How can a farmer access the internet using only SMS?

Last month we introduced you to WeFarm – a knowledge sharing platform from the CafeDirect Producers Foundation – which allows farmers in developing countries to harness the power of the internet using only a ‘dumb’ phone.

A recently published infographic shows that while 75% of the world’s population have cell phones, just 30% have internet access. In rural parts of Africa, Asia and South America these statistics are even more extreme, underscoring the need for initiatives like WeFarm.

But how does WeFarm allow farmers in different countries, and speaking different languages, to share knowledge about farming practices using just SMS messages?

The mechanics of WeFarm

Broadly speaking the WeFarm platform consists of three main elements.

A back-end

This is the domain of Conker Group. A database stores questions and answers as well as some basic user data for farmers and translators. The database allows identification of users who might be able to answer or translate a given question. Relationships between the data can be used to identify when new questions might be solved by answers to previous questions. This is done via a series of rules in a layer of software that sits between the back-end and…

A front-end

The front end displays questions and answers to users based on their status, expertise and preferences. Questions can be asked, answered and rated here – by users with internet access – but it’s chiefly used by translators to provide translations of Qs and As. Manifesto is working on both the front-end (using the designs provided by Someone/Else) and the layer of logic that sits between the database and front-end.

A mobile-messaging platform

Telerivet provides the interface between the database and the mobile phone user. Their software parses the message sent by the farmer (to a dedicated WeFarm number) and transfers it to the WeFarm database. Here it is matched and distributed to up to 80 farmers based on their registered areas of expertise.

The journey of a WeFarm question

The diagram below describes the journey of a new question (and a corresponding new answer) through the WeFarm platform. Here’s a description of each stage:

1. A farmer – let’s call him Jacob – wants an answer to a question e.g. ‘How do we control coffee rust disease?’ He sends it via an SMS message, in English, to WeFarm. Telerivet picks up the message and conveys the question to the WeFarm database.
2. A call from the API updates the web interface, allowing web users to answer or translate the question.
3. A translator sees the question and translates it into Spanish.
4. The database stores the question in both languages and the front-end displays both versions, inviting answers and further translations.
5. Another farmer – let’s call her Maria – has provided user data that indicates she might be able to answer this question (it’s in her field). WeFarm pushes the question out to her via SMS (through the Telerivet API).

6. Success! Maria does have an answer to the question and sends it, in Spanish, back to WeFarm.
7. The front-end displays the answer to the question – but it’s in a language, Spanish, that Jacob can’t understand.
8. Another translator finds the Spanish answer in her feed and translates it into English.
9. The English answer is pushed to Jacob via SMS, helping him with his coffee rust disease problem.
10. Question and answer, in both languages, are stored in the database. They’re available to all farmers via the front-end and for translation into further languages. They can also be matched with the same, or similar, questions asked in the future, giving farmers access to knowledge that would otherwise be unavailable to them.

What’s next?

The launch of WeFarm is getting pretty close now. Stay tuned for more posts on user testing in Kenya and a more in-depth look at building a front end for this exciting project.