Today, many companies are prone to serious forms of attack from customers, not just hacking, though there is a relationship between this and the move to more remote forms of doing business, whether Internet or call centre. These attacks are attempts to commit fraud - money laundering, illegal trading - or to exploit loopholes in credit or insurance products. These attempts are helped by technological advances because companies have fewer face-to-face opportunities of validating identities, credit-worthiness and the like, and also because the perpetrators often work in teams, using Internet and mobile telephony to communicate quickly with each other in ways authorities find hard to track.
Need for Protection
The response of government authorities has been to impose upon companies - particularly those in the financial services sector - a series of requirements in terms of 'know your customers' and 'track and if necessary unwind your transactions'. These requirements incidentally sometimes conflict with data protection and industry-specific requirements designed to protect honest customers. For example, money-laundering monitoring requirements include the need to retain and be able to retrieve information, track and report to the national authorities. For this, record keeping and reporting products are designed to track and report information to regulators or other national authorities, where the detailed requirement, such as a threshold value, is specified. So, financial services companies are now investing in a number of areas designed to support their war against the bad customer. These include very advanced software, designed to pick up new patterns of bad behavior, the data storage required to ensure that everything can be tracked and where necessary is reversible, and the systems integration work (often including specialist middleware) to ensure that data from various sources can be brought together for 'detection' work.
What is interesting about these developments is that they are some way in advance of recent developments in database marketing in companies with very large customer bases and/or very high volumes of transactions. Rules-based products are used to monitor transactions to identify and filter out potentially suspicious transactions that are outside the norms for an account and customer (or other group). These 'outside limit' or 'out-of-character' transactions are flagged and routed for manual review by bank compliance, audit, or risk management staff. Transactions that appear, after review, to be inconsistent with a customer's business are reported to the appropriate government entity. To achieve this, expert and/or intelligent systems identify non-linear trends and relationships within transaction activity, including associative patterns among accounts, customers, relative to peer or other groupings. These systems examine all possible combinations among transactions, rather than looking at the individual transaction record itself, and apply risk scorings to suspect transactions. There can be significant differences between systems and results depending on the architectural approach and the intelligence technologies used. The most advanced types of software can be programmed to warn about virtually any kind of risk - including imminent system failure due to lack of capacity.
Value Vs Risk
In the financial services industry, value and risk are not far apart. In call centres and Web sites, where the customer is invisible to the supplier, systems have assumed the role of detecting bad customers. The pay-off to good risk management has always been much larger than that to value management, at least in the short run, simply because the costs of failure are so large. However, with increasing volumes of business being done remotely, the need for smarter systems to help companies extract much higher value from customers is increasing. For this reason, banks are starting to investigate how the software and processes they use for detecting bad customers more quickly and accurately can be used for detecting value more quickly. One reason for this is that these products are not cheap, so banks want to extend their use. The products are very advanced, and the very latest products have shown dramatic levels of success, in terms of reducing the incidence of false positives and successful detection of negatives. They do this by creating 'sentinels' which take on much of the role of the data analyst, freeing analysts to work on strategies. For the data analysis industry, these products represent both a threat and an opportunity - an opportunity if used to liberate analysts to focus not on implementing models but on understanding why they work, but a threat if analysts try to compete with them.
Building a Fraud Intelligence in CRM Apps
One of the definite value add beyond having Financial CRM Vertical application is to have Behavioral & Transaction Detection Intelligence which eventually would lead to detect Fraud and ability to differentiate a good customer from Bad Ones. The CRM industry is up for a revolution and customer prefer more of Virtual Services than to be served in person in lieu of enablement of Web2.0 Self Services, Mobile Computing and Kiosk terminals etc. Real Time Decisioning and Intelligence is of great value to render fraud free services to customers ranging from High Net worth to Moderate Value customer.
Loving P&C
DC*
No comments:
Post a Comment