Friday, November 10, 2006

Surveying web portals and their usage

In the first half of 2006, Euforic carried out a study for CTA on its four ‘portals’ – Agritrade, ICT Update, Knowledge for Development, and Spore. The aim was to get a better picture of the niche and position of each portal with a view to improving their effectiveness.

Agritrade is a web-based news and alerting service for specialists in ACP-EU agricultural trade negotiations – it also has email and print outputs. ICT Update is a web magazine for people working with information, communication and ICTs in agriculture and rural development - it is also distributed in print, email and sometimes CD-ROM formats. Knowledge for Development is a web-based resource for policymakers in science and research – it is integrated in a wider programme of meetings and other activities. Spore is a print magazine for people working in ACP agriculture and rural development - whose web ‘portal’ gives access to the full text content and archives of the print version. Its web site is more a web interface/archive of the print magazine than a portal.

Some points emerging from study were:

People responding to the survey were generally positive about the usefulness and value of the portals. According to the user surveys, the most popular features of the current portals are the email newsletter, full text documents/publications, links to information resources, and feature articles. Generally ‘staying informed’ was the most cited reason to use the sites.

The online surveys attracted a high proportion of respondents with African, Caribbean or Pacific nationality. This indicates a healthy ACP ‘reach’ in terms of the email products. Web usage statistics suggest a much smaller ACP audience – though this is difficult to measure. Other audiences are also being reached. The Agritrade site, for instance, is regularly visited by people from key ‘target’ institutions in Europe – such as the European Commission and Parliament, and the ACP Secretariat. Spore has a significant non-English language audience.

We also looked at some ‘behind the scenes’ challenges:

Capturing and making sense of web site usage statistics is always difficult. Massive ‘hits’ do not always translate into many visitors. A particular issue was to identify and understand usage from the ACP countries that CTA is mandated to support. As the portals move towards tools like RSS, this usage also needs to be tracked.

Another challenge is to look at ways to organize and structure content in the various sites so it can easily be shared and re-used and can be accessed across the entire CTA web ‘family’ - and indeed with external partners. A related issue was to make more of the ‘hidden’ content in the portal archives visible and accessible.

It was also important to optimize the technical performance of each portal – loading speed for example; as well as positioning in web search engines that will increase site rankings and drive visitors to the sites. A particular issue is to ensure that portal content is also well-positioned in ACP information and communication channels that may, for example, be off-web or more face to face or paper based (and not just on Google et al).

Some particular learning points of this study are:

1. Although we were looking at ‘web portals’, it is clear that the interactions among different tools and media and especially the mix of complementary information and communication channels is an important factor in usage. For the CTA portals, the associated email, paper, and face to face products seem to strongly reinforce and extend the use of web products.

2. Getting high visibility on Google and related search engines and ranking tools is ‘accepted wisdom’ in Europe and North America and the private sector. In the development sector, the target communities may not be ‘on’ Google and it is therefore as important to be very ‘visible’ in the many online and offline networks where these people meet and exchange and find information. It may be more appropriate to optimize the sites to reach smaller niche ‘markets’ in the so called ‘long tail’ than to compete for global audiences. This means thinking beyond the web and looking for different indicators and ways to ‘rank’ visibility and ‘clickthroughs.’ A high ‘ranking’ in an ACP-EU trade expert meeting may be more valuable than a high ranking on Google – though it is perhaps easier to focus on a single Google or Alexa ranking than on dozens of individual ‘niche’ rankings.

3. Users from most ACP and other developing countries hardly show up in web statistics. Partly because there is a smaller Internet population and digital divide issues; but also because much ACP usage is ‘hidden’ in usage that comes via ‘freemail’ or ‘freeweb’ services such as hotmail or yahoo. Some 70% of email subscribers to the CTA email alerts were from such addresses. It is reasonable to conclude that much web usage and many users from ‘.com’ domains are likely to be individuals from developing countries using such ‘free’ services. The ‘usual’ web statistics are likely to under-rate usage from these countries.

More information on the CTA portals at:,,, and from the Euforic dossier on information, knowledge and communication.