Whatever happened to the $14.8bn enterprise portal market?
It was Merrill Lynch that kicked off the mega-hype surrounding a new category of business software that came to be commonly known as enterprise portals. Triggered by Merrill Lynch's Enterprise Information Portals report, first published in November 1998, particularly its exciting market forecasts of a $14.8bn market for portal software and services by 2002, there has been a subsequent surge of interest in portals that has led to well over a hundred companies, from start-ups to established IT vendors, embracing some form of the "portal" concept.
With hindsight, Merrill Lynch couldn't have got it more wrong. The numbers crunched may still be feasible, given that the modern enterprise portal draws in potential revenue from a wealth of different technologies that have experienced rapid growth. But Merrill Lynch's report could not have foreseen how the landscape has changed. Since the enterprise portal market formally emerged over six years ago, the reality of the market is in stark contrast to Merrill Lynch's vision of a lucrative investment opportunity. A clear leader in the enterprise portal market has yet to emerge, the vendor landscape has consolidated significantly, and portal technology has undergone a rapid commoditization.
From a peak of about 80 or so serious players, the enterprise portals market has now been whittled down to about 20 or so mainstays. This rapid consolidation gives strong credence to the emerging consensus that enterprise portals are a discrete market sector that cannot support a sustainable business model. The survivors offer portal technologies that fall into two segments: portal infrastructure and portal application suites. The potential leaders are likely to have a stake in both segments, meeting the specific, but differing, needs of building the underlying portal infrastructure (application servers, data integration) and "assembling" functionally rich portal productivity suites (content management, collaboration) from different enterprise applications.
The rate of commoditization of portal technology has also been alarming, one of the fastest in the software industry. The main reason for this has been the entry of large IT platform providers such as IBM, BEA Systems, SAP and Oracle, which are all providing portal services as an add-on feature or component of Web application servers. IBM and BEA, which effectively "own" the application server market, are most likely to come out on top, and both are starting to make their presence felt because of their technology fit, market presence and sales execution. But they also face a challenge from the major ERP providers--PeopleSoft, SAP and Oracle--and several other vendors, notably Microsoft, Sun Microsystems, and Vignette that have positioned themselves and their new and existing technologies in the portal market. These vendors are not far behind, and with proper execution, can gain competitive market share as well.
The pressure from the large platform plays has taken its toll on the so-called "pure-play" portal vendors--Epicentric, Corechange, Plumtree, Viador, to name but a few.
In the early stages of the market, many pure-plays found themselves competing not against each other, but with in-house IT departments who leveraged existing Internet infrastructure (notably corporate intranets and URL systems) to build enterprise-like portals from scratch. Most of these pioneering portal vendors have now been acquired, fallen by the wayside, or are nearly out of business. Once the dust from Merrill Lynch's report cleared, there was a suspicion that these start-ups are merely upstarts, destined to be overtaken by larger, more conventional software suppliers that have added portal functions to their core products. In 2002, pure-plays have been pressed to consider suitable exit strategies under a barrage of financial and market pressures. All eyes remain focused on publicly held Plumtree Software as the sole surviving challenge from the "pure-play" camp.
Where Have All the Pure-Plays Gone?
Over the past three years the portal market has been consistently peppered by consolidation. One the most significant developments has been the disappearance of the pure-play vendors. Most have exited the market, first by IPO and more recently by merger & acquisition or financial ruin. Others continue to operate, but in a much diminished capacity.
Arguably the first major consolidation in the enterprise portals market occurred in May 2001, when SAP bought TopTier for its HyperRelational Navigation technology, which fuses structured data, unstructured data, and the Web using the concept of Web hyperlinks. TopTier's technology subsequently formed the core of SAP's portal offerings. SAP even restructured itself to take advantage of the new market opportunities by setting up an independent division called SAP Portals that was responsible for driving the adoption of its portal strategy to the SAP ERP customer base and beyond. However, in a dramatic U-turn, the division was dismantled and SAP's portal technology was once again brought into the corporate SAP fold.
Since then, there has been a string of other consolidation activities. Some of the early consolidation activity in the market was driven by technology acquisitions. The later ones were driven primarily by financial concerns, and the fate of Epicentric and Corechange are good examples of this.
The lack of more than one original pure-play offering signals that this technology is settling into a place as a middleware feature and is not likely to continue long term as a separate market outside suite offerings that include combinations of collaboration, content management, user management and integration capabilities.
This early consolidation was driven primarily for technology gains, rather than financial worries. Some of the main vendors and events are highlighted below.
DataChannel was acquired by Netegrity, a business security management software company, in October 2001 for approximately $50m. The acquisition was part Netegrity's grand strategy to build single sign-on (SSO), access control, portal and provisioning into a unified and "secure relationship management platform". A year later, DataChannel, whose DataChannel Server product includes bi-directional data access and integrated content management with search, workflow and version control, dropped off the radar screen of the mainstream portal market.
InfoImage, which focused on providing "high-velocity" decision-making portals, was purchased by Serviceware Technologies, a provider of Web-based knowledge management solutions for customer service and support. InfoImage had previously filed for bankruptcy protection, and following its takeover, elements of the Decision Portal technology were absorbed into Serviceware's broadly focused knowledge management (KM) offering.
XML-based portal specialist Sagemaker was acquired in April 2001 for $16.5m by Divine, a collaboration and content management vendor, and subsequently became victim to Divine's own financial woes which led to a Nasdaq delisting, a filing for Chapter 11 bankruptcy protection, and the sale of its assets to Golden Gate Private Equity in April 2003. Divine's woes largely stem from its inability to effectively execute a consolidation strategy after acquiring a host of companies and technologies over the past three years.
Sequoia Software, a developer for XML portal technology, was bought by Citrix in March 2001 for $185m. The move is part of Citrix's strategy to deliver Web, Windows or Unix applications to any kind of device, wired or wireless. Sequoia's XPS portal software is included as part of Citrix's NFuse Elite and newer MetaFrame Secure Access Suite enterprise access infrastructure software offerings.
Sterling Software, which bought the MyEureka line business intelligence web portal technology originally developed by Information Advantage, was itself acquired outright by Computer Associates in May 2000 for about $4bn. The Eureka portal product is now branded as CleverPath Portal, and is languishing as part of CA's vast and unfocused product set without making much impact or headway in the market.
Viador broke new ground back in 1998 by launching BI-Portal, the industry's first decision support-focused portal offering a range of Web-based BI such as reporting, ad-hoc query, and multidimensional (OLAP) analysis, held an IPO in 1999, and since 2001 has fallen into financial trouble. The company's stock-price crashed, forcing the company into a merger agreement with MASBC Acquisition Corp in September 2002, and was once again spun out as a privately held company. The company is hanging on by its fingernails and is now repositioning its software offerings as a BI solution rather than an enterprise portal.
In December 2002, Epicentric, another of the original portal pioneers, took the hint and finalized its own acquisition by Vignette, a content management software provider that was itself looking for a new lease of life following a string of poor results. Epicentric had concentrated on providing an infrastructure for the integration of applications and services for refining how content is presented to users. Its approach is to supply a set of "federated" interface components that make it easy to plug in third-party applications, content, and services to the portal. The company has traditionally focused on the B2B and B2C (e-commerce) portal market.
Vignette confirms that Epicentric was running at around $9m a quarter prior to the acquisition, which makes the $26m (rising to $32m under certain conditions in subsequent payments) that Vignette paid for Epicentric sit at around one times revenue. One suspects that Plumtree would kill for that kind of valuation instead of its value of cash plus one-sixth revenues. But it would prefer it in its share price, not in a merger and acquisition offer. Unless Plumtree does sell out to a bigger player or to the type of asset chasing opportunists that bid for it in September, then it cannot expect such a valuation, unless the business suddenly and unexpectedly gets back into growth.
Corechange develops a portal framework built around the Microsoft platform and offers credible content, collaboration and community technology. It was bought in February 2003 by Open Text, a provider of collaborative knowledge management solutions.
The move was hardly unexpected. Throughout 2002 Corechange had largely been preoccupied with raising funds for survival. The company had also failed to close a few pivotal deals or create a unique claim or position in the market, though it was building up a presence in the healthcare sector. The tiny portal specialist tried in vain to go public as early as May 2000, when it filed a share registration statement with Nasdaq. Still a privately held concern, this document was the only source of hard financial information about Corechange. It stated that the company reached a revenue peak during 1998 of $5.6m, while losing $7.4m in its reported quarter ended March 31, 2000. Hence the company had been running at a miserly $1m a quarter while losing almost $2.5m in the process. Little wonder that it decided to pull its IPO.
Apart from the blue chips in its customer base led by General Electric, there had been little to suggest that Corechange could continue to challenge the new portal order. Venture capital backers may well cash in their investment sooner rather than later if they sense that the tide is turning against specialist firms. But getting back the full $75m invested in Corechange, at a time when the world values Plumtree below this figure, is just not feasible.
A likely buyer for Corechange would have been a vendor seeking to plug a technological gap in their own products, and Open Text stepped in with its check book in a bargain $3.6m deal. Open Text is leveraging Corechange's CorePort engine to drive what it calls "contextual continuity"--which specifically refers to the ability to move from one portlet to another and still retain context. Open Text says it will continue to market Coreport as an independent portal alongside its own Livelink-specific Unite portal offering. But in the future both products will most probably be integrated into a single portal product.
Following Open Text's acquisition of Corechange, the lone remaining pure-play company in the market is Plumtree Software. As a founding member of the portal market, Plumtree remains a perennial favorite (in terms of mindshare at least) and, despite growing competition, typically ends up on many RFP short lists.
However, Plumtree is struggling to redefine itself in the wake of massive consolidation in the market combined with continuing pressure from infrastructure vendors--notably IBM, BEA Systems, and (increasingly) Microsoft. As these large platform vendors, with their equally large installed bases of software solutions like application servers, integration servers and databases increase competition by bundling portals with their other offerings, it is unclear whether the access, collaboration and content managementcentric approach Plumtree has opted for will stand up.
But unlike other portal pure-plays that have easily succumbed to direct onslaught of the infrastructure players, Plumtree is not in any danger of exiting the market soon, functionally at least. Plumtree has broadened its portal capabilities while maintaining simplicity to remain viable. Also, collaboration, content management and communities continue to gain importance as the core business value delivered by the portal. Hence, companies that need a bipartisan offering--one that aggregates URL-addressable interfaces without additional coding from ERP, CRM and other applications while facilitating community-based activities and interactions--will favor Plumtree for portals that manage information, help users collaborate or federate other portals together.
Plumtree ultimately faces strategy choices as either a platform/infrastructure player, or through vertical market focus, as a provider of specialized portal technology and tools. Whichever route it chooses, Plumtree faces several key issues and challenges, not least providing more robust integration, messaging, caching and transaction protection capabilities, building up a strong professional services organization that can support enterprise-scale implementations at a high enough margin to keep investors happy as well as build vertical market-specific or solution-oriented portals rapidly, and figuring out how to shorten sales cycles.
Architecturally, Plumtree must also compete against strong J2EE platform incumbents--remember, Plumtree's ace card is Microsoft's ASP/.NET. To effectively challenge the likes of IBM, BEA and the other platform portal vendors, Plumtree must therefore focus its cross-application server platform message on companies that need Java and Windows applications to coexist in the portal. Plumtree needs to show how it can aggregate Java applications running as services with those built on .NET in a way that minimizes the perception that its approach is somehow "proprietary". While Web Services standards will help promote this capability, problems like how Web Services will handle transactions and security need to be solved before Plumtree's different architectural approach (compared to the more J2EE-centric platforms) will no longer matter.
Executing such a strategy will be difficult at Plumtree's size, particularly given the flattening in revenue growth it has experienced since its IPO. This leads to the question of the company's long-term financial viability. Plumtree is valued at only $10m more than the cash it holds in the bank, well under one times last year's revenue or even one times this year's diminished sales. Last May, Plumtree went public in one of the few sectors of the technology markets able to mount an IPO. Before the IPO Plumtree had been growing at over 100% per annum, from $34m to $80m in the fiscal year to December 31. Since then Plumtree's revenue has started to shrink. Last September the company even had to fight off an audacious, unsolicited offer for its business from Sutter Capital Management that would have taken it into private hands once again. That offer valued Plumtree at $5 a share, making the offer worth just under $147m. Plumtree, not surprisingly, rejected the offer outright. But its shares are languishing at almost exactly half that value, despite the fact that it has cut back quite successfully in order to break even.
Given the state of its finances, it is hardly surprising that certain parts of the market believe that it needs to be acquired. But who could land it? IBM is always in the frame, even though its current size and momentum in the portals market does not require it to buy market share. Vignette has already made an acquisition, and other emerging portal players like Broadvision simply not big enough yet. That leaves Microsoft, Sun, BEA, or Sybase as the only major portal companies in the market that have the right type of market capitalization to buy Plumtree outright for shares and some cash without batting an eyelid. All these players have the motivation to gain the market lead that the buy might give them.
Under Fire From Platform Plays
The encroachment of larger IT infrastructure-oriented vendors has caused many of the pure-plays to run for cover and go on the defensive. Throughout 2002, a variety of software development forces pulled portal technology further into the corporate IT infrastructure "stack". Hence, it has been the tall order of integration and functionality required that has prevented the pure-plays from getting too far ahead in the market. However, these capabilities are well within the scope of the larger, more established IT providers that have now waded into the market with broad and comprehensive portal infrastructure platforms and frameworks. These vendors can compete in the long term because they can deploy more resources than smaller or financially weaker vendors, especially now that portal products are defined more by the "middleware" they package, and less by the quality of their presentation frameworks, an area where most have drawn fairly even and which has become highly commoditized.
Companies such as IBM, BEA Systems, Sun Microsystems, PeopleSoft, SAP, Oracle, and Microsoft are among the leading high-end players that have embarked on portal strategies, most of them with homegrown technology. Throughout 2002 many of these companies have convinced portal users in significant numbers to go for a trusted, brand-name supplier and buy their portal software as a part of a wider enterprise-wide purchase. The pure-play specialists cannot hope to compete with the marketing might and resources of these players.
The major application server/platform vendors find a prominent position on portal short lists, as do the leading ERP application providers. This is because organizations seeking to implement portals want to enter into a strategic or long-term relationship with dominant vendors.
IBM became a formidable contender when it made WebSphere Portal Server its sole portal offering, repackaged it with its application server, and combined Lotus, Tivoli and DB2 technologies in a supporting role. The company had previously offered two seemingly overlapping portal solutions: IBM EIP and Lotus Development (K-station), which led to considerable confusion among customers and partners alike.
Because of the close links between application servers, portals, collaboration and data integration, IBM is strongly positioned to capture market leadership. It claims have garnered more market share than any pureplay supplier, and many hundreds of installations (though it refuses to say precisely how many).
Given its size and its ability to market and sell, IBM challenges the rest of the field on breadth and comprehensiveness of its offering. IBM offers the broadest range of native technology, from collaboration, content and knowledge management, analytics and application development. IBM also supports a range of software partners and is not averse to integrating with other, sometimes rival, technologies. This openness stems from the vast resources of IBM's Global Services professional services/consulting division, which possesses the size, focus and breadth of resources to build, architect, and assemble almost any type of portal. IBM added over 50 new portlet development partners for WebSphere last year, bringing the total up to 400. These portlet partners are ISVs or information provides that encapsulate their applications or information offerings in a "portlet" that can be embedded, surfaced, and managed through the IBM portal. Key application partners include Business Objects, Bowstreet, JD Edwards, Cognos, RSA, and WebEx Communications. IBM also maintains content partnerships, including the likes of FT, Factiva, Reuters-Dow Jones information subsidiary, Dialog, and Newsedge.
The technical challenge of integrating these applications and content providers is not the issue here. Rather it is the marketing message that it creates. IBM now sees portals as a core focus, both in terms of a new market opportunity, and as a way to bring together its various technology divisions (content management, analytics, collaboration, systems management) into a single manageable solution. But simply bringing these disparate technologies into the WebSphere brand is not enough. Organizationally, IBM continues to struggle to get its vast development divisions to share technology and work well together. For example, in its portal offerings, many of the components are not as well integrated as they should be--simplified integration with development tools and systems management are lacking, making Visual Studio and Tivoli products look like an afterthought. With the Global Services organization behind it, IBM is also giving out a message to the market that building a portal is now point-and-click and IBM will install it and look after it for you.
BEA's portal activities in 2002 mirror IBM's initiatives. BEA positioned its portal software under the umbrella of its competing application server WebLogic. BEA has over 700 "portal" customers, and recently claimed that its portal license revenue exceeded the quarter's license revenue of Plumtree. BEA says that about 25% of portal-related software revenue comes from its consultancy partner channels--such as Deloitte and CSC.
BEA maintains a competitive position in the portal market because it effectively "sells back" into its application server installed base. It also commands a separate value with portals providing a so-called "front end" to application servers, a market BEA currently dominates along with IBM.
BEA can make a strong claim for delivering the most efficient, productive, and tightly integrated Web application development platform in the portal market. BEA's application server, integration server, tools and portal server integration are peerless in terms of seamless integration. This translates to reduced cycle times between design, development, implementation and deployment for externally facing, high-volume portal sites, with a resultant lower cost of ownership. While its offering is not as strong as IBM's in terms of collaboration, content management or information retrieval (search, classification, taxonomy generation), it does maintain several key OEMs or co-selling partnerships to cover these capabilities adequately.
BEA is certainly strongest in the build-your-own-portal category. But it risks being labeled a "coders-only" portal solution. BEA clearly recognizes this and provides strong support for standards, a rich set of APIs, and tag libraries, and a unified installer that make portal coding less complex to minimize this. Additionally, BEA is also starting to forward a more agnostic approach to integrating technologies from third parties that specialize in information management, collaboration or content management.
BEA's long-term strategy is focused on moving away from simply pushing its application servers and into a broader base of application middleware that includes greater synergy between application servers, portals, and integration servers. The company certainly has the size and the track record to become a dominant player in the portal market, and many buyers will be confident of selecting WebLogic as a foundation for their enterprise portal initiatives. But in order to achieve market leadership, BEA must continue to execute on technology leadership, its integrated road map, and building its partnership and sales momentum.
PeopleSoft's portal strategy centers around the concept of an employee-facing portal called B2E. This strategy focuses more on assembling pieces of applications and processes together in a unified, and more productive employee "Webtop" than on an application server development platform. The real proof of ERPled portal solutions such as PeopleSoft will be the ability to drive adoption outside the non-ERP installed sites. So far, PeopleSoft, SAP, and Oracle have all met with limited success, but their respective installed bases provide a lucrative market in themselves, in the short term at least.
The foundations for PeopleSoft's portal strategy can be traced back to the release of version 8 of its ERP suite. Version 8 effectively rearchitected all the PeopleSoft applications and tools to run on the Web. This made its applications ideal for portal deployment. Its philosophy that the application owners, rather than portal developers or administrators, should customize and manage ERP applications will resound nicely in the ears of organizations that do not want to embark on an expensive portal project.
PeopleSoft's portal technology is sound, but the introduction of its AppConnect suite of products needs to be differentiated to avoid like-for-like comparison against the larger platform vendors. AppConnect is an umbrella technology that includes the portal, PeopleSoft's data warehouse/BI products, and an integration product. The goal of AppConnect is to make PeopleSoft's technology core to the business platform. Whether as the de facto user interface to its ERP application, or as a key component of core infrastructure, the PeopleSoft portal is key to this strategy.
SAP's portal offering is not yet a direct replacement for more established portal offerings like Plumtree--though the company hopes to get there in the next two years. Like PeopleSoft, SAP is also focused on B2E productivity portals. In 2001, SAP merged the TopTier acquisition with the mySAP work space into a revised portal product that included new search, classification, content management and rudimentary collaboration capabilities.
As things stand today, SAP's portal is targeted at those organizations with a strategic commitment to SAP and primarily run SAP's ERP systems. These customers typically turn to the SAP portal where supporting process automation (in SAP terminology, "unification") is a key requriement.
Overall, SAP's portal strategy has been haphazard to say the least, and has been accompanied by a confusing strategy of changing brands and product releases. With the 5.0 release SAP was poised to challenged for leadership in both breadth of technology offered and vision. But an ill thought-out merger between SAP Portals and SAP eMarketplaces, and then the absorption of the subsidiary as a business unit into the parent SAP AG, meant that plans to roll out new portal technology and the Cross Apps (xApps) solutions slowed, and initial implementers struggled with the first-generation technology components. Version 6.0 attempts to address the immaturity issues. In order to speed-up adoption, SAP now offers its portal as the front end to employee and manager self-service packages. But if organizations require more robust content management, collaboration or application development tools they must "federate" out to other portal technologies. SAP's new developed content management, search/classification and collaboration have been developed internally and are still relatively immature.
However, there are a number of merits to SAP's portal. SAP is particularly strong at integration to its own R/3 business objects and its BI products. It also has a large set of business adapters and both a Java and ABAP-based development environment to extend the portal in different directions. SAP also recently announced support for a new integration platform called NetWeaver that supports both J2EE and .NET development. While this potentially strengthens its integration capabilities, the alignment of NetWeaver and the portal is still not crystal clear at this time.
The main influences on Oracle's position in the portal market have been its relational database strengths, an initial focus on Web-based application development, and the Oracle application server. The Oracle portal works well with its own BI/reporting tools, collaboration suite (Oracle Collaboration Suite) and its own ERP applications, as well as providing tight integration with its application server and database. Like other ERP portal providers, the Oracle portal is primarily aimed at Oracle's own installed base, particularly those that want to implement a single, simple-to-use interface to standardize Web access across multiple internal applications and unstructured information with a strong transactional component.
Oracle customers were developing their own portals using Oracle tools and technology before the 9iAS version of the portal product first emerged in 2001 as a credible portal offering and challenger. It leverages its enterprise features like scalability, ability to distribute, manageability, caching and integration to backend applications that also use its enterprise platform. Its portal provides wizard-based portlet/portal layout design tools. It also features distributed site generation/management, as well as improved Web content management features. More wizard-like tools for assembling data views are expected in an upcoming release.
While its portal technology is credible, Oracle has struggled to clearly articulate its portal story so that audiences can identify their needs with what is actually being offered. With its efforts focused on Collaboration Suite, ERP applications, 9iAS and the like, Oracle's portal offering lacks the full extent of its marketing and sales muscle, and often appears as a user interface, rather than portal suite in its own right. Oracle also takes a radically different approach with its application server positioning. IBM bundles its application server with its portal offerings, but Oracle reverses the emphasis, and highlights the value of the application server first, with the portal offered as a feature. Oracle's strategy of charging the same for the portal as the portal and application server combined provides an appealing cost-effective single platform offering for Oracle purists, but this might seem somewhat inflexible to the best-of-breed shops.
Because the installed base of portals is small but growing, there are several vendors that are challenging for market share.
Due to its size and affinity to the desktop, Microsoft could be dominant, except that its Sharepoint Portal is more of a lightweight document management system than a portal. Only those organizations that have a strong commitment to .NET technology and less immediate need for complex server-based application development and integration will find the Sharepoint Portal and Team Services combination useful for departmental or smaller-sized organizations.
Team Services is Microsoft's team collaboration product that allows portal users to create ad hoc workspaces, manage group activities and produce shared deliverables from within the portal. However, the combination lacks the transactional support, integration capabilities, tools or analytics needed to compete with the dominant vendors. For example, application integration requires coding WebParts and presentation layer development that is not insignificant today. Integration with other Microsoft technology like Microsoft Content Manager is also accomplished separately.
To compete for a dominant position among other portal products, Microsoft needs to articulate how Sharepoint will come together with Jupiter. Microsoft's long-term strategy for Content Management Server, Commerce Server, BizTalk and .NET. However, this strategy is likely to take 18 months or more to develop. Scalability is also an issue. Microsoft's portal version also suffers from scalability and user management limitations that make it ideal for departments, but difficult to manage in enterprises. Microsoft-centric organizations considering Sharepoint as foundation portal technology should therefore be prepared for a long cycle until it reaches parity with other platform portal offerings.
Overall, Microsoft seems to view its portal more as a "tool" than a platform, and it prices it inexpensively, even in low volume. Microsoft reports seat deployments in the millions (measured by counting enterprise license deals that involve the portal in addition to other software), but the company struggles to demonstrate customer success beyond departmental settings. Microsoft's focus on .NET and competing in server-side application deployment leaves few resources dedicated to the portal. This could well change once .NET is more broadly deployed, creating the installed base into which more advanced portal technology could be deployed.
Sun Microsystems, as a leading IT hardware and software platform company, should not to be dismissed from consideration in the portal market. Sun brings a broad set of technology to its portal offering. Yet it often does not make it to the top of buyer short lists and is not seen as a front-runner in the market, mainly because portal buyers align with application servers or back-office applications where Sun is not strongly positioned.
The challenge Sun faces is how to turn the components of the portal into something more distinctive. Sun's re-branding from the iPlanet to the SunONE umbrella is a step in the right direction. But repackaging and reported integration issues between some of its non-core components still make it difficult to predict Sun's long-term intentions with SunONE.
With a more solid portal vision, Sun could easily gain a unique position in the portal market with a combined user identity, user management, collaboration and secure access story. For example, its portal and identity servers can front-end a range of portal services, from application, directory and Web servers to Windows, mainframe or X-windows applications, and reduce administrative costs dramatically. New additions to the collaboration capabilities such as secure instant messaging and online awareness also enhances the user experience. These capabilities lend themselves to a highly scalable, distributed architecture that can grow with individual departmental needs.
With financial problems in 2002, Vignette did not take advantage of previous acquisitions or reposition its technology sufficiently to become dominant. With the Epicentric portal, multi-site content management and application integration, Vignette is now taking more of a "suite" approach to address both the content management and portal markets. After struggling with its own portal technology, Vignette has finally decided that buying a pure-play portal vendor (Epicentric) will allow it to execute on this vision than continuing to build on its own. At the same time, Vignette spent 2002 moving its core product to a pure J2EE-compliant platform.
Following the acquisition of Epicentric, Vignette now challenges the portal market with a rich combination of Web site deployment technology, content management, integration and a rules engine. Organizations that require both strong content management and distributed/delegated portal site management should consider Vignette for their portal projects. While giving up on its own portal development, Epicentric now provides better delegated portal administration and management over the multiple sites V7 can generate. The Vignette Application Builder (renamed from Epicentric Foundation Builder) will also appeal to business users that want to rapidly develop portlet-based applications (without writing code) for data access, reports display and simple form-based inputs.
In the long term, Vignette is focused on developing a low-cost alternative for building and assembling composite portal-like applications. Fourth-quarter financial numbers, cash in the bank, and expected profitability in 2003 all indicate that Vignette is in no danger of disappearing. However, the company needs to improve its collaboration capabilities and demonstrate the value of acquired technologies in integration solutions if it is to move into a more dominant position.
Niche Players & Opportunists
Several vendors such as Broadvision, ATG, Sybase, Tibco, and CA are aiming to capture unique niche opportunities with portals specialized around sophisticated Web site management, enterprise application integration, content/reporting integration and application development. Others, notably Hummingbird and Citrix, are veering toward marketing their portal technology using the term portal to describe Web-based user interfaces.
Both these sets of niche/opportunist vendors are likely to reduce their efforts in the mainstream portal market in 2003, and some will exit the market altogether. Notable differences between them and the challengers listed in the previous section are based less on technology and more on factors like size, market visibility, partnerships, strategy execution, and financial stability.
BroadVision and ATG
Over the past couple of years, both BroadVision and ATG have successfully delivered technologies to create sophisticated, online business-oriented Web sites for the internal enterprise. These vendors have repackaged and repositioned their portal offerings to leverage their Web content management, personalization and ecommerce rules engines. At the same time they have also added collaboration, user management and integration into their portal offerings. The problem however was that in 2002 portal deployments did not grow to the size or complexity that would demand the sophisticated combination of these rich feature sets. Hence, both companies have endured serious financial setbacks in 2002, which has made it difficult for them to market their portal offerings that aggressively.
BroadVision has struggled with its business model and portal positioning, staying out of the limelight for most of 2002 to divert attention from its financial woes. Following a recent repositioning, the company has subsequently found itself in a very lonely market, competing against eCRM players with finished applications that run as specialized customer portals.
Meanwhile, ATG has solid portal technology and deployments. But the company's rate of growth in sales and installations is slower than other portal players. ATG is also suffering from its fair share of financial troubles. With layoffs and increased focus on expenses, this trend appears to be reversing, but it will take more than a few successful quarters for ATG to fully recover.
Sybase and Tibco
Sybase and Tibco have gained considerable momentum in the enterprise application integration market, largely through their respective business process modeling, transaction-data mapping and message-based architectures. Sybase also adds database, data integration, and wireless technologies to this mix. Tibco's strength in messaging and integration adapter breadth are key to its success. It provides the portal technology for the Yahoo! portal's instant messenger offering for the enterprise.
Sybase's technology for portal page assembly from its OnePage acquisition, and Tibco's messaging bus architecture are the unique differentiating features of their products. Organizations with sophisticated application integration requirements that wish to expose business processes via customized "zero-footprint" client interfaces will find Tibco and Sybase's portal offerings attractive. But neither vendor offers portals with strengths in content management or collaboration.
While the demand for integration functionality and portals is converging, neither Sybase nor Tibco occupies more than a niche position in the portal market because they focus on adding portal technology to "front end" their middleware solutions where their traditional strengths lie. In both cases, portals provide the interface through which users access the platform functionality.
Computer Associates recently re-branded its entire product line, and now features its Cleverpath portal as part of an overall information delivery platform that combines structured and unstructured information retrieval, analysis and management.
The Cleverpath portal essentially provides the main interface to CA's analytic and reporting tools, rules engine, image/content archiving, and systems management technology, a group of technologies where portals are not necessarily the primary demand driver. To counter this perception, CA is focusing its message around how Cleverpath can enhance the value of information by delivering it in the context of a business process where taking action on the information is key.
CA's strategy is to provide combined offerings. Its assumption is that customers tend to buy its portal product if they have purchased or are inclined to buy another CA product first. But relative to its vast product portfolio, CA's portal division only produces a fraction of overall revenue. Despite not-insignificant spending on advertising and marketing, prospects, system integrators and competitors do not generally list CA in their top 10 portal providers. CA is likely to reposition its portal product as a process-centric interface to its vast portfolio of software products--and not as a portal to challenge the infrastructure or application segments.
Hummingbird has a wide collaboration, document management and reporting product offering branded as Hummingbird Enterprise, with these components well integrated by a portal called Hummingbird EIP. However, has been painfully slow to demonstrate a successful portal deployment strategy outside its installed base. The reason for this is simple: Hummingbird's strength is as a combined suite and not as a standalone portal. Hence, organizations generally select Hummingbird's portal based on how it fits with their full offering, and use the portal component as a unified access point to suite capabilities. However, as Hummingbird continues to adopt a more open Web services model, the issue of whether it has a portal or not becomes increasingly moot.
Hummingbird promotes the concept of a "Smart Enterprise Suite" to differentiate itself in the portal market. But demand for collaboration, content management and information retrieval capabilities in the portal is unlikely to develop without some supporting application integration and tools. While a Smart Enterprise Suite may include a portal, implementations increasingly need the value of integration, application access and transactional support, which are not among Hummingbird's core competencies. Moreover, most large enterprises will already have some form of search, content management, and collaboration technology in place. These companies will be reluctant to replace it with a single package, but rather will look for a portal to work with what is already in place.
Citrix acquired portal technology when it purchased Sequoia. Since then the company has been struggling to position its remote access portal offering uniquely in the market. Citrix's XML-based portal technology nFuse is ideally suited for organizations that want to offer access to both non-Web and Web-enabled applications jointly from the same interface. Hence, nFuse Elite is less of a portal and more of a portalenabling technology for existing applications. But without the integrated development capabilities of Bowstreet and others, Citrix's portal technology is more like a tool suited for those who want to enable existing applications rather than build new ones. As things stand, nFuse will have little impact on the mainstream portal market.
Silverstream and Bowstreet
Silverstream and Bowstreet represent a niche-oriented category focused on portal developers. Both these vendors provide portal development tool/IDE technologies oriented toward code writers who require tools to help them build "portal-like" application interfaces from scratch for custom Web-based applications. From a pure development perspective, Bowstreet and Silverstream therefore simply offer an alternative to more mainstream Web developer tools, such as BEA's Workshop, Borland's JBuilder, IBM's Visual Age, or Microsoft's VisualBasic/.NET Studio.
If all the claims of the leading portal vendors are true, then every one of the Global 2000 already had at least one enterprise portal package. And if market growth were to be maintained, which seems unlikely, by next year they would all have at least two. This scenario is supported by Forrester, which predicts that 60% of companies will have a portal initiative underway this year, and also Meta Group, which expects over 85% of the Global 2000 to have selected a portal framework by 2004. Meanwhile, Gartner's magic quadrant, true to form, gives all the larger portal players its usual bland endorsement.
Chances are that Plumtree will bow out gracefully from the pure-play market in the next year or so. On the other hand, every new market yields a single survivor, and perhaps Plumtree is the one. With $63m in the bank, and a last quarter run rate of around $68m annualized, Plumtree does not have to choose a partner any day soon. It can feasibly survive on its current cash burn, possibly forever if nothing else changes.
Regardless of the Plumtree's fate, more changes are in store for the enterprise portal market. Chances are that by the end of 2003, none of the major analyst firms will even be citing portals as a separate market as they make their predictions for 2004. The resultant pieces will not disappear. Rather, they will be subsumed into other enterprise software markets, and forecasts. The financial messages coming from the pure-play vendors certainly points in this direction, and investors tend to agree.
From a technology perspective, portal choices will simplify in 2003, splitting into two primary segments: "build your own infrastructure" from platform technologies, and "assemble the components" from existing applications. The first segment relates specifically to Web application development, specifically forwarding portal platforms, frameworks, and tools to primarily build custom, transaction-based portals. Here, the focus is on tight integration for supporting transaction portal applications such as customer self-service, supply chain and partner management, and online selling. The second segment relates to pre-configured, out-of-the box features such as process workflows, content management and collaboration, usually in support of internal B2E portal sites or more specific applications such as e-commerce or information management.
The size and market strength of the incumbent (infrastructure) vendors points to the major software development platforms absorbing portal technology, along with application and integration servers. Consolidation down to two or three major players is unlikely during the next year or so. This is because there is still strong demand, particularly from departments and small and medium-sized enterprises (SMEs), for configurable portal solutions that do not require heavy investments in programming to develop, install and manage. For example, smaller players such as Plumtree will continue to experience a degree of mindshare, if not necessarily market share.