In Search of Systems Integration

If you are looking for perfect integration in bank systems, may I suggest you add these items to your dreams in your pursuit of the "Noble Ideals Prize" -- world peace, made-in-the-U.S. energy, clean planet, government designed for citizens, zero deficit, 5 percent unemployment, fair taxation for all, politicians who govern for the people, crime reduction, better education, an inexpensive chicken in every pot, your style of happiness, and a 75 degrees cap in July and August.

The 70 percent goal is not just a guess. It is derived from the fact that a bank system is made up of two parts -- the core plus a variety of ancillary applications. The core should be 100 percent integrated because to have deposits, loans, financials and customer database in separate silos is like dividing 50 states into four countries. Hmmm. Then the ancillaries could make up the balance from a variety of specialized vendors based on the preferences of bankers. The extent of ancillary best-of-breed choices impacts the degradation of the 100 percent. In the case of my clients, the average number of foreign app vendors was seven.

Following are some facts that I picked up based on client projects:

1. Large banks violate every rule regarding integration, but they can't help it. Every application is a stand-alone that evolved as a result of in-house development projects, specialized vendors preferred by a user department, acquired as a result of a merger, built-to-specs by a software developer, or a carry-over from a time so far back that nobody knows where it came from, but is being maintained by the bank's in-house COBOL crew, all of them equipped with walkers.

2. At the other end of the spectrum are the very small banks (under $200 million in assets). They are the most likely group to claim legitimately that they have total integration. They buy a package deal from a thoroughly checked out popular vendor, and as some menus say on their daily specials, "No substitutes please." There are some vendors that built integration into their system from day one. I believe that prize goes to the founders of Information Technology, Inc. (a company acquired by Fiserv in 1995). Don Dillon and Dale Jensen built a true turnkey product, so tight with integration that bankers loved them or hated them -- "We decide what's good for you." In time, however, ITI loosened the reigns, and gave customers choices. Score one for liberty, subtract one for integration.

3. The mid-tier banks are somewhere in the middle in both size and style. They go with the integrated core but then it's best-of-breed the rest of the way. That is not so bad as long as the bank understands that "interface" is not "integrated."

Even after defining the boundaries, there are still some gray areas. For example, some ancillaries are like salt and pepper on every dining table. Harland Financial Solutions, for example is certainly not the largest core vendor but their origination products are found in 40 percent of all financial institutions. Their names are industry standards, LaserPro and DepositPro. Other examples include Asset/Liability Management (ALM) systems. ProfitStars (a Jack Henry company) and what I'll always remember as Sendero (a Fiserv company) practically own the ALM space.

So even if some foreign apps defy the strict rules of integration, it's no different than when U. S. Immigration Services looked the other way as rocket scientists like Wernher von Braun entered the U.S. right after WWII. Or when doctors from India are now allowed in as long as they settle in rural areas that have no doctors.

Integrated systems are indeed a good thing, as long as you don't believe every vendor's sales pitch, even when they use the adjective "fully." And above all, as long as you don't force a vendor to degrade their integration by introducing your best-of-breed that no one ever heard of. Seventy percent is all I promised my 321 clients, and they are still using the same 100 percent they bought.