ICTworks

CrowdGlobe has analyzed both Ushahidi and Crowdmap data as well as its user base. The Ushahidi platform, which means, “witness” in Swahili, is a free and open source tool that integrates information collection features with a live map.

Ushahidi, the company, subsequently launched Crowdmap, a hosted version of the Ushahidi platform, which is easier to use since downloading the software and installing it is not necessary.

When the CrowdGlobe research project was launched in October 2011, a total of 12,795 Crowdmaps had been created in over 100 countries. This presented CrowdGlobe researchers with an ideal first use-case for the project. The aim of this first report is to develop a better understanding of how Crowdmap (and Ushahidi) have been used and to analyze the data they have generated over recent years.

Mapping the Maps took advantage of statistical analysis, quantitative content analysis and exploratory surveys. The quantitative analysis revealed that 93% of the 12,000+ Crowdmaps analyzed had fewer than 10 reports while 61% of Crowdmaps were identical to the default Crowdmap setting, i.e., they had not been customized or used at all.

This “long tail” distribution of Crowdmaps follows a power law distribution, a common feature in many online platforms, as well as in a number of occurring phenomena. Crowdmaps with 21 to 10,000 reports were selected for further analysis, resulting in a data set of 585 maps. About 30% of these focused on North America while 18% focused on Western Europe and 16% on Africa.

On average, these Crowdmaps had 814 reports but the median number of reports for this set of deployments was substantially lower, which is not surprising considering that Crowdmaps follow a power law distribution.

CrowdMap Distribution

About 30% of these 585 Crowdmaps focused on North America while 18% focused on Western Europe and 16% on Africa. When the analysis is broken down by region, the relative frequency with which these themes emerged in the regional deployments differs dramatically.

For example, taking into account recent events, it is not surprising that the most common themes that emerged from the 63 deployments in the Middle East and Northern Africa pertain to: Crime and public safety issues (43%), human rights abuses (40%), emergency-related infrastructural issues (30%), and political organization (25%). The distribution of themes in the 79 Western European deployments, on the other hand, paints a very different picture, with entertainment and leisure appearing in 32% of the deployments, followed by non-emergency infrastructural issues (25%), and media reports (23%).

It is also not surprising that the 16 deployments from the Caribbean region, 12 of which hailed from Haiti specifically, heavily featured issues related to the occurrence and aftermath of a natural disaster (63% and 50%, respectively), emergency-related infrastructural issues (63%), health and medical-related issues (50%), and crime and public safety issues (38%).

CrowdMap Usage

The quantitative analysis also revealed that 93% of the 12,000+ Crowdmaps analyzed had fewer than 10 reports while 61% of Crowdmaps were identical to the default Crowdmap setting, i.e., they had not been customized or used.

This “long tail” distribution of Crowdmaps follows a power law distribution, a very common feature in many online platforms such as Wikipedia. To this end, Crowdmaps that had between 21 to 10,000 reports were selected (yielding 585 “actionable” Crowmaps) for the more in-depth analysis carried out above.

On average, these Crowdmaps had 814 reports each but the median number of reports for this set of deployments was substantially lower, which again is not surprising considering that Crowdmaps follow a power law distribution.

CrowdMap Users

In addition, surveys were sent to all 12,795 Crowdmap users to better understand how they used the platform and to assess their experience. A total of 276 users responded to the survey (~2% response rate). ).

While this is only a convenience sample, the results are nonetheless informative and relevant. About 80% of respondents are men and the average age of a user is 40 years old. As for educational background, an impressive 43% of users have a post-graduate degree and a total of 84% have at least a college degree. Approximately 53% of users responded as having no prior experience in using online mapping technologies, which is not surprising since the technology is still relatively new.

The most important reason why Crowdmap users used the Crowdmap platform was to create a map with aspecific purpose or event (40%). About 63% of these users launched a map to cover an event in the city in which they live.

CrowdMap Testing

About 50% of Crowdmap users created a map for non-direct use, i.e., for training or demonstration purposes. Furthermore, about 14% of users noted that they were not intending to create a map in the first pace but were simply curious.

In other words, about to 64% of Crowdmap users who responded to the survey were not launching a map for operational, project-focused purposes, which in large part explains the power law distribution reported above. That said, around 24% of users said that Crowdmap was either too confusing or too technically challenging for them to work with or that they simply did not understand how to aggregate data from other sources for their map.

In addition, approximately 6% of users revealed that they “gave up” trying to create a map when they realized it would take too much time to set up or maintain. In sum, about 30% of 276 users who responded to the survey found the technology counter-intuitive and too time consuming.

CrowdMap Promotion

In terms of users who did launch a Crowdmap, about 16% felt that they were not able to generate the required public awareness vis-a-vis their map to make it as effective as they had hoped.

Around 19% noted they were able to make their map almost as effective as they hoped while 31% replied that they simply were not able to garner sufficient interest in their map. This finding also explains the power law distribution described above.

Of those users who responded to the survey and felt they had been successful in raising awareness of their maps, about 23% noted that they had done so via engagement with community organizations, civic groups and/or professional organizations.

Around 20% said they had successfully built public awareness by engaging with social media, SMS, etc. Only 6% of users said they used traditional news media to build awareness of their project.

Conclusion

The findings from the quantitative analysis and surveys provide the first evidence-based analysis of crowdsourced data of its kind. In addition, the results supply actionable feedback to Ushahidi software developers on what they can do to improve their platforms and substantially increase the number of Crowdmaps that gain more traction and possibly greater impact.

As it turns out, Ushahidi Inc. has already been implementing a number of important changes since this research over half-a-year ago. For example, the Crowdmap platform now includes a set-up wizard that guides first-time users through the customization of their platform. In addition, Ushahidi has recruited a full time Community Manager who has held more user-group meet-up’s in 2012 than in all four previous years combined.

Furthermore, the company recently launched an informative wiki for Ushahidi users, which provides more documentation and supporting material on best practices and lessons learned. Moreover, Ushahidi has also launched a novel partnership with TechChange to offer a unique and dedicated, hands-on course on how to use Ushahidi & Crowdmap and create successful deployments. Much of the content developed for this course will be shared on the Ushahidi wiki.

Finally, the team has carried out a full usability review of the Crowdmap platform with significant changes in store for the interface to render the tool more user-friendly and intuitive.

These important improvements, taken together, are bound to generate Ushahidi/Crowdmaps that gain both more traction and visibility. In sum, this report provides an important baseline study—and indeed the only one of it’s kind—which could serve as an important comparison if this research is replicated in the coming years.