‘How Africa Tweets’ report is a good benchmark for future analysis
The other day, Portland Communications released a new report on how Africa tweets that improves upon a similar analysis from 2011. Two years ago, the goal of the study was to establish a “benchmark for measuring the evolution of Twitter use in Africa.” This time around, researchers examined Twitter trends at a more granular level than before (ie. by city instead of country). They also looked at the individual messages to gauge language and subject matter.
The information quickly made the rounds online, with many people focusing on which cities tweet the most, which languages are most popular, and what users discuss most often. This data is all very interesting, but the majority of information should come as no surprise. But it is indeed a great benchmark and the infographic is stellar.
Casual observers would be wise to consider that many observations are more useful when compared with past and/or future data. Alone, the numbers are interesting but less actionable. For one, it makes sense that larger cities have more Twitter users. It follows that large South African cities, often with high levels of internet access, have high levels of social media activity. The same goes for Nairobi within the East Africa region. Looking at activity by time of day is almost a moot point since global internet usage trends always show a peak in the evening.
Moreover, the analysis is inherently limited to only geo-located tweets from the twenty most populous African cities. With that in mind, there is almost no point in focusing on absolute numbers of geo-located tweets since the ratio of them to all tweets is an estimate. Comparing the number between cities is more useful, but again, who knows what share within a city is geo-located. Furthermore, not all Africans live in this cohort of 20 cities so the trends are not true for all countries and do not represent secondary cities or rural areas. (To be fair, though, the majority of tweets from Africa probably did come from these cities.) In other words, the dataset is a great sample of African Twitter activity at the end of 2013, but it is just that – a sample.
Despite the aforementioned critiques on how some have hyped the findings, the data is a sound benchmark for African online communication. It will no doubt prove extremely useful when Portland repeats the analysis at a later date. Other tidbits like how 76% of tweets were in English, French, or Arabic are thought-provoking. So is the suggestion that Accra produces more tweets than much larger West African neighbor Lagos. Also, both the 2013 and 2011 reports mention a relatively low number of politically-inclined tweets. The recent publication notes a heavy prevalence of football discussion which could explain users’ priorities.
Don’t get us wrong: the infographic is beautiful and fun to read!