Two weeks ago, I blogged about the language that vendors use, and yet, I committed a faux pas by not explaining the dual meaning of integration. One meaning has to do with the task of the acquiring vendor's CFO. His job is to eliminate redundant positions, consolidate real estate and leases, slash overhead, consolidate data centers, consolidate software licenses, regulate purchase agreements, neutralize human resources policies such as benefits plans, and try to get the acquired company to adopt the acquirer's culture. Except for the last goal, the task of consolidating the business aspects of two companies takes about 90 to 120 days after the merger is finalized. It's a tough piece of work but doable, especially since good CFOs are constantly looking for ways to cut costs. The other meaning of integration has to do with how the disparate systems of each company will be meshed harmoniously, and that's the job of the CIO plus hundreds of techies. That task takes about five years and lots of money. Even after it's done some people will never be satisfied because "foreign" systems are never consolidated as well as single-designer systems.The reason this is an important thing to understand is that the interchangeable use of the word "integration" causes confusion. For example, investment analysts want to know that an acquiring company has integrated the business pieces. Bankers, on the other hand, couldn't care less. They care about the benefits they will derive from the integration of systems. The difference between 90 to 120 days and five years tells it all.
If this blog were a Harvard Business School case study, two companies would take center stage-Fiserv and Fidelity National Information Services (FIS). The reasons should be obvious. Fiserv's middle name is acquirer-more than 150 in 23 years. FIS wouldn't exist today if it not for acquisitions. The company never owned anything organic.
The complexity of integrating disparate systems can be demonstrated in part by just a few considerations that developers view as sacred. Once recognized, it then becomes rather clear why systems that were created years before a merger, in completely different "camps," can't possibly be suited for perfect integration today. Please note that "perfect" is a word that has never been part of my criteria in the selection of a system for any bank.
A Few Characteristics of an Integrated System: • The adherence to strict standards created by the chief architect of a system; (usually unique because industry standards don't exist). By the way, does any company have the title Chief Systems Architect (CSA)? In Automation in Banking - 2008, only one company lists a close facsimile-Jack Henry & Associates with Chief Technology Officer (CTO). • Tran codes that cross boundaries of all applications within a suite so that meanings become common; • Field designations (data elements) that are common in all applications; • DBMS rules-There shall be one and only one. Can you imagine integrating two applications, one that uses Oracle and the other Sybase or DB2? Worse yet, what if one used Informix? • Data entry rules using a consistent set of integrity tests; • Data security rules that follow a prescribed set of intensity tests, unlike for example, the variety of TSA screening procedures at various airports even within the U.S.; • Common programming conventions and languages; • Standard user manual formats; • Standard systems documentation.
The cost of integrating systems is very high, but never revealed. Fiserv paid $4.4 billion for CheckFree. Owning the #1 Electronic Bill Presentment and Payments system is a very good thing, but it also brings with it greater challenges such as integrating it with 18 Fiserv-owned core systems. I don't believe Fiserv acquired CheckFree just to leave it on an island without some connection to its core systems.
On the surface, it looks as though FIS made 14 acquisitions in five and a half years. But if one peeled back the onion a bit he/she would find that InterCept and Aurum Technology had about 19 acquisitions that were still dangling as "island" systems. All told, FIS is a company consisting of not 14 companies but 41 companies in terms of systems owned.
I don't think acquisitions are wrong or that they should stop. I just believe that after the businesses come together, one should be able to see a clear picture of the resulting system mosaic as opposed to an abstract image of lots of pieces glued together.