Weather, moods and money: exciting new things for African digital retailers, where the skies are your friend

With the virtual trial room, reduced product return rate to nearly zero, and purchases of products available with 3-D trial is 400% higher

TAMPA, FLORIDA: E-commerce represents a small, but rapidly growing segment of the African retail market: a recent study by McKinsey highlighted that by 2025, e-commerce could account for 10% of retail sales in the continent’s largest economies, translating into some $75 billion in annual revenue.

The growing popularity of online shopping, particularly among the African middle and upper classes, is driven by some of the usual things – increasing internet access and smartphone ownership, busy lifestyles, and a preference for convenience.

But there’s also some unusual drivers for the soaring popularity of e-commerce, that, on the other hand, make African urban life particularly exasperating: nightmare traffic jams, bad roads and a lack of parking – instead of putting yourself through all that, the easier thing is letting someone deliver your items to you.

In South Africa, a 2011 MasterCard worldwide survey revealed that 51% of those with access to the internet were already shopping online. Currently, in Nigeria, the online shopping site Jumia has 100,000 customer accounts and its sales are increasing by 15% a month.  

The question for online retailers is how to maximise revenue, particularly given the unique opportunity that doing business online offers: volumes upon volumes of data – most of which marketers have no idea what to do with.

One of the year’s biggest digital marketing conferences just ended  in Tampa, Florida, where the focus is how Big Data analytics can help marketers and e-commerce professionals learn, predict and guide customer engagement on a scope and intensity never yet seen before.

Conference convener IBM, a leading global company in business-focused tech solutions, was showcasing its famed cognitive computing platform named “Watson”.

Dark, messy, unstructured

“We are at a new dawn for commerce,” said Harriet Greene, IBM’s general manager for the Internet of Things, commerce and education. “Until recently, we just didn’t have the computing power to analyse this information. Cognitive computing allows us to work with larger, more diverse, dark, messy, unstructured data. From this we can discover new patterns and drive new insights,” she said.

Take, for example, a young urban African who decides to take up a new hobby – say cycling, yoga or boxing.

The first thing they are likely to do is type something in a search engine, for example “yoga poses for beginners”, or “good cycling routes in Nairobi”.

They may even post on their Facebook page or Twitter profile questions to their friends asking for recommendations on where to attend a session, find an instructor or buy equipment.

What cognitive computing does is detect that this is a beginner and so already open to buying things; the site then responds with visual, targeted advertising on bike deals or yoga mats at a location nearby, combining data not just from explicit search terms that they made in the search engine, but also their browsing history and the unstructured chatter that they carried on via social media.


As they become better at their new hobby, the system’s self-learning algorithm adjusts its offerings and fine-tunes the experiences over time.

Another exciting solution is being deployed by online retailer, based in India and whose representatives were present at the IBM Amplify conference.

The site’s storefront is now using a 3-D virtual trial room, powered by IBM and integrated business partner Metail. The feature enables customers to input their body measurement and virtually try on any garments before purchase.

The average return rate in the apparel industry is 20%, mostly because customers are not happy with how they fit. Since launching the virtual trial room, has reduced the product return rate to nearly zero, and purchases of products available with 3-D trial is 400% higher than the industry norm.

The possibilities go even further – IBM recently acquired the Weather Company, the world’s biggest private weather forecasting firm.

Looking to the skies

The Weather Company produces forecasts for 2.2 billion locations across the world, every 15 minutes – a total of nearly 26 billion forecasts every day.

With the acquisition, IBM is now the largest weather-based and second largest location-based company in the world, the IBM Amplify conference was told.

“Weather is a critical yet largely misunderstood data source,” said Greene. “Weather influences when and where people go to shop, how long they stay out, their mood and behaviour, even how much they are willing to spend at any particular time.”

Studies (from temperate countries, but perhaps generalisable to Africa) have shown that cold weather reduces dexterity, muscle strength, blood flow, and balance; rainy weather increases carbohydrate cravings and sensitivity to pain; and sunny weather makes you want to spend more.

“Using this very targeted and location-specific insights, marketers can craft more relevant campaigns, and fundamentally change the way marketing is done,” she added.

Related Content


blog comments powered by Disqus