‘Big Data’ comes to Africa

CEOs claim it will revolutionise African business. Philanthropists insist it can help eradicate poverty. ‘Big Data’ enjoys a lot of hype. “[It] is all about weeding through huge volumes of data and finding the gems,” says Cory Wiegert, IBM’s director for software product management. Collecting and interpreting big data can provide businesses with unprecedented insight. As Mr Wiegert puts it, “companies become more informed and they make better decisions and wiser investments.”

While African businesses are tapping into big data, the region has lagged the global trend. An IBM report focusing on Nigeria and Kenya reveals that 40 percent of businesses are in the planning stages of a big data project, in comparison with the global average of 51 percent. Twelve percent of Kenyan and Nigerian firms have big data projects live, just shy of the 13 percent global mean. Although the majority of firms in Nigeria and Kenya claim to be capable of having such projects, formal training in management remains low, according to the report, suggesting that the adoption curve has yet to hit a serious upward trend.

The biggest African firms are, naturally, more likely to have big data projects. But in an intriguing twist, smaller firms are more likely to be dabbling than medium-sized firms; 43 percent of small firms said they were in the planning stage for big data projects, compared with 24 percent of medium companies. Mr Wiegert of IBM argues that smaller African firms can benefit from these projects as much as larger companies. “We don’t think that size affects the value that a firm can derive from big data. There are plenty of African entrepreneurs basing their business models on collecting and innovatively deploying big data.”

Stories of how big data is boosting businesses are starting to emerge. Mr Wiegert says this includes all industries. “There are businesses engaging with big data and they don’t even know it. Every sector is grappling with volumes of information that they have to process and make into something.”

The South African insurance firm Santam has started using predictive analytics to make it easier to process claims and spot fraud. As a result, it says it can deal with legitimate claims within an hour; 70 times faster than previously. Being able to tap into big data also means that Santam’s claims adjusters have to visit clients for low-risk claims less frequently, which has helped slash operational costs. The company claims to have saved $2.4m within the first four months of launching the programme.

Meanwhile, another South African health insurance firm, Discovery, has pumped money into an analytics programme to interpret complex health-related data and predict its clients’ problems in minutes. The new operation has also sharpened Discovery’s fraud-detection tools; it can now mine big data from pharmacies and health providers to spot potential phantom drug prescriptions, something that the previous system could not do.

Discovery hopes that it will soon reap the rewards it has accumulated over the years through a programme called Vitality, which rewards policyholders for going to the gym, for a health screening or buying healthy food, and has been collecting data on participating customers.

“The wellness behaviour of consumers, and the way in which incentives drive behavioural change generate significant volumes,” says the CEO of Discovery, Jonathan Broomberg. “Discovery has invested heavily in analytical capability to interpret and derive value from the data collected.”

In January, IBM announced a partnership with South African lender Nedbank to come up with a predictive modeling system so that the bank can integrate real time big data gathered from social media into its main banking infrastructure.

“Say someone rings our customer service hotline, is unhappy with the service they receive and then tweets about it. We now have the ability to look up the banking profile of the person by linking their social media profile with their banking profile,” says Craig Smee, the head of social media communications at Nedbank.

In future, the lender hopes that banking and social media big data will become seamless. “A person’s contact number and social media profile would be linked to their banking data so we would never have to ask for that data when interacting with them over social media. We would have it straight to hand,” says Mr Smee.

For Nedbank, the end goal is to build a single, holistic picture of their customer. For a long period, departments in banks existed in silos, and lacked a singular view of their customer. This made accurate analysis of customers – how they respond to marketing, their views on customer services, their use of accounts – virtually impossible.

By storing, integrating and analysing big data, this can change. “You’ve got to gather all of these masses of data and then simplify it so you can have a single, meaningful conversation with the customer,” says Eugene Liebenberg, head of retail banking at Nedbank.

Data for development

The business community is not alone in approaching the opportunities of big data – development actors are also seeing the benefits. Wolfgang von Loeper, a former farmer, is three months away from launching MySmartFarm app in South Africa. It leverages agriculture-themed data, including climate, soil information and disease, to help farmers make the best decisions about  irrigation, pest control and fertilisation.

“The African smallholder farmer is producing a quarter to half of their potential. Big data and knowledge transfer could help to triple or quadruple their production, allowing them to way surpass their own sustenance requirements,” Mr von Loeper says, adding that big data can also help farmers reduce costs. “If farmers have access to detailed information about irrigation then they can irrigate only when required, and avoid wasting a precious resource,” he says.

Accessing the agriculture-themed big data necessary to power the app is difficult, however. “For much of the data, someone has IP over it and then it usually is not readily available. If anything, it’s becoming more closed, even with regards to research institutions funded by government. That is because the government requires these institutions to search for additional funding through other sources and, one of them is [through selling] data,” Mr von Loeper explains. “So when we get started, we basically need to start from scratch. Nevertheless, as soon as our product starts accumulating data, we’ll source additional open source data. For example international open source weather data helps us a lot already.”

Opening up big data is proving tricky on a global level. In a sign of a possible shift, during a G8 open data conference in 2012, many countries pledged to make new agricultural data publicly available. G8 leaders agreed to share agricultural information with African countries and work towards setting up a global agriculture platform, which would grant Africa unprecedented access to relevant big data.

But there are other problems. Some firms are sceptical about the business justification. “The most common question we hear from firms is ‘what’s my return on investment?’ Companies don’t want to have to wait four to five years to feel the rewards,” says IBM’s Mr Wiegert.

It also seems that many African companies are yet to understand the security benefits of big data projects, which allow firms to access new information on potential risks. Thirty percent of respondents in the IBM survey said they were unsure about whether big data projects had security benefits and 22 percent claimed that they did not.

On the other hand, big data could stir up tensions about security. The collection, storage and interpretation of data about consumer habits is likely to touch nerves about privacy. Cyber criminals will also be keen to get their hands on big data, according to Mr Wiegert. “The inevitable is going to happen so on top of thinking about how we are going to stop it, we need to ask what we do when it does happen, what safeguards can we put in place to ensure it can’t happen again,” he says.

Despite the challenges, when it comes to development in Africa the potential of big data is huge. “It has advantages that the First World would take for granted – from health insurance for the unbankable to sustainable farming,” Mr Wiegert says. “And the appetite to take advantage of the rise of big data is definitely there.”