Blog article
See all stories »

Want More from Big Data? Get Real!

I know it sounds ludicrous, but there are banks which indeed process several thousand transactions per second! Blink. The counter notched up another 5,000 transactions. Oh wait, that’s an additional 20,000 in the time you took to read this sentence.

Welcome to the reality of Big Data, which is growing at a scale that is almost unreal. Being part of a highly information intensive business, banking institutions are among Big Data’s biggest stakeholders. So the fact that most banks process the bulk of their Big Data offline, outside of real time, by which time another million events have gone by, is most perplexing.

Sure, there are “good” reasons for this, including systemic inadequacy and the seeming impossibility of mopping up data as it occurs, from a myriad of channels. But they don’t matter. What does matter is that the lack of online real time intelligence is denying banks several advantages – agility to respond faster to changes in customer behavior; deeper insights into operational, customer and counterparty risks; the ability to test and fine-tune marketing campaigns; optimized mapping of customer type and communication channel; proactive fraud management; and higher operating efficiency. 

A few banks have seized the initiative by deploying high-performance analytics solutions, which act in real time on their Big Data. There’s a Minneapolis-based bank, which uses analytics extensively to identify changes in customer behavior, and leverages that insight to sell more or prevent attrition. First Tennessee Bank uses customer analytics to test assumptions and build product propensity models. It also uses predictive analysis to forecast revenue and Return on Investment on marketing campaigns before putting them out to market. A retail bank in South Africa uses customer analytics to improve debt management and collection. And another U.S. institution has dramatically reduced fraud at its branches with the help of facial recognition technology and video/ customer data analytics. Way to go!




Comments: (2)

Simon Harper
Simon Harper - Morgan Harper - London 12 October, 2012, 09:30Be the first to give this comment the thumbs up 0 likes

I'd agree Kishen - banking tends to be tame when using data as an asset. Some aspects of governance are adopted through operations but not through business development where the potential value seems to be seems too much of a risk to management. Not quite a dog-fooding industry ;-) .  Much seems based on transactional data rather than the bigger picture of 'other' data impacting people & decisions which is being captured by the second and not necessarily used!!

Looking at telcos in the UK, they are honing in on the value: >100k events per second per telco shaping not only propositions to be relevant to customers but also value from aggregated data to 3rd parties. 

Advertising, product offerings, wastage, fraud, relevance all add up to £MMM let alone operational improvements. Waiting to see the value add examples around data & its management (as an asset) as you illustrate go mainstream globally within banking... 

Ketharaman Swaminathan
Ketharaman Swaminathan - GTM360 Marketing Solutions - Pune 14 October, 2012, 10:32Be the first to give this comment the thumbs up 0 likes

@MAKishenK: Nice article. It stands out by giving concrete examples. Let me add 'credit card realtime fraud detection and prevention' as another use case, which is perhaps the oldest and most widespread use of predictive analytics in BFSI. I'm eager to know if you expect the "false positives" problem to worsen when processing shifts from the present batch / sample data to realtime / big data?

Now hiring