(from Professor Denert 1)

This is a unique book in its field. I know of no other, at least not in the German language, that deals with IT management, such as Dr. Karl-Rudolf Moll has done here. We have many books about project management which deal with the planning, implementation and control of software development. IT-management though is more than just this. It is something more specific. Namely, how the various units of an organization should be organised and guided so that each of their IT needs are catered for? There are many management books that deal with purely organisational issues. None though deal specifically with IT and come with such expert experience as this book. The practical experience and insight that Karl-Rudolf Moll has from his management job as CIO of the Bavarian Hypotheken- und Wechsel-Bank AG in Munich is summarized in this this book.

Data and information processing are fundamental to the businesses of banking, insurance and travel agencies. For manufacturing industry, information processing can also play an important role. Its significance though varies between industries, i.e. car production has a much higher dependence on information systems than say the steel industry. The underlying differences in the role of IT in a company greatly influences the management and design of its IT systems. A conglomerate with multiple production centres can (and should) decentralise their IT systems and data processing units. A commercial bank though must operate its databanks centrally. The thought of doing it on a branch by branch basis is absurd. The flight reservation industry also needs its systems to be operated centrally. There can be no arguments made as to why it should be done otherwise.

Buzzwords such as downsizing, client server applications, open systems and the death of the mainframe platform are the current hot topics in the IT industry, at least according to the popular industry press. Sometimes, however, they end up throwing out the baby with the bath water. Sure enough many applications did not need to be hosted on a single supercomputer, and were better suited to a decentralized Unix style platform. It can also prove useful to introduce a Client/Server-Architecture straight to the workstation. But it is almost never reasonable, to divide up one database (a logical unit in itself) amongst several machines in different locations. For many information dependent firms it is necessary to operate one large data-centre, or small number of lesser ones. This is not because their IT management hails from the technical "middle -ages" but because there are rather good reasons for these platform choices:

  • Booking systems (bank accounts, flights, hotels etc.) need a central database (ultimately a single bit is all that matters - and you cannot distribute a bit).
  • The management of certain procedures can be configured in a more efficient manner centrally, for example the printing, collation and the distribution of documents, e.g.  Insurance policies. (Postal charge optimisation can only be achieved in the situation of large postal deliveries.)
  • Batch processing, that is typically performed across an entire data body, can only be efficiently executed on a central computer.
  • The data safety, that a bank must have in its data processes can only be guaranteed through operating a central data centre. The structural, organisational und staff arrangements are also much too costly to provide across multiple locations.

Those who are not convinced of this reasoning will think otherwise after having read this book, especially chapter 5 - Data Centre Operation. It will then be clear that Karl-Rudolf Moll is not behind the times in keeping faith with mainframe computing. He sees that IT management in the banking world is a highly complex task which belongs to a large Data Centre supported by mammoth databases. At the same time a decentralised infrastructure can tend to the customer support elements such as  communication technology, servers, PCs, counter-terminals and cash machines.

Ernst Denert

1 non-authorized translation