# Vignette 14: Togo's radio mining
When COVID-19 lockdowns began in Togo, UNDP could no longer collect data through conventional methods to understand community needs. "We were like in the blind," reflects Yem Ahiatsi, Head of Solutions Mapping at UNDP Togo's Accelerator Lab. The Country Office had no way of gathering sufficient data for decision-making or responding to government requests for information.
It would have been easy to use online surveys or social media as a data source for conducting a perception study. However, that approach would exclude people in isolated areas without internet access. The Accelerator Lab team therefore turned to an unconventional data source: Togo's rich radio landscape, with over 75 public and private stations serving as forums for community expression. "This channel has the advantage of eliminating response biases, where respondents frame their opinions to show themselves in a favorable light to society or the interviewer," explains the team. By mining radio conversations, they could reach a representative population while capturing more authentic perspectives.[\[1\]](#endnote-1)
Working with the Centre for Web Observation and Analysis (COAWEB), the Accelerator Lab analyzed content from 32 interactive radio broadcasts on Radio Lomé (the national station) and 12 community radio stations with strong regional audiences. As a data source, broadcasts in five languages (Ewe, French, Kabyè, Mina, and Tem) were used to ensure inclusivity across Togo's diverse linguistic communities. Each broadcast included street interviews and call-in segments where people shared experiences about COVID-19, from misconceptions about the virus to impacts on their livelihoods. Participants were informed that their contributions would be used to understand trends and advise the government.
This approach revealed deeper insights than traditional surveys. The team found that radio conversations provided more inclusive and candid perspectives than social media could have provided, with callers expressing themselves more freely and sharing more detailed opinions. Over 200 participants contributed through street interviews and call-ins, reaching an estimated audience of 5.2 million people, representing the majority of Togo's population.
Converting radio conversations into analyzable data proved challenging. "The biggest challenge was the audio transcription," Yem reflects. "It was done manually. It took a lot of time and even the analysis kind of required specialized skills and tools." The team had to dedicate significant additional resources to transcribe broadcasts, particularly those in local languages, which extended the project timeline by several weeks.
The insights gathered informed both COVID-19 response strategies and subsequent radio programs designed to combat misinformation. The team shared their findings with health authorities and NGOs involved in pandemic relief, which helped shape more effective communication and intervention strategies based on the actual concerns of the community.
{% hint style="info" %}
#### **Key takeaways:**
* **Identify inclusive data sources:** Consider how your data collection approach might exclude vulnerable populations and seek alternative channels that can reach diverse groups
* **Include multiple languages:** When working in multilingual contexts, design data collection strategies that accommodate linguistic diversity to prevent excluding important perspectives
* **Ensure informed consent:** When collecting data through public forums like radio shows, clearly communicate to participants how their contributions will be used
* **Plan for resource-intensive processing:** When using non-conventional data sources like audio recordings, allocate sufficient time and expertise for processing raw data into analyzable formats
* **Engage specialist partners early:** Collaboration with organizations possessing relevant expertise (like COAWEB for media analysis) can significantly enhance the quality of data collection and interpretation.
{% endhint %}
***
## Notes
1. For more information about this radio mining initiative see PNUD Togo (2021). [↑](#endnote-ref-1)
---
# Agent Instructions: Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter:
```
GET https://bas-leurss-personal-organization.gitbook.io/undp-accelerator-labs/doing-r-and-d/5.-r-and-d-practices/capturing-data/vignette-14-togos-radio-mining.md?ask=
```
The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.