Keeping Up With Customers: A Requirement, Not an Option

February 25, 2008

Up-to-date records are vital, and Customer Data Integration tools automate the process. These software packages provide competitive advantages can be lifesavers.

In a typical war movie, no matter how hot the battle action gets it’s generally pretty easy to tell the good guys from the enemy. That’s the essence of dramatic storytelling.

The reality of battle is bit messier than that. Often it is difficult to tell who is who and what is happening at any given time. It’s semi-organized chaos, which gives rise to the term “the fog of war.”

While not a matter of life or death, financial institutions today face a similar situation – what you might call “the fog of data.” They are constantly collecting data about their customers, the market, financial regulations, etc. This data is then funneled into various applications where it is parsed and stored until needed.

Unfortunately, like old Army K-rations, the integrity of data tends to degrade over time. What was once true may not be true anymore. For example, a once-legitimate account may no longer be part of the client’s business but is still linked to it. In other cases there may be underlying factors that can affect risk. For example, a regulated mutual fund company doing business with an institution might be regarded as a low-risk client. One or more of its accounts, however, might be high-risk because of geographic location or other considerations, and thus could have an impact on the client’s overall risk profile.

Whatever the specific cause, the net result is an inaccurate picture of the client that can have an adverse affect on the institution’s ability to manage risk. This, in turn, can create potential violations of anti-money laundering (AML), Know Your Customer (KYC), and other regulatory requirements.

While organizations often will make an effort to cleanse the data in order to address these and other concerns, usually this work involves intensive, manual, one-time fixes that address issues within specific institutional operations. It does not carry forward, and it does not cross over into related areas. In addition, this work lacks the necessary controls to keep the problems from recurring.

What is needed is a way of centralizing and automating the data cleansing effort to create an approach that is sustainable and continuous, with data cleansing integrated into data gathering. Such a system will help improve the consistency of client data, remove inaccuracies, and ultimately improve risk management and overall business performance. All of these benefits can be found in customer data integration (CDI) software solutions.

Starting with good information

In movies about modern warfare, at some point the chief strategist will be looking battle intelligence, perhaps a set of satellite photos and say: “This intel is 24 hours old! Get me current data.” He knows that a great plan based on the wrong information is doomed to failure.

The same goes with the new wave of enterprise data management (EDM) initiatives. No matter how well an application is designed, if the information going into it lacks integrity, pre-existing inadequacies are carried over to the new system.

Cleansing the data as the organization migrates to the new system will solve the problem – at that one point in time. But that can be a massive operation. And once the project is completed, there is a high risk that the data will fall back into disrepair.

CDI software creates a long-term, sustainable solution. It cleanses the data before it is moved into the new EDM system in less time and at a lower cost than can be achieved through other means – including shipping records offshore for manual review. It also eliminates manual intervention by people not familiar with the context of the data, eliminating potential risks to data accuracy.

Where it really pays off, however, is in creating a set of probabilistic and deterministic rules that can be applied to the data going forward. The automated system is constantly comparing the data to third-party reference data from a trusted source to confirm the accuracy of information, such as client, account, and hierarchy. A good application will also create immediate reports to expedite and improve the accuracy of the remediation process, saving months of work. The net result is a clean set of data for the EDM system to apply, helping reduce risk due to data that is old or out-of-date.

Getting to the cause

While no financial institution sets out to accumulate bad data, the fact is it happens more often than anyone would like to admit. In some cases it is a case of simple operator error – client data from one account being associated with another that has a similar name. Sometimes it’s due to the natural course of business, such as when a once-legitimate account is no longer part of the client’s organization but is still linked to it. There could also be duplicate account entries in the client’s systems, perhaps caused by a merger or acquistion. In that case, these duplicates could contribute to an inaccurate view of the hierarchy structure by overstating or understating an account’s contribution to the client risk profile.

All of these situations are difficult to see with typical legacy systems in the course of doing business. These systems typically lack edit and confirmation capabilities, allowing accounts to be opened without entering all the required information.

The problem is these inaccuracies can result in incorrect risk ratings, which exposes the institution to violation of AML requirements (along with the associated risks and penalties). Inaccuracies also usually have a ripple effect on multiple downstream systems that rely on the data.

A CDI system is able to scan client records, identify those that require remediation, and assist with the fix. Take the case of a small group of records requiring remediation within a larger group. Probabilistic matching allows the organization to identify problematic records quickly through an iterative evaluation of attributes that yields meaningful record groupings.

An example would be a client whose accounts have a similar name and address but have various account codes. The challenge is to identify the relative few records that, while similar to the others, are different enough to stand out and are not grouped due to subtle variations in the attributes. The CDI software’s probabilistic approach allows the organization to identify these situations accurately, without creating and maintaining complex rule sets. This method simplifies the entire process, making it easier to use today and tomorrow.

Seeing through the fog

In the “safe” world of movie battles, the information is clear and the good guys always win. But in the real world of financial institutions, it’s not so cut and dried. And when the fog clears, organizations can find themselves facing serious fines and other consequences. To follow the military analogy, institutions can win the battle of capturing customer data, only to lose the war of risk management and compliance.

CDI software helps cut through the fog of data by dramatically improving the cleansing process, in the short- and long-term view. It creates a repeatable, automated data remediation solution that increases data accuracy while reducing time and costs. This, in turn, improves compliance with AML rules while preparing data for transition to a new EDM client data platform.

In short, it provides the air cover for the organization’s ground troops, clearing the way for better client data management while reducing risk. This is the key to helping the good guys win.

Scott Schumacher, Ph.D., is senior vice president and chief scientist at Initiate Systems, Inc. He is responsible for the research and development of algorithms for Initiate’s data integration solutions. His business experience includes leading advanced surveillance algorithm projects commissioned by the Department of Defense. He holds advanced degrees in mathematics from the University of California, where has served as an associate professor.

Posted by Carl · Filed Under Uncategorized