Portal Proliferation

For many organizations, enterprise portals have proved the ideal way to organize information and applications around the needs of individual users. Many users have adopted portals as the de facto application platform within the enterprise.

Changing expectations in the marketplace have created a demand for portals and Web applications that deliver highly detailed information from back-office systems, such as inventory level, customer-specific pricing data, order shipment status or order shipment date. And there's one-upmanship in the marketplace. Any CIO able to deploy an effective system that is superior to a competitor's finds himself driving increasing customers' expectations of the portal and the Web application at the same time as he is forcing his rivals to dramatically improve the quality of their Web experience. Derome says that one-upmanship is the rule rather than the exception, and will only increase.

Yet that very proliferation of Web applications and portals is forcing many companies and their CIOs to deal with the challenge of consolidating numerous Web application assets and consolidating many different portals. They are learning the hard way that trying to consolidate those different systems without exploding their integration costs, or driving huge labour expense to tie those systems together, is a significant challenge.

"When portals were first proliferating in the marketplace oftentimes decisions were made at a business-unit level," Derome says. "You would have a sales team building a customer service portal, and then maybe your supply chain team building a portal for suppliers, and then maybe your HR team sponsoring some kind of an effort to build an employee self-service Web application. And each of those was oftentimes built on a different infrastructure, or a different vendor platform.

"When a CIO takes a look at those three different systems, and decides that it makes sense to consolidate the three on a single platform, they have two choices: They can either integrate the three on a fourth platform - a new platform - or else go through a rewrite exercise where instead of integrating the three they would write a new application and bring all of those pieces together on a new platform. It is typically a fairly complicated exercise."

Both choices are problematic, and the problem is by no means insignificant. When US InfoWorld launched an ambitious effort to complete an in-depth review of the seven major portal servers, long-time contributor Mike Heck spent many months doing everything from digesting mounds of documentation, to installation, configuration, integration and creating test portals. He found by far the greatest challenge involved setting up portlets, or portal applications, to talk to one another. "The promise of portlets is that you plug in the portlet and it will just run." As so often happens, reality falls short of promise. "It's not plug and play just yet," Heck says.

Still, InfoWorld concludes, given the potential benefits that portals can provide, all that heavy lifting is probably worth the effort. Heck's own experience at Unisys reinforces his view that it is common for organizations to go through multiple portal initiatives before they get it right. "You can expect a few false starts," he says. "But there's benefit to that. No pain, no gain."

Portal integration costs are an important yet rarely quantified variable in any portal implementation, Derome says. Costs differ from vendor to vendor, and at a higher level, from vendor segment to vendor segment. At the same time, pressing issues such as the impact of budding Web services and portlet integration standards, and pressure from large application and infrastructure vendors, is transforming the dynamic of the Web application interface market and forcing stand-alone vendors to modify their models. Companies, meanwhile, are increasingly recognizing the urgent need for new approaches to these technologies.

Throughout the late 1990s and through to the end of 2001 most of the activity in the marketplace was less considered, less strategic and more a simple reaction to the technology change. Companies today are taking a far more cautious, measured and calculated approach.

That change in approach has not come before its time. There are two factors that have encouraged companies to take a more considered approach, according to Derome. First, because of technology implementation failures and a spectacularly high failure rate for technology projects, finance is now increasingly involved, with its obsession with applying real rigour to technology investment decisions. The inevitable outcome is that that rigour has now filtered through the portal and Web application markets, as companies give fresh consideration to issues like total cost of ownership and return on investment, rather than reacting to the new technologies and the promise that new technologies provide. The new focus has significantly changed how people buy Web applications and portal technologies.

The second major shift in the marketplace is maturation on the vendor's side or on the supply side. The vendor community is no longer promoting technology for the sake of technology. Instead, vendors are looking into their customer base and trying to understand what specific business problems they can solve. As they increasingly position their technology to solve those very specific and unique business problems, they force changes in the way that the end user or the customer buys the technology.

Given the complexity of the market, and the continual emergence of new solutions for managing and presenting content and information, integration has become a critical factor. At the same time the presentation-layer applications offered by Web content management and portal vendors pose difficult questions about the future of the market.