*% [40-50]*% Oracle [20-30]*% [20-30]*% Microsoft [5-10]*% [5-10]*% Fujitsu [0-5]*% [0-5]*% Hitachi [0-5]*% [0-5]*% Sun [0-5]*% [0-5]*% Others [10-20]*% [10-20]*%
777. IDC's definition of application server middleware includes application server software platforms as well as transaction processing monitors ("TPMs"). Those two sub- segments have closely related functionality that may make the product offerings substitutable in some instances. For 2007, the TPM-only market shares as reported by IDC were IBM [70-80]*%, Oracle [10-20]*% and Sun [0-5]*%.481
778. Table 6 below indicates IDC's worldwide market share estimates for Oracle, Sun and their main competitors in relation to application server software platforms.
475 Application server middleware is software that allows multiple applications on a computer system to perform a variety of tasks: communicating with a common database, managing interactions, prioritising use of system resources.
476 Enterprise portals are web-based user interfaces allowing information, people and processes to be integrated across organisational boundaries.
477 Integration and process automation middleware are events-based middleware.
478 ESB is a piece of message-oriented middleware ("MOM") that is used to connect enterprise applications and systems together using real time messaging.
479 BPMS is a category of process automation software that models, executes and monitors processes or workflows which connect information systems, packaged applications, people and organisations with each other.
480 Form CO, page 104, table 7 based on IDC, Worldwide Application Server Middleware 2007 Vendor Shares, August 2008.
Table 6: Application Server Software Platform shares 2006-2007482 Market shares % Vendors 2006 2007 Oracle [30-40]*% [30-40]*% IBM [20-30]*% [20-30]*% Microsoft [10-20]*% [10-20]*% Fujitsu [0-5]*% [0-5]*% Micro Focus [0-5]*% [0-5]*% Sun [0-5]*% [0-5]*% Others [10-20]*% [10-20]*%
779. The notifying party submits that Sun's application server software platform only marginally adds to Oracle's position and that the merged entity will remain subject to competition from a number of players, including IBM, Microsoft and open source competitors (such as JBoss and Apache Tomcat). The notifying party further submits that Oracle and Sun are not particularly close competitors for application servers and that their application servers serve different needs.
780. According to Gartner483, application server software products are predominantly based on either .NET, which is subject to strict bundling with Microsoft's offerings, or Java Enterprise Edition, which is in use by various vendors. Solutions based on one of those two competing platforms484 are dominant, yet still compete to some extent with older TPM and object request broker products and emerging products such as PHP, Ruby, Java advanced intelligent Networks (JAIN) and Java Service Logic Execution Environment (JSLEE).
781. Although IDC's market share figures indicate that the increment in Oracle's market share resulting from the proposed transaction will be marginal, it could be presumed that market shares based on revenues may not be fully indicative of the constraints exercised by Sun's open source products on Oracle. However, IDC's market share figures for application server software platforms do only indicate a very modest reduction for Sun following Sun's recent move to offer its application server product, Glassfish, as open source: The decline in Sun's reported market share from 2006 to 2007 was from [0-5]*% to [0-5]*%.485 Even the higher of these two figures would only add marginally to Oracle's market share. In addition there is no indication whatsoever that the "real" market share of Sun's offering had increased more widely but was not captured due to its availability as open source.
782. The market investigation has also confirmed that from the customers' point of view, Sun's products are not the closest substitutes for Oracle's products in relation to the application server segment. In most instances when one of the parties is named as a company offering a close substitute to the other's products, other competitors are also indicated as companies offering close substitutes. In fact, the market investigation has confirmed that the market is very competitive with many active viable vendors, and that the merged entity will continue to face competition from a number of market players such as IBM, Microsoft, SAP and Fujitsu even in the (somewhat artificially narrowed)
482Form CO, page 105, table 8 based on IDC, Worldwide Application Server Middleware 2007 Vendor Shares, August 2008.
483 Gartner, Magic Quadrant for Enterprise Application servers, 2Q08, ID # G00156200, p. 2. 484 See Case COMP/M.5080 – Oracle/BEA, paragraph 11.
application server software platform segment. Significantly, the market investigation has also indicated that open source providers' products, such as Red Hat's JBoss products, exercise an important competitive constraint in the application server segment. Furthermore, none of the customers or competitors indicated that they expect Sun's middleware products to develop to such an extent that they will become stronger competitors to the other players' products in the application server segment (or in any event stronger competitors to Oracle's products).
3.1.2.2. Enterprise portals
783. Table 7 below sets out IDC's worldwide market share estimates for Oracle, Sun and their main competitors on the enterprise portal segment.
Table 7: Enterprise Portal Software shares 2006-2007486
Market shares % Vendors 2006 2007 Oracle [20-30]*% [20-30]*% IBM [20-30]*% [20-30]*% Microsoft [10-20]*% [10-20]*% SAP [0-5]*% [0-5]*% CA [0-5]*% [0-5]*% Sun [0-5]*% [0-5]*% Others [10-20]*% [10-20]*%
784. The notifying party submits that Sun's market share in relation to enterprise portal software does not significantly strengthen Oracle's market share in this area and that the merged entity will remain subject to competition from a number of players. The notifying party further submits that Oracle and Sun are not particularly close competitors in enterprise portal products and that their enterprise portal products serve different needs.
785. Oracle's share of the enterprise portal software segment is just over [20-30]*%. Moreover, IDC's market share figures indicate that the increment in Oracle's market share resulting from the proposed transaction will be small. This finding appears robust even taking into account the fact that market shares based on revenues may not be fully indicative of the constraints exercised by Sun's open source portal product. There is no indication of any kind from the market investigation or elsewhere that Sun's enterprise portal product could be more widely deployed than indicated by IDC's market share figures. The market investigation has also indicated that Oracle's and Sun's respective portal products are not close substitutes in the enterprise portal segment. Moreover, the market investigation has confirmed that the market is very competitive, with many active viable vendors, and that the merged entity will continue to face competition from a number of significant market players, primarily IBM, Microsoft, SAP and Computer Associates, in the enterprise portal segment. Furthermore, none of the customers or competitors indicated that they expect Sun's products to develop to such an extent that they will become stronger competitors to the other players' products in the enterprise portals segment (or in any event stronger competitors to Oracle's products).
486 Form CO, page 107, table 9 based on IDC, Worldwide Enterprise Portals Software 2008-2012 Forecast Update and 2007 Vendor Shares: A New Landscape, July 2008.
3.1.2.3. ESB Software
786. Table 8 below sets out IDC's worldwide market share estimates for Oracle, Sun and their main competitors on the ESB segment.
Table 8: ESB and connectivity middleware shares 2006-2007487
Market shares % Vendors 2006 2007 IBM [20-30]*% [20-30]*% Oracle [10-20]*% [10-20]*% Software AG [10-20]*% [10-20]*% TIBCO [10-20]*% [5-10]*% Sun [5-10]*% [5-10]*% SAP [0-5]*% [5-10]*% Microsoft [5-10]*% [5-10]*% Others [20-30]*% [20-30]*%
787. The notifying party submits that the proposed transaction will not significantly impede effective competition in the ESB software segment. In particular, the notifying party submits that IBM will remain the market leader after the transaction, whilst other competitors will continue to impose competitive constraints on the merged entity. The notifying party submits that the ESB segment is experiencing growth and the open source "mule" and JBoss products by MuleSource and Red Hat have been identified by IDC as strong enough to be considered good alternatives to commercial solutions488. Furthermore, the notifying party submits that Oracle and Sun are not particularly close competitors in ESB software and that their ESB software serves different needs.
788. In addition to the relatively small combined share of the parties' products (even if taking into account the fact that the numbers reported for Sun's market share may not be entirely accurate due to the open source nature of some of its product offerings), the market investigation has indicated that from the customers' point of view, Sun's products are not the closest substitutes for Oracle's products in relation to the ESB segment. In most instances when one of the parties is named as a company offering a close substitute to the other's products, other competitors are also indicated as companies offering close substitutes. In fact, the market investigation confirmed that the market is very competitive, with many active viable vendors, and that the merged entity will continue to face competition from a number of significant market players such as IBM, TIBCO, Software AG, Microsoft and Progress Software in the ESB segment. The market investigation has indicated that open source providers' products, such as Red Hat's JBoss products, also exercise competitive constraints in the ESB segment. Furthermore, none of the customers or competitors indicated that they expect Sun's products to develop to such an extent that they will become stronger competitors to the other players' products in the ESB segment (or in any event stronger competitors to Oracle's products).
487 Form CO, page 109, table 10 based on IDC, Worldwide Application Deployment Software 2007 Vendor Shares: Growth Continues to Accelerate, August 2008.
488 IDC, Worldwide Enterprise ESB Software 2008-2012 Forecast Update and 2007 Vendor Shares: A New Landscape, July 2008.
3.1.2.4. Process automation middleware (BPMS)
789. Table 9 below sets out IDC's worldwide market share estimates for Oracle, Sun and their main competitors on the BPMS segment.
Table 9: Process automation middleware shares 2006-2007489
Market shares % Vendors 2006 2007 Oracle [5-10]*% [10-20]*% IBM [10-20]*% [10-20]*% ACI Worldwide [10-20]*% [10-20]*% TIBCO [5-10]*% [5-10]*% Software AG [5-10]*% [5-10]*% Adobe [0-5]*% [0-5]*% Sun [0-5]*% [0-5]*% Pegasystems [0-5]*% [0-5]*% Microsoft [0-5]*% [0-5]*% Others [40-50]*% [40-50]*%
790. The notifying party submits that Sun's market share in relation to BPMS products does not significantly strengthen Oracle's market share in this area and that the merged entity will remain subject to competition from a number of other market players.
791. In addition to the relatively small combined market share of the parties' products (even if taking into account the fact that the numbers reported for Sun's market share may not be entirely accurate due to the open source nature of its product offering), the market investigation has indicated that from the customers' point of view, Sun's products are not the closest substitutes for Oracle's products in relation to the BPMS segment. In most instances when one of the parties is named as a company offering a close substitute to the other's products, other competitors are also indicated as companies offering close substitutes. In fact, the market investigation confirmed that the market is very competitive, with many active viable vendors, and that the merged entity will continue to face competition from a number of market players such as IBM, TIBCO, Pegasystems, Software AG and SAP. The market investigation has indicated that open source providers' products, such as Red Hat's JBoss products, are also seen as exercising competitive constraints in the BPMS segment. Furthermore, none of the customers or competitors indicated that they expect Sun's products to develop to such an extent that they will become stronger competitors to the other players' products in the BPMS segment (or in any event stronger competitors to Oracle's products).
3.3. Conclusion
792. In the light of all the above, it is concluded that the proposed transaction will not lead to a significant impediment to effective competition in relation to the overall market for middleware or any of its possible sub-segments.
489 Form CO, page 111, table 11 based on IDC, Worldwide Process Automation Middleware 2007 Vendor Shares, September 2008.