The literacy assumption embedded in every text-only platform
Most digital platforms assume that the user can read fluently in the language of the interface. For consumer applications and urban professional tools, this assumption is reasonable. For agricultural advisory platforms in rural Africa, it is the mechanism by which the platform excludes the producers it exists to serve.
The exclusion compounds. A producer who is not comfortable reading in French cannot use a French-language interface. A producer who is uncomfortable reading at all cannot use any text-based interface, regardless of language. National extension programmes are designed to reach exactly these producers, and text-only platforms quietly fail them while looking successful on the analytics dashboard.
Audio as the primary modality
Audio-first design treats the spoken word as the primary modality for content delivery, with text and visuals as supporting layers. The discipline is more demanding than it first appears. Audio content has to be authored, recorded in studio quality, edited for clarity, organised in a structure the user can navigate without reading, and delivered in a way that works on the connectivity available to the target audience.
Each of these requires real institutional work. Local language audio requires native speakers, agronomic accuracy review, and editorial governance. Audio navigation requires careful information architecture so users can find content by topic, crop, or season without depending on written menus. Audio delivery requires technical engineering for adaptive streaming, so playback degrades gracefully on weak connections rather than failing outright.
What we delivered for ICAT and FAO Togo
PANEOTECH integrated multilingual audio content as a first-class feature of the e-Agriconseil+ platform delivered to the Institut de Conseil et d'Appui Technique under the Pro-SADI programme commissioned by FAO Togo. Studio-quality audio versions of advisory content are available in Ewe, Kabye, Moba, and Tem, the four major working languages of rural Togo, integrated directly into the mobile application alongside the technical sheets.
Adaptive streaming adjusts audio quality to the bandwidth available on the user device, keeping playback usable on 2G connections and recovering smoothly on intermittent links. The content management back-office includes the editorial controls ICAT administrators need to manage audio metadata, including language tagging, so the institution can extend the audio catalogue independently as new content is produced.
The institutional lesson
Audio is not an accessibility upgrade bolted onto a text-first platform. For rural advisory programmes, it is the primary modality through which the platform reaches the population it was funded to serve. Treat it as structural, design the institutional content production capacity around it, and the platform earns the kind of adoption that text-only systems cannot.
