Trends in Content Integration: From a Hub-and-Spoke Model to Content Mashup Model
InfoManagement Direct, June 27, 2008
Content never stops growing. Research firm IDC estimates that by 2011, enterprises will be responsible for 1,530 exabytes of digital information - a 10-fold increase in only five years.1 As enterprises gain both more content and additional kinds of content, they also face significant fragmentation. According to a 2005 study by Forrester Research, 78 percent of companies have more than one content repository, while 43 percent have more than six.2 This number has probably jumped significantly since then, driven by new forms of digital content as well as corporate mergers and acquisitions.
Advertisement
Providing easy access to all this content becomes more challenging every year, but it’s essential for business success. IDC found that the typical information worker spends close to 25 percent of their time searching for and accessing content – and can waste up to three and a half hours per week looking for information that is never found.3 Meanwhile, the Web has led both information workers and consumers to demand seamless communication and collaboration – forcing the enterprise to support real-time sharing of content and ideas. In this new environment, deploying a comprehensive content integration system that permits enterprise-wide searching and content access is a must.
A variety of different solutions have emerged to address these needs, from basic consolidation (migrating all content to a single enterprise content management, or ECM, system) to a “hub-and-spoke” model that connects repositories based on different ECMs through common connectors. The most flexible and powerful strategy for today’s enterprises, however, is the new content mashup” approach.
Hobbled by Hub-and-Spoke
Although consolidation offers some advantages – such as the ability to easily search for content through a single interface and the need to administrate only one system – it is ultimately far too limiting for large enterprises. No matter what companies claim, in practice ECMs rarely handle all kinds of content well, which creates problems when working with unusual files. Combining existing repositories is also time-consuming and expensive. It’s never-ending, because new acquisitions require further migrations. Finally, consolidation makes organizations heavily dependent on a single vendor, which could cause problems if the industry transitions to a new platform.
The hub-and-spoke solution was proposed as an alternative to consolidation. In this model, standard connectors are used to connect disparate repositories to a master content integration platform. Users can then access content through a single interface, such as an enterprise portal. Hub-and-spoke should allow different divisions or geographies to maintain separate content repositories, while providing users with the benefits of a single sign-on and unified search.
Unfortunately, hub-and-spoke has fallen short of expectations in many respects. While the Java-based JSR-170 and JSR-283 standards were developed to encourage the creation of widely accepted, off-the-shelf connectors, they have not been broadly adopted - forcing companies to custom-build their own. Applications from alternate platforms such as .NET or C++ also won’t work effectively with either standard, which is a particular concern for the many companies using Microsoft Office SharePoint. The basic architecture of the hub-and-spoke model raises further concerns as well because users search and access content through the master content integration platform, it can quickly become a performance bottleneck, particularly given the swift pace of enterprise content growth.
Maximize Content with Mashups
The content mashup strategy addresses many of the concerns raised by hub-and-spoke. Like Web application mashups, the content mashup brings together content from a variety of sources so they can be used through a single comprehensive tool. It accomplishes this task with an service-oriented architecture (SOA) approach that takes cues from peer-to-peer (P2P) file sharing networks and Web 2.0 sites like Flickr, whose database of images and open protocols have inspired hundreds of applications.
Page 1 of 3.
