FEATURE/FUNCTION is one of the most underappreciated, yet critical, subjects in all of banking technology.

The words themselves are innocuous enough. "Features" are those attributes of banking application software that serve the needs of the business. For example, the ability of mortgage processing software to keep a current property appraisal amount and date - as opposed to the original appraisal - is a currently needed feature.

Function means the relative capacity of the software to carry out the purposes of the software. For example, the ability to provide combined statements to retail depositors is a well-established feature. But, the number and type of different accounts that can appear on the statement is a function. The volumes that can be processed - as well as seemingly mundane things like field lengths - are also functions.

Application systems constantly require additional feature/functionality. These enhancements usually come from the business unit and reflect business needs and trends.

For example, one major savings bank we had many commercial and residential mortgages in distressed areas where real estate values had fallen significantly.

This situation drove an expensive investment in two system-based capabilities to reduce the risk. First, the bank flagged accounts where foreclosure would be more likely since the current appraisal value was below the mortgage balance. And, they had to track two balances: the original and still legally due balance, and the new, reduced balance based on the current appraisal.

Feature/functions can be complicated to those who don't know the business. Consider this pending requirement at one institution: "Flag the [IRA] retirement record for customers who receive distributions from other financial institutions and suppress yearly notices for flagged customers over age 70 1/2. Create a report of all notices sent to decedent and beneficiary pension accounts five years after the death of the participant." It's not easily decipherable except by the people in the know - but for them its day-to-day business.

Banking feature/functions have never been categorized in any useful way. There is no reliable way to describe accurately the universe of feature/functions and measure how big the different types might be. Nor do we have any kind of theory that states what types are gaining currency as time goes on.

A brief review of another retail bank's feature/function enhancements did allow us to extract some general categories that described a lot of today's investments in feature/functionality:

* Automate the offering of a new product or feature visible to the customer. Example: offer a bonus percentage paid on monies deposited into a new time account.

* Do something to replace or reduce a manual back-office effort. Example: produce automatic interim statements on commercial accounts; order new mortgage payment coupon books simultaneously while doing escrow analysis.

* Perform new calculations with prior or new data. Example: a new effective yield adjustment option for FASB-91; spread mortgage escrow shortage or overage across a certain number of months.

* Put new controls on what events or transactions are allowed. Example: require that a trustee must disburse money debited from an escrow account within a fixed number of days.

* Provide information to the customer, regardless of medium or frequency. Example: add unapplied balance to yearend statement.

* Provide information to anybody not a customer. Example: create file maintenance history for [mortgage pool] investors.

* Provide additional internal reporting, especially through the on-line screens. Example: display two available balances on screen, one for customer available and one for inclearing available.

* Add fields so that additional data is kept and/or processed. Example: store customer's seasonal phone number in the relationship customer information file.

Feature/functionality is cumulative and constantly seems to advance. New feature/functionality is almost always added to, but does not replace, previous feature/functionality. Of course, some transactions or data categories are sometimes eliminated, but the net effect is that total work done by the applications goes up. There are just constantly more customer products and features, more things automated, more calculations, more data fields, more controls, and more reports.

In fact, feature/functionality is like an advancing frontier, constantly moving into uncharted territory. Once today's needs are adequately incorporated into feature/functionality, banks move onto tomorrow's needs. This frontier is very large, but has individual territories governed by the individual businesses served by each application program. For example, the global custody business is very far removed from retail mortgage lending. Both are seeing the same need for additional reporting functionality. But the specifics of what needs to be reported are totally different.

Thus, at each application's frontier, we find new feature/functions that represent the cutting edge in terms of what technology can do for that business. By knowing what the currently "hot" feature/functions are, banks can help measure the trends in the business. By knowing how their feature/functionality compares to others', they can know whether they are ahead or behind.

Feature/functionality has only a loose relationship to technology. A business requirement often makes no statement about how that requirement is to be fulfilled. A calculation must be done in a certain way, independent of whether it gets done on a midrange or mainframe computer.

Some reporting features vary based on the reporting channel. A paper-based customer statement or letter is going to have different data and be coded very differently than an electronic format report. An important function of all applications is whether they have interfaces to, or are integrated with, other applications. The ability to do this is very much related to the underlying technology.

The choice of technology is not based on the required feature/functions per se, but more on the volumes required, the speed with which new features are to be produced, the availability and/or security of the system, and the cost to build the feature/functions.

Basic changes in information technology are profound and they are leading to new possibilities in creating feature/functions. One of the biggest problems with older mainframe legacy systems has always been that new feature/functionality couldn't be added fast enough - or with reliability. Yet, one of client-server technology's biggest potentials is the speedier creation of new feature/functionality.

Thus, as client-server (and other new technologies) penetrate banking in the years to come, we can expect the total amount of banks' application program feature/functionality, already large, to become positively massive.

No one knows exactly how many features and functions banks have. A conservative estimate might be in the hundreds of thousands, depending on the exact definition of a feature. It's reasonable that this number grow by a factor of 10, into the millions, in the next decade as client-server becomes firmly established.

We can expect this growth in feature/functionality to create its own problems, as banks struggle to control their new frontier. For example, a runaway acceleration of feature/function might end up deluging customers and third parties with too many reports - with conflicting data, to boot. So, a new feature might arise that filters reports before they go to the recipient. Also, it becomes harder and harder for bank employees to learn and remember all the features and functions. New features will be required just to manage human interaction with the deluge of existing features.

We may expect the businesses themselves to continue to be fertile sources of new feature/functionality. Another example: Credit card issuers would like to have more than one "bucket" for grouping and pricing transactions appearing on cardholder statements. ATM withdrawals could be totaled and tracked separately from other types of cash advances. Certain purchases might be isolated and given a "90 days same as cash" pricing. Such examples can be repeated almost endlessly across banking's approximately 150 different lines of business.

The bottom line is that the frontier of expanding feature/functionality deserves more attention than it typically gets. Feature/functionality is the manifestation of technology in the business, yet there is neither an accepted theory of how feature/functions come into being nor is there a measure of how much feature/functionality already exists. A better understanding of this area can only grow in importance as new technologies enable banks to vastly increase the total amount of feature/functionality embedded in the industry's software.

Mr. Teixeira is president of the Tower Group, a consulting firm in Wellesley, MASS.

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.