All projects

The Digital Agency for International Development

PreventionWeb - Evaluation

An evaluation of UNISDR's PreventionWeb platform for the disaster risk reduction community

Nov 2011 to Apr 2012



The UNISDR PreventionWeb platform provides information and resources to disaster risk reduction professionals and communities looking to improve their resilience against natural and human-made disasters.  

Working with M&E specialists, ITAD, we conducted a wide-ranging evaluation of PreventionWeb (PW). We sought to answer the following questions:

  • Is PW reaching the right people?
  • Is PW meeting their needs? 
  • Is PW providing the optimum user experience? 
  • What impact is PW having? 
  • How effective and efficient is the management of PW?

The main recommendation of the evaluation report was that PreventionWeb should look to become a knowledge broker, targeting particular stakeholder groups with learning products  rather than just being a repository of information. This would involve a site redesign, moving from a library model to persona or action-centred navigation, which seeks to tailor pages, pathways and/or packages of content to particular user groups.



We interviewed many stakeholders in Africa, Asia, South America and Europe. These included town mayors with responsibility for disaster risk reduction, NGO leaders in the sector, people working within the UN system and academic specialists. All of these people used PW as an information source and most felt that PW (and UNISDR) had played a significant role in defining the disaster risk reduction domain and providing platforms for knowledge exchange.



Aptivate's primary focus was evaluating the user experience and establishing who was accessing PW using what platforms.

Our data collection tools included a web survey, interviews and focus  groups, web stats analysis, task‐based usability  tests, expert reviews and automated validation tools. Wherever possible we triangulated data from quantitative and  qualitative sources, providing a strong evidence base for our findings and recommendations.



We used the web-based Survey Monkey tool to send out questionnaires and we went deep into Google Analytics to slice and dice PW's usage data into different segments (e.g. OECD and non-OECD users; visitors by continent; mobile users who spent less than five seconds on the site). 

We also used Skype Call Recorder on Ubuntu to (you guessed it!) record VOIP interviews - we did of course ask the permission of the interviewees. We found this tool simple and non-intrusive - it is always clear when you are recording and the resulting sound files are well organised with each caller panned to opposite ends of the stereo spectrum.