In the recent past we have experienced ground up re-invention of business models. At the heart of these changes are forces dictated by regulatory environment, technological changes and shifting consumer dynamics. The impact of this dynamic shifts are defining the disruptive state of the financial and technology sectors. We’re witnessing an epic, fundamental shift in how technology integrates with, alters, and improves society and its functions.
One such technology that is driving the rapid shift is data. So intense it is that the word “data” has become a deal-killer at cocktail parties. I know from personal experience. I have the data. But data just grows like a weed anyway. It’s so indiscriminately collected and warehoused, like some bland, uninteresting residue that companies dump into the cloud as they transactionally churn away endlessly.
These however is a view held by those who still don’t appreciate the value of data and it’s wrong. Data isn’t indiscriminate. The information logged into all these memory banks are exactly the things that matter. That’s why they’re being recorded. People think data’s boring because they’re overlooking the fact that data is experience it’s a long list of prior events from which it’s possible to analytically learn. Data can mean anything and everything. In its most abstract, it means nothing in particular, but in the particular, it always means something valuable and interesting.
Safaricom is defining the way data is creating meaning. And they are using that data to churn out new products cross-sold to their customer base. The recent Fuliza Mpesa overdaft is such a product born out of this data insights and fueled by data. In the wake of its launch, Safaricom’ s CEO Bob Collimore announced that the product had excided its expectations and that the product hit 1 million subscriptions within one week of launch and its loan book hit a 1 Billion Mark with underwriting form KCB and CBA banks within the same period.
The Fuliza performance signal we are entering a new era where technology trumps regulations. Fuliza has been able to demonstrate that through partnerships with lending institutions, one can bypass the barrier of regulatory confines and deploy products that are rich in data analytics trends. This is a trend that will give banks a run for their money as they are at risk of ceding their customer market share to Telcos.
But how are telcos able to achieve these feat yet banks have been there much longer? Big Data Analytics is at the heart of this transformation. The earliest use of large data sets does not differ substantially from how big data is being used today. What has changed is that computers have become more powerful, computing resources have drastically dropped in price, many different sources of data are now available, and several big data technologies exist that allow for efficiently managing and extracting information from large data sets. The telco companies by default are technology driven compared to banks. This means they have invested in latest cutting edge technologies like Hadoop, map reduce that analyze large data sets for decision making compared to banks.
The first aspect that differentiates the two sectors is Volume of data. The financial and credit scoring industries are no strangers to data. These sectors have been accustomed to data ever since credit bureaus first started gathering consumer credit information a few decades ago. However, they seem to be struggling in the ability to analyze big data which unlike the traditional data that they are accustomed to, is more unstructured. Voice Analytics, Call Data Records metadata, Social Network Analytics are examples of unstructured data that Telcos have at their disposal and have the ability to analyze for lending determination. Telecommunication providers have for example a massive transactional database where they record call behavior of their customers. relationships of networks can be built through understanding of frequency of calls between them and their duration.
The second aspect is the velocity of data. This is the aspect of big data refers to the rate at which data is being collected, stored and analyzed. The traditionally banking and credit score industry does batch processing of data. Hence why the approval loop of your loan application takes hours if not days. In the mobile lending era, which ride on the crest of telco infrastructure, the data generated by consumers on their mobile phones, is produced and processed at a faster speed than the more traditional data-collection techniques. With the rise of dynamic databases that are updated on a, minute- by-minute basis, credit applicants are approved in a matter of milliseconds.
Another aspect of the velocity of big data sources is the ability to build predictive models at a much more rapid speed, and use much more data than before big data sources were available. Banks have traditionally used statistical predictive modelling techniques on sampled historical data to predict what is going to happen next. With the big data storage and powerful analytical software that is now available, traditional methodologies to build predictive models may not be as relevant as compared to previous years, which is why telcos are lending money even when ones history is sparse.
Telcos have more variety types of data they collect. The more modern data repositories need to deal with both today’s rapid speed of data collection, as well as its greater variety. Banking data is mostly structured. Telco data on the other hand is currently being captured in both structured and unstructured formats this includes information from sources such as mobile apps, telecommunications sources, social media data such as connections on LinkedIn, website clickstream and voice response logs. This increase in the variety of data sources allows a telco to get a more detailed view of the person, compared to the more traditional data sources such as credit bureau. Unstructured data sources such as social media data can, for example, verify not only that the applicant really exists, but what types of social connections the applicant has, aspects of the applicant’s personality- using OCEAN model and whether the applicant is deserving of a loan.
Banks are increasingly in a catch 22, erosion of its market share, IFRS 9 looming and declining revenue legacies. These forces are seeing the need of Banking institutions being driven to collect almost precision level data of their customers and find creative ways of funding of capex intensive big data infrastructure that is needed to properly collect, store, and analyze big data sources. Culture transformation programs to eliminate data silos that will poise the banking sector into digital posture away from the traditional regulated risk averse posture. Lastly recruit highly trained data scientists who are able to churn models from unstructured sources.