Nokia Data Gathering (now Microsoft Data Gathering) is a survey and analysis tool for web and mobile. Many international NGOs and governments use it for monitoring and evalution. The previous interface was outdated and difficult to learn, prompting a redesign.
I researched how people used the tool through contextual inquiry and user interviews, and completely redesigned the interactions and look and feel of the server-side user interface.
A usability evaluation, and wireframes and mockups of the redesign.
The redesign was successfully launched for organisations in over 20 countries, and improved usabillity (especially for users with low literacy levels).
Microsoft Data Gathering (formerly Nokia Data Gathering) is a mobile app for collecting information from the field. NGOs and governments worldwide (including the UN FAO, World Vision, Plan International, and government health departments in Brazil) have used it to conduct surveys, monitor their programmes, and gather data on everything from health to education results. The project was a Corporate Social Responsibility initiative by Nokia.
The software allows users to create and deploy surveys, manage field workers, and view and analyse survey results through a web interface.
Without an adequate user base, the project would struggle to justify continued funding from Nokia's CSR manager. By continually improving the system, the development team was aiming to meet evolving user needs and demonstrate the tool's usefulness. Design became a concern after the tool had been in use for several years, and I was brought in to lead a redesign of the server-side interface for a major new release.
After reviewing the tool, I made some preliminary observations of potential usability problems. However, because this tool was already deployed and being used by thousands of people, feedback from actual users was critical. I traveled to Manila, Philippines for a workshop organised by Nokia, where I conducted interviews with key users.
The research backed up my earlier assumptions - creating surveys was complicated and confusing, and analysing collected data to understand what was happening on the ground was cumbersome. Additionally, many NGOs using the tool were based in developing countries and had noticed that their data collectors were making mistakes when completing surveys.
I set to work on improving the survey builder and data analysis features, starting with wireframes. I focused on simplifying the existing interface, stripping out unnecessary elements and designing easy drag-and-drop idioms to help users quickly build what they needed. These early wireframes show some of my concepts of how the survey creation tool could function.
Collecting data correctly is half the story - analysing what's been gathered adds meaning. From the interviews I learned that most users just need a simple snapshot of how their survey has been completed. More complicated analysis always took place in external tools. I decided to keep the visualisation tools extremely basic - a map view and two types of chart.
During the multi-day workshop in Manila, I created quick wireframes and got several rounds of feedback from key users and stakeholders. I decided to do this in order to save time - I could get insights from 20 users from 10 countries in 3 days, something which may have taken me weeks of travel or scheduling calls.
Once I had the interactions and flows designed and validated, I moved on to the visual design. The interface needed to be pragmatic, visually clean, and focus on interactivity (lots of rich visual modeless feedback).
One of the main challenges was designing a way for users to easily add skip logic to their survey questions (for example: “if the user chooses this answer, skip the next question”).
Often, data collectors entered information incorrectly, for example in fields that didn't need any data based on a response in a previous field. This required manual cleanup from administrators - a boring and consuming task.
If we could get the skip logic interactions right, it would add powerful functionality for survey builders by helping them collect more structured, reliable data, and save time and money for project managers.
I decided that a straightforward rules-based system for each question would be best. Since many users were based in developing countries around the world, the interactions needed to be as simple and universal as possible.
After being implemented, my redesigns significantly improved the user experience for new data managers by improving learnability and ease of use.
The tool was mentioned by the World Bank in study on mobile applications for the agriculture and forestry sector, and continues to be available open source to this day: microsoftdatagathering.net.
I also redesigned a version of the data collection app for windows tablet.
Looking back on the project, I would have conducted more usability testing with prototypes. Since the new changes were so different to what existed previously, some more validation of the major functions would have been valuable.
I would also have created a style guide to help the developers continue building the app even if I wasn’t available.