insight2impact (i2i) is exploring how data can be used by financial service providers to create client value and enhance firm value simultaneously.
We interviewed Morné van der Westhuizen and Alex Shabala to understand the role of data in decision-making at Zoona, an African mobile payments operator.
This topic will be further explored during the invitation-only roundtable discussion that i2i will be hosting during the Chief Data & Analytics Officer Africa 2017 conference on 5 July 2017.
How is Zoona making use of alternative data to inform decision-making and product design?
If you consider that more than 70% of the population in Zambia is not formally banked, then it is easy to see that the reference documents typically associated with credit scoring are often inadequate in that country context.
We have a large footprint in the country so, in designing for our 3.9 million consumers in Zambia, we rely on other forms of data and we are forced to be creative about it. Network analysis can be a useful tool – we can estimate how trustworthy you are based on who you know, specifically if you interact or transact with people we do know more about.
Geospatial data is particularly valuable to us. We use the data to determine financial inclusion and market penetration, but the problem with the geospatial data that is readily available for the markets in which we operate, is that it is not granular enough.
When we are moving into a new area, we send out an advance team to scope the landscape. They log physical landmarks such as markets and bus stops. As money flows traditionally follow bus or taxi routes, it can be helpful to get into a local taxi and travel on key routes. The frustrating thing is, if you are fortunate, you may see a small, hard-to-notice sticker in a taxi that indicates the route map. In Africa, data often lives on paper and is not digitalised.
What is your biggest data-related headache?
In addition to some of the issues mentioned above, one of our biggest frustrations is that, often by the time we get access to data, it is already outdated. Mobile money data is useful but it is anonymised, which makes it difficult to analyse. Much of the data we are interested in is locked into our competitors – in our market the mobile network operators are competitors in the mobile money space but we are not their competitors in offering mobile calling services.
So, the mobile operators aren’t willing to share and the government statistical services in some countries are not advanced enough. For instance, one of our team who was visiting Lusaka happened to spot a billboard that advertised population stats from the most recent census. However, the same statistics were not available on the website of the national statistical bureau.
Do financial service providers (FSPs) in Africa have different data concerns and challenges compared to FSPs in other regions?
There are gaps in the available data available that wouldn’t necessarily be a problem in other countries because governments or other organisations take ownership and release that data. For example, if we were using Google Maps in certain markets, we could map things down to a street corner, but Google Street View isn’t a thing in many African countries (apart from South Africa). Africa’s data environment is not as well-mapped or well-maintained.
In Africa, 40% of online transactions are via USSD and an app that is 100Mb might take a really long time to download in some parts of Malawi. It is important not to get caught up in “new, shiny tech” and its potential for data collection, as any deployment of new technology needs to be balanced with the reality of operating in specific environments.
How important do you think hyperlocal data is? In terms of your client segments, do local (micro) differences matter?
For Zoona, our agents are our immediate clients and they in turn deal directly with the consumers of our services.
We have developed a performance measurement system that includes specific targets for each agent in the network. These targets are customised for each outlet and are based on a range of data points including historical business data, geospatial data and financial inclusion data, etcetera. We are differentiating our targets for agents spaced only 1km to 2km apart!
We are actively pursuing new opportunities for collaboration in providing financial access to the unbanked. Moreover, we make an attractive partner because we are willing to go to markets where others do not want to go and so our penetration rates and local knowledge tend to be higher.
In terms of the provision of disruptive financial services to the underserved, which player is the one to watch?
We think Zoona is one of the most innovative operators in Africa but Barclays Africa has an amazing approach to data science and is demonstrating that not all banks are dinosaurs. MyBucks is doing some intriguing work incorporating artificial intelligence and Capitec is also interesting to watch.
Having said that, I (Morne) talked to French microfinance company Advans recently. In my opinion, they are still “in preschool” when it comes to data analytics and innovation and yet they have a sizeable loan portfolio of €719.9m, much of it with SMEs in Africa. So, you don’t always need to be data heavy or disruptive to successfully gain market share.
How are you incorporating consumer insights observed by your agents in different markets into your overall business model?
Historically, we haven’t always been great at this but we are currently doing a lot of work on capturing consumer insights.
However, our new products are not designed by the head office team, they are designed by the people, for the people. We recently introduced Sunga, which is a savings type product. Within 30 days of launching there were already 60 000 consumers with estimated $500 000 in savings with Sunga – proof that this approach is a winning one.
Morné van der Westhuizen is the Head of Data Analytics and Science at Zoona. Having studied a B.Comm. Mathematics (Actuarial Science) at the University of Stellenbosch and a B.Comm. Management Accounting through UNISA, he then worked as BI Consultant to some of the biggest insurance and investment companies in South Africa. Morné has been responsible for setting up Zoona’s Data Analytics function from scratch, building it up to a team of 10 that has become core to the company’s functioning.
Dr Alex Shabala is a Data Scientist at Zoona. Prior to joining Zoona in 2016, Alex was a Postdoctoral Research Fellow at the University of Cape Town. He has a Bachelor’s in Physics and Mathematics from the University of Tasmania and a Ph.D in Applied Mathematics from the University of Oxford.