When I was at the MERL Tech (Monitoring, Evaluation, Research, and Learning) conference recently, I heard something that resonated with me and my work at GeoPoll: using mobile technology - whether it be for financial transactions, reaching communities, or collecting and sharing key information - is no longer a subject of the future. It is happening right now. At GeoPoll, my core focus is helping our international development partners navigate this new arena effectively. I work with donor-funded organizations to help them better understand and use SMS and IVR as a means to collect data so that they can become more knowledgeable, more efficient, and more cost-effective in their work. Many days, I act as a consultant on mobile surveys, guiding our partners on what works, what does not work, and how to approach using mobile surveys in developing countries. Along this path, some themes have emerged, and so I am sharing the “Top 5 Things to Think About” when considering incorporating mobile data collection into development programs:
- The Objective. With any research study, the most important first step is identifying the primary purpose. Clearly laying out the objectives and the information needed at the end of the study helps narrow in the focus. Keep it simple and targeted to start and build then from there.
- The Audience. After identifying the “what” for your study, the next major thing to think about is who you are trying to reach. At GeoPoll, we have an existing database of more than 200 million users indexed by key demographics such as age, gender, and location. These mobile subscribers often serve as the sample for the research we conduct. However, some clients are looking to reach a more targeted group of people, such as project beneficiaries. In these cases, clients can provide GeoPoll with the mobile numbers of a select group, provided they have permission to contact those individuals. Sometimes our partners find collecting the mobile numbers of beneficiaries or other groups is easy; other times, it is a time consuming process that needs to be accounted for when determining the timeline for launching mobile surveying activities.
- The Frequency. The appropriate frequency for engaging the target audience in mobile surveys varies. Sometimes quick snapshots or needs assessments are all that is needed. Other times, on-going monitoring over several months better serves the overall objective. GeoPoll conducts thousands of mobile surveys each day, and we have learned that frequency matters. As far as respondents go, there is usually a learning curve at the onset of a data collection project – the more you engage in mobile surveying, the better respondents understand the process, expectations, and outcomes of the engagement. At the same time, there is a need to be mindful of the respondent experience and cognizant of potential user fatigue. We find that, with the right balance, conducting surveys on a daily, weekly, or monthly basis allows for more granular data which can be disaggregated to see specific points in times as changes occur. This long-term tracking has not traditionally been done in development projects and can add immense value to your research.
- The Limitations. The goal of mobile surveying is not necessarily to replace face-to-face surveys, and mobile surveying and traditional surveying methodologies can often work together to overcome limitations of each method. For example, GeoPoll recommends a mobile survey consist of 15-20 question per module. This means that keeping questions targeted and simple is critical, particularly since you will be reaching respondents of varying backgrounds. Despite this, complex survey instruments can be adapted to the mobile phone. We have worked with multiple partners on surveys with 80+ questions, breaking the survey instrument into modules and sending one module at a time. Whether you choose this method, or combine mobile surveys with face-to-face research, mobile is an incredibly useful tool to complement other forms of data collection, and additionally can serve as a valuable mechanism for data triangulation.
- The Data. Last but not least, how the data will be used is perhaps the most important of all. This can be as simple as what format the data will be presented in – GeoPoll can provide raw datasets in several formats, or build an online dashboard which visualizes data and can be shared with donors and local project staff. Beyond the nuts and bolts of transmitting the data, thinking about how it will fit into the broader decision-making process is critical when launching mobile data collection activities. For example, is there already an existing process in place where decisions are made based on the data? Is there buy-in needed from government counterparts or local partners as data is collected? Regardless of the intricacies of your project, timing the data collection to coincide with reporting timelines and/or decision-making processes will ensure the data is used in a meaningful way, whether it be helping projects make course corrections, or conducting an evaluation after a project has concluded.
Interested in reading more about GeoPoll's work with development organizations? View our case studies page, or click below to contact us.