Login to view full entry:
Agro360: Voice Messages in Bangla versus Local Dialects
BAN -19 -1545Last modified on December 19th, 2025 at 10:39 am
-
Abstract
mPower is a Bangladesh-based social enterprise that specializes in technology-based development solutions. As part of its agriculture program, Agro360, mPower sends text messages with crop management recommendations and weather alerts to farmers to improve the efficiency of their cultivation of their crops and to promote practices that mitigate the impact of climate change on their crops. For Rabi 2019, mPower partnered with PxD to understand the effect of sending voice messages in the local dialect on farmers’ engagement with the recommendations.
Even though Bangla is the official language of Bangladesh, farmers in different parts of the country speak local dialects. According to the Cultural Survey of Bangladesh, there are 16 regional varieties of Bangla spoken in the target districts of the program. Even though farmers understand Bangla, they may feel more comfortable with messages in their local dialect.
We tested farmers’ engagement with the service when the voice messages were delivered in their local dialect, compared to messages delivered in traditional Bangla as they had been during the previous seasons. The results show that farmers engage more with the voice message when it is sent in their local dialect, as reflected in the listening rate and the completion rate. -
Status
Completed
-
Start date
Q1 Jan 2019
-
Experiment Location
Bangladesh
-
Partner Organization
mPower
-
Agricultural season
Rabi
-
Experiment type
A/B test
-
Sample frame / target population
Mungbean and watermelon farmers registered in Agro360 and assigned to receive voice recommendations
-
Sample size
4,518
-
Outcome type
Service engagement
-
Mode of data collection
PxD administrative data
-
Research question(s)
Do voice messages in local dialects increase farmers’ engagement with the service?
-
Research theme
Agricultural management advice, Message narration
-
Research Design
The study randomized 4,518 farmers into two groups to evaluate the effect of language on the farmers’ engagement with voice message advisory services. The treatment group (n = 2,360) received messages in the local dialect, while the control group (n = 2,158) received messages in standard Bangla. Randomization was stratified by farmer union, crop type (mungbean or watermelon), and planting date, to ensure balance in key characteristics across treatment arms.
Because watermelon farmers joined the service later, they received fewer messages on average than the mungbean farmers.
-
Results
Farmers engaged more with the voice message when it was sent in their local dialect. Voice messages in the local dialect had a 3.8 percentage point (pp) increase in the listening rate, defined as the percentage of the call a farmer listened to, over the control mean of 73.2%. Voice messages in the local dialect had a 7.0 pp increase in the call completion rate, defined as the likelihood of listening to the full voice message, over the control mean of 48.5%. There is no notable statistical relationship between the language of messages and pick-up rates.