Information analysis: Business yet to respond to big data challenge
Posted: 22 October 2012 | Author: Charles Batchelor | Source: Financial Times
The banking and insurance industries are used to handling vast volumes of data relating to payments and loans, policies and claims. But how are they responding to the challenge of “big data” – the current buzzword in the field of IT used to define the sheer amount of information institutions have to deal with? Some experts think they still have some way to go.
The amount being generated now threatens to overwhelm the ability of conventional software tools to capture and process the information in a reasonable time. IBM, the technology and consulting group. calculates that every day 2.5 quintillion bytes of data are produced – a quintillion is 1 followed by 18 noughts – with the result that 90 per cent of the data in the world today has been created in the past two years alone. Alongside structured data, including sales figures and share price movements, is an increasing amount of so-called unstructured data, such as emails, posts to social media sites, digital pictures, videos and text messages. Rashmi Joshi, assistant director forensics and fraud at Ernst & Young UK, says: “Organisations are only making use of 20 per cent of the total data available. Most of the unstructured data are not being analysed yet. There is only a limited number of technologies to sift through it.”
Edwin van der Ouderaa, managing director in financial services analytics at Accenture, adds: “A lot of what financial services organisations are doing is actually normal analytics. Big data is a question of petabytes [10 to the power of 15] and exabytes [10 to 18] of information and the technology you need for that is very different. I have not seen a lot of banks that have introduced big, mature applications.”
Steve Shelton, head of data services at BAE Systems Detica, thinks investment banks are farthest advanced in using big data because “their business model involves good analysis and doing it fast. The retail banks are more focused on driving value out of the transactions and payments data they already have rather than mining series of unstructured data”.
The insurers offering personal lines – motor and home insurance – have not had a need for a big data response, though the introduction of in-car devices to record driver behaviour will provide masses of data that they will need to process, says Mr Shelton.
Much of the big data development work in financial services organisations has been focused on improving internal procedures of risk assessment and fraud prevention. However, there are some examples of companies that have applied data analytics to directly serving customers.
ZestFinance, a US lender to borrowers with poor credit ratings, used big data technology to assess its customers. People who took a long time reading its terms and conditions, established by tracking software such as cookies, were judged a good risk. The company has since moved out of loans but sells its analysis to other lenders.
Visa, the credit card provider, went into partnership last year with Gap, the clothing retailer, to alert customers by text message to discounts in nearby stores. With the customer’s permission, Visa uses the data from its global transaction processing network to see where the customer last made a purchase in same district.
Although the banks process the payments that underpin these transactions, they face similar privacy issues to those that have caused problems for Google and other internet search engines, and it is unclear how they might benefit from the information they acquire.
There are also other barriers that have held up the sector’s response to the challenge of big data. There is a shortage of the statistical, IT and data analysis skills needed to collect and interpret the information. These, in turn, need to be paired with detailed knowledge of the business and context in which the data are to be used.
A further challenge is the fragmented nature of governance in many large organisations. Accenture’s Mr Van der Ouderaa says a lot of these technologies cross barriers. “Departments such as risk, treasury and finance have to learn to work together.”
Paul Thomalla, managing director for Europe, Middle East and Africa at ACI Worldwide, a supplier of electronic payment software, says: “The larger banks may have 20 or 30 different methodologies for making payments. We talk to the banks about unifying their systems. At the moment social media are not even in the picture.”