> Dialogue > Dialogue

Big data – structuring the unstructured

| 2013-06-24 | Hits:

coreThe concept of big data has been making its way into debate among industry participants, but what does it mean? Julian Wilson, Europe’s strategic alliances director at FreedomPay, a company working on capturing data and turning it into business intelligence, describes big data as an opportunity for banks.



What does big data mean?


Big data is the availability of so much unstructured digital information that it cannot be handled by our current approaches to information technology. Big data today, in a bank context, may mean having knowledge of the clients that can be discerned through the point of sale (PoS), what they buy, where and when. A big data example might be inferring meaning from that extra information.


How can banks use this information?


In recent decades many banks have been driven by the reduction of costs and increased efficiency. Many of us feel as though our banks are run by algorithms and call centers. We only seem to visit the branch when there is a problem. But technology may offer the solution to the problem it helped create. Big data offers banks the opportunity to re-engage with their customers. This statement must of course be broken down into specific objectives which can be measured, for example, the number of customer interactions with the banks, inbound dialogue, satisfaction ratings, transaction volumes, new service take-up and revenues.


How do you translate big data into valuable business intelligence?


People who work with big data are data scientists, not business analysts. PoS, card transactions, social media, news, weather, client location, football results, CCTV footage are all examples of big data. Big data's promise is to allow you to infer a correlation or to ‘spot a connection’. These ‘relationships’ feed into analytics, the goal of which is to predict outcomes such as what people want to buy, for example. Predictions go into marketing and end up as campaigns, which in turn cause decisions resulting in more data to capture and learn from.


What types of business may a bank pursue with the data?


I don’t think there is one answer to that question for all banks. I think that each bank will find different answers that sit most comfortably with their clients and their views. I believe, and these are my views, that I have a different relationship with a retail outlet than I do with a bank. I wouldn’t like to see my bank sending offers and promotions, even though it may have the ability to do that. What a bank could create are horizontal services, which would facilitate its community, such as retailers, to have relevant promotions and offers for themselves. Also, social media, which is a form of big data, must be embraced as a means of listening to, learning from and communicating with clients. The reason is simple – you will hear objective ‘truth/perception’ more quickly and at a lower cost than by recording calls between call center staff and clients.


Would banks sell the data?


They could sell the data subject to appropriate permissions being granted. But while selling data is an option, it is the low value ‘easy option’. It’s a bit like missing the profit opportunity by selling corn versus selling corn flakes. The banks should, in my opinion, assume the costs of doing the harder things with big data, such as aggregating, integrating, inferring correlations, etc. and then make that information more accessible to more people.  Help your business and personal clients to ‘ask questions of that data’.


Could it infringe on people’s privacy?


There is a lot of talk in the industry about who owns the data. I think that’s misleading. What should be asked is what permissions exist for that data. The laws governing permissions for use of and ownership of data are still evolving, and it is important that all banks seek and secure the necessary ‘permissions’ from their clients for the use and rights.  However, banks have an excellent opportunity to offer a range of ‘identification or association’ options, from anonymity to pseudonymity to fully-identified.  When I buy a domain name on the Internet I have a pop up option asking me if I’d like to pay to have my ‘details’ hidden from the transaction.  Why can’t this model work on the high street?


So where should banks start when tackling big data?


Of course not all banks’ starting place will be the same, nor will their strategies. With that said a vision for the bank is critical to provide a context for decisions. Starting points could be combining the card transactions history with Level 4 PoS data - what people bought - to learn what drives transactions. I would also advise banks to embrace social media to augment the ways they engage with their clients.