The starting point for the session was the number of large and diverse financial inclusion data investments made over the last decade by several prominent donors. These data investments have largely been driven by donors’ needs to inform programming and to measure the effectiveness of their interventions. Many of these data investments have played a critical role in informing local decision-makers in the design of more effective policies and services, especially in data-poor countries. The session on funding systemic data solutions focused on how these data investments can be made more sustainable. Sustainable initiatives are those that would continue to operate or continue to have an impact after the funder or implementers withdraw. In addition to sustainability, greater collaboration between funders and other data users, as well as making data more accessible and relevant to users were discussed.
What to invest in?
Two approaches to identifying data gaps were highlighted:
- The first approach requires a good understanding of how achieving impact within a given market or for a given partner works. Further it needs a process of validated learning, whereby you start out small to test and learn, and drop aspects over time which are found not be useful.
- The second approach combines country-specific research with local stakeholder engagement to highlight the country-specific data needs and requirements.
Is it worth it?
It is easy to assume that data investments will be used given the number of uses there are for data. However, it is difficult to estimate whether the potential uses of the data justify the cost. One approach identifies an estimate the cost of the top three uses and decide if these are worth the investment. Another approach is to identify where costs can be decreased. For example, costs can be decreased and investments made more effective if complementarities are found with existing initiatives and where there are opportunities to leverage existing infrastructure (either in collecting, housing or disseminating). Investing in capacity-building can also lead to cost decreases in the future when new capacities can be leveraged.
How to make it last?
The session highlighted the importance of thinking about data as a product. Often, the focus of data initiatives lies more with data authenticity than with the tailoring of the user experience. Thinking about data as a product involves integrating traditional business concepts that are focused on the user experience into data initiatives in the development space, where these concepts have not been largely applied. Examples include prototyping to verify demand at the launching phase, using marketing at the dissemination stage, or doing in-depth usage analysis that goes beyond tracking basic website traffic. The discussion also highlighted important factors to take into consideration when thinking about sustainability and impact:
- Data quality. It is important that respondents understand the survey questions and context. Local language surveys help, but the translation process can often undermine the value of a local language survey. It is therefore important to check that the correct meaning and context have been retained during translation. Updating or revising data collection templates can ensure that consecutive rounds of data collection remain accurate and relevant.
- Dissemination. Data have to be made available to potential users in a timely manner in order to achieve impact. The format of dissemination is also important (for instance, infographics can be more effective than reports).
- Focusing on the user experience. The importance of simplicity, visual engagement and guidance were highlighted. In other words, it is important to think through the user experience. If your audience is not informed, the interface should nudge or point them to the right information.
- User understanding – Highly public data could be sensitive to publish especially if collected data does not correspond with comparable publicly available figures (even if discrepancies are only due to differences in methodologies). For instance, the difference between household versus individual-level surveys. Making complimentary data and research available alongside the data can help user understanding.
- Local needs. Incorporating the data needs of local (and other) stakeholders in the design of your data initiative can create more buy-in and make data more relevant to local users. This process can involve data scans, which local stakeholders can react to, or it could set up a local steering committee, which could create buy-in and ensure that local inputs are integrated into the data collection process. Motivating or mobilising local actors are also a key objective of certain data investments.
- Comparability. The definitions of variables need to be transparent for data to be comparable across countries. It is important to be flexible and adapt questions in consecutive rounds if it becomes clear that variables are not comparable if, for instance, there is a misinterpretation of the local context.
Trade-offs in investment decisions.
Taking local requirements into consideration were highlighted as an important component of sustainability in data initiatives. However, this has to be balanced with cross-country comparability and donor monitoring and programming requirements. For example, the standardisation of questionnaires allows for scale but reduces flexibility to adapt to local requirements. Further trade-offs may be required if data for donor reporting purposes are different to those required to inform and change the behaviour of local financial inclusion stakeholders. Available resources, questionnaire space and the length and time of a questionnaire also have to be taken into consideration. As a result of cost considerations, donors may also have to choose between a broad set of countries with a narrow set of questions, or a narrow set of countries with a broad set of in-depth questions.
How to deal with trade-offs.
Communicating clearly upfront with stakeholders about the limitations of questionnaire space and the need to balance trade-offs can help reduce the requests for unique local questions to be inserted in the questionnaire. Further, to balance cross-country comparability with local needs, a standard baseline questionnaire can be used to ensure comparability. Modular variations can then be adapted based on local use cases or specific needs. The use of Computer Assisted Personal Interviewing (CAPI), as opposed to pen-and-paper interviewing (PAPI), can help to maximise questionnaire space. Removing unused questions identified through user feedback also helps.
How to disseminate data.
In planning for the session, participants also highlighted some of the challenges of hosting and disseminating data. For example, requesting an institution whose main competencies are to collect data to also host and disseminate this very same data can be problematic. Timeliness of dissemination was also highlighted as an issue, as data usually have a limited shelf life for usefulness. Making the methodology and the information on the quality checking process available alongside the data is important for users to understand the data and to allow local stakeholders to replicate surveys locally. The value of engaging in post dissemination activities with local stakeholders was also highlighted in terms of enhancing the impact of data investments (for instance, providing financial education if the data highlighted this as a local need).
Going forward i2i will continue to engage with CGAP’s Measuring Market Development donor working group. Members of the working group will also be invited to join the i2i community of practice (CoP) to input into the measurement and data quality work streams of the facility.