Development Gateway has changed significantly in response to external challenges and the emergence of new web-based technologies. Their experience highlights some of the lessons learned in the process that may be relevant for web 2.0 initiatives.
With more than 260,000 registered users and 800,000 unique visitors per month in mid-2007, the Development Gateway is the largest searchable repository of development information and tools on the internet.
The concept of the Development Gateway was first discussed at the World Bank in 1999. Under James Wolfensohn’s leadership, the Bank positioned itself as the ‘Knowledge Bank’. It would promote knowledge sharing and put growing emphasis on the use of ICTs as tools for disseminating information both within the organization and across the development community.
The Development Gateway Foundation was established in 2001, funded by the Bank, bilateral aid agencies, developing country governments and private companies. The Foundation then set about creating a collaborative web portal offering development content and tools for information exchange. The approach to knowledge sharing focused on capturing and making available relevant information through the ‘topic pages’ (now dgCommunities), which also provided discussion and commenting tools to enable sharing and collaboration among its members. The idea was that experts and users would contribute resources that were then to become the basis for the sharing of explicit knowledge.
The Gateway selected experts to act as ‘topic guides’, contributing knowledge resources, monitoring and ‘approving’ user-contributed resources and facilitating discussions. The issue of quality control was addressed through a ‘guided community’ model based on deferred publishing – content submitted by users went live only after it was reviewed and approved by the topic guides. Initially these guides were paid consultants, although they were later replaced by volunteers, in line with industry trends.
Country gateways, another core initiative, were locally owned projects designed to provide online and offline technical services for e-government, small enterprise support, e-learning, e-health, and online community building. They were seen not only as local versions of the global portal, but also as nodes of the broader network with the Gateway at its core. They were expected to adapt to local conditions, languages and needs, while establishing systems to access, maintain and disseminate local knowledge.
Since its inception, many have argued that the Gateway’s ambitious objectives were difficult to operationalize, and that it tried to cover too much ground. Gradually, the Gateway strategy moved away from ‘knowledge sharing’ towards ‘web-based tools to make aid and development efforts more effective’. This new focus was supported by two of the Gateway’s original applications. The Accessible Information on Development Activities (AiDA) was set up as an online database of development projects and activities provided by over 200 agencies.
More recently, the Gateway, in collaboration with the government of Ethiopia, launched the Aid Management Platform (AMP) to assist governments and donors in planning, monitoring and reporting on international aid flows and activities. Transparency of public sector transactions, in turn, was promoted by dgMarket, an online service that posts tenders for government contracts funded by the Bank and other agencies, as well as national tenders.
Vision versus implementation
Like most pilot projects, the Gateway generated criticism. Staff were concerned about the overlap with the Bank’s own website and tools, while civil society organizations argued that the close link between the Gateway and the Bank would result in a Bank-centred vision of development at the expense of southern knowledge. There were also concerns that the Gateway would create unfair competition for other aggregators of development knowledge. As the Gateway ended its pilot phase and the Foundation became legally independent, however, these issues became less prominent.
A more generic criticism centred on the evolution of the dgCommunities and the country gateways. While the dgCommunities attracted around 50,000 registered members by mid-2007, they were primarily passive users, contributing or accessing and viewing existing resources. They did not utilize the collaborative or interactive capabilities of the topic pages to pose questions and/or obtain answers on specific issues. Rather than a dynamic space for sharing knowledge and solving problems, the dgCommunities became a useful, but static repository of development knowledge. And while the country gateways were successful in providing local content, the issue of the quality and depth of their information assets remained a problem. Also, integration of the local and global knowledge sources, an original objective of the programme, failed to occur. Data sharing mechanisms enabling a two-way flow of information between the Gateway and the country gateways were never fully implemented.
The financial model of the network also remained open to criticism. While the Gateway provided initial funding to local groups (directly to Country Gateways and indirectly to associated research and training centres), it was expected that the local Country Gateways would become financially independent, but few have done so. The Gateway no longer provides overall financial support to local Country Gateways; rather, selective support for local deployment of Gateway’s core products, such as local project databases, is occurring. Moreover, the integration of the local research and training centres established as part of in-kind contributions from partners (in South Korea, India, China and Rwanda) into the network remained limited at best.
From the beginning, the Gateway was able to offer to its partners (e.g. the Country Gateways) value-added services, such as technologies (the original technology strategy relied on open-source solutions) and standards for automating content exchange and syndication. Recently, the Gateway portal started to move towards using data syndication and aggregation features such as RSS feeds to enable users to keep track of the latest additions. But the question remains as to whether this has come too late, and the Gateway missed an opportunity to establish a niche in the area of information collaboration technologies.
Nevertheless, the number of unique visitors has steadily increased, reaching 800,000 per month by mid-2007. Similarly, within the dgCommunities, the number of registered users increased from 10,000 in 2003 to 50,000 by mid-2007. dgMarket subscribers increased from 7000 (2003) to 34,000 in 2007, and annual sales revenue increased fivefold in just four years, reaching $625,000 in 2007.
Looking back, looking forward
The Gateway has changed significantly in its pursuit of better targeted business objectives and adapting to external challenges. Many lessons have been learned as the project has evolved. The Gateway was influenced not only by the enthusiasm and spirit of innovation of the Internet era, but also by the ambitions and overreach typical of projects initiated in the midst of the ‘dot-com bubble’. But its ability to remain flexible and to respond to new developments (e.g. social networking, blogging) has been constrained by the complexity of its governance and institutional design – involving bilateral and multilateral agencies, governments, private sector and NGOs. Still, as a recent user survey indicated, the Gateway receives high marks for its utility as a tool for improving ‘development knowledge’ and networking. Looking back, perhaps more effort in organizing content, automating content upkeep (e.g. checking for broken links, web crawlers, etc.) and developing question-answering algorithms would have helped further energize its impact.
Looking forward, the issue of long-term financial sustainability will remain a significant challenge. With the exception of dgMarket, all the Gateway’s products and operations are dependent on donations. It is ironic that the separation from the World Bank – an important step in establishing the credibility of the initiative as a ‘neutral’ space for the debate on development – has hindered its ability to leverage funds from public and private donors. Accordingly, the initiative has been operating under a hybrid model that combines the original mission of promoting the dissemination of development knowledge and ICT tools as public goods, and a revenue-generating initiative (dgMarket). The cultures and requirements of these different components of the Gateway are not always easy to coordinate.
Eight years after its conception, however, the ‘dream’ of Jim Wolfensohn continues to evolve and to have an impact.
Denisa Popescu is an information analyst, and Carlos A. Primo Braga is senior adviser at the World Bank. Both were involved in the pilot phase of the Development Gateway portal. Comments by J. Daly, J. Garrison, and M. Fleeton are gratefully acknowledged. The views expressed in this paper are entirely those of the authors, and do not necessarily reflect those of the World Bank, its Executive Directors, or the countries they represent.
Development Gateway Foundation (2007) dgCommunities Member Survey 2006 Results.
A. Jha et al. (2004) . Bretton Woods Project.
World Bank (2000) Knowledge for All: A Strategy for Global Partnership. Washington, DC: World Bank.