Every day, over a billion Africans engage in buying or selling goods and services within the Micro, Small, and Medium-sized Enterprise (MSME) sector.
While precise data on a day's trade volumes is scarce, the African Trade Report 2022 estimated the value of the continent's goods trade with the rest of the world at $1,180.45 billion in 2021 - an impressive increase of 31.4 percent from 2020. Importantly, 80 percent of this trade occurs within the MSME sector.
A distinctive feature of Africa's MSME sector is its unique linguistic landscape. Unlike other regions where European and Asian languages dominate, African languages form the backbone of communication in these enterprises.
As digital trade and e-commerce continue to drive global markets, it is clear that incorporating African languages, idiomatic expressions, and cultural nuances into digital tools for e-commerce could boost intra-African business and cut across food stalls, retail markets and jua kali (informal) manufacturing hubs.
By embracing these elements, we can unlock economic opportunities, enhance competitiveness, and celebrate the cultural richness of the continent.
Africa’s youth, who constitute more than 70 percent of Sub-Saharan Africa’s population and largely communicate in these local languages, represent a key demographic for digital product development.
Ignoring this group in creation of language technologies and digital tools exacerbates exclusion from e-commerce, widens the digital and wealth divide between the Global North and South, and contravenes the United Nations' commitment to ensuring that sustainable development leaves no one behind.
Equally critical in Africa's informal trade, is the role of women who account for over 70 percent of transactions. These women conduct business primarily in local languages.
For instance, a trader selling chilli oil in Kigali could potentially collaborate with suppliers in Tanzania to meet market demand in Chad or South Sudan. However, such cross-border commerce requires language technologies that accurately facilitate communication and transactions.
Developing these tools starts with recognising the importance of local languages in accessing public services and earning livelihoods.
These languages should provide the foundation for e-commerce solutions.
As the majority of commerce on the continent is anchored around the MSME sector and six out of 10 people in Sub-Saharan Africa live in rural areas, the digital revolution must scale with considerations for this significant population and their natural languages of communication and transaction.
An MSME keen to engage in a continental market or scale digital applications and services for a larger market must recognise the power of spoken language, the data generated daily on the continent, and its potential to bridge access to services and trade.
Looking at the digital technologies powering e-commerce in the 21st century, one might be led to assume that a handful of global languages suffice for trade and communication.
However, current data reveals that the most widely digitised languages do not represent the majority of the world's population, for instance, only about 1,000 out of 7,000 languages are digested as reported by Translation Commons Language Digitisation Initiative.
Languages that lack digital representation are sidelined from advancing their use and richness.
Without digital availability, there is limited access to resources such as dictionaries, grammar models or advanced technologies such as machine learning and speech recognition systems. Such languages are further underrepresented in research and development, with insufficient data to train conversational AI systems.
The Africa Next Voices Project is one initiative addressing this gap leveraging public-private sector collaborations.
A consortium of academia in Kenya and tech-innovation based research institutions has come together to build text and voice data sets, ensuring a foundation for dictionaries, training of language models such as machine learning and speech recognition, and conversational AI systems.
This project underscores the need for collaboration in transforming the daily data generated as speech, text, or multi-modal formats into digitised data sets that can leapfrog the continent’s digital capacity.
Globally, more than three billion people speak languages considered "low-resource" or minimally digitised, accounting for approximately 68 percent of the global population.
For instance, English, spoken by about 430 million people worldwide as either a first or second language, represents less than 10 percent of the global population, while other European languages combined account for only 23 percent.
Yet, these languages continue to shape the digital landscape. English contributes 67 percent to language model development, and other European languages contribute 24 per cent, leaving the three billion speakers of low-resource languages with only a six percent share in this digital realm.
This disparity highlights the urgent need to incorporate African languages into AI and other digital tools driving global e-commerce.
MSMEs and tech entrepreneurs, who generate this kind of data by design (‘operational design’), must prioritise collaborations to harness the power of language technologies.
Failure to do so risks marginalising younger and larger populations globally, particularly in Africa, from the benefits of the digital economy.
The writer is a digital commerce advisor at Tony Blair Institute of Global Change
