Written by David Tebbutt, Director 01/91 - scanned


When the computer was first invented it was used for very specific, scientific, tasks. Systems were designed by the scientists, who knew what they wanted, and the electronic engineers, who knew how the computer worked. By the time computers started being used in business, this cosy arrangement had become impractical and companies had to start employing computer programmers to implement their business systems on the computer.

Through the fifties, sixties and seventies, the programmers became the high priests of computing. They understood the computers very well and could make them perform the most amazing tricks. The trouble is that these tricks weren't necessarily of any benefit to the company. A huge communication gap existed between programmers and company management. Many companies and manufacturers spotted the problem and thus was born a new `breed of person who supposedly understood both management and programmers. This person was the systems analyst. This was a great idea in theory but, in practice, it merely gave the programmer an opportunity to earn more money. The programmer/analyst was born.

Before long, data processing departments appeared, headed by a DP manager and containing a mix of analysts, programmers and operators. Some were very good, but most were headed by ex-programmers whose primary understanding was of the computer system, rather than of business. For years they tended to tell management why certain jobs were `impossible' and why urgent amendments to existing systems couldn't be started for `nine months at least'.

When the personal computer arrived towards the end of the `70s, managers and users of the central computing services saw a chance to break free of the DP department's stranglehold. At last they could have computer power on their own desks to help them with their own work. Packaged programs such as spreadsheets and databases opened up opportunities for local processing beyond their wildest dreams. This combination of personal leverage and freedom from the tyranny of the central computing function, drove millions into the arms of the PC sellers.

During the second half of the eighties many companies had realised that isolated computers on individual desks were no longer the most effective use of personal computers. They started to be networked together. At first this was done to share expensive resources such as laser printers and high capacity disk drives. At first, the drives were divided up in such a way that each user had their own chunk of storage and only shared information with others of the network by copying files from one user's storage area to another.

The opportunities for genuine sharing were there, it's just that the software writers had to do a bit of catching up. Before long work groups were able to share common files which meant that all members of a group could be working on current information. Other developments included electronic mail, which enabled all participants in a network to send messages to each other. The advantages of this are that (1) the message is held electronically so later retrieval is instant and (2) unlike a telephone call, messages can be sent even when the recipient is not at the computer. If the network has a connection to the outside world, then it is possible for users to exchange messages with other computer users on other networks.

The result of all this activity is that users have grasped control of their computer systems. Sometimes they have an `information centre' within the company to support these personal computer users. The data processing department quite often still has control of the central mainframe or minicomputer. And senior management still has little control of the computing systems which have cost the company so much.

So what happens next? Well, we are on the verge of a completely new era in computing. We have a situation where companies have many `islands' of computing. These are the individual networks with their associated information bases. They are the mainframe and minicomputers which are, in most cases, still quite separate from the personal computer networks. And we have a recession upon us, in which companies need to compete more effectively to grab market share.

As long as the computing systems are used to satisfy local needs, at the departmental level or according to the wishes of the computing department, they are not going to make much difference to the company's success in the outside world. All that computing power is sitting there waiting to be harnessed to the- furtherance of the company's greater goals. It doesn't have to be used simply to speed up traditional tasks.

The Leeds Building Society knows this. It appointed the head of computing to the board and the managing director committed 20 per cent of his time to ensuring that the Leeds information management system furthered the Society's corporate objectives. Once building societies found themselves competing with other financial institutions they had to find ways of attracting and keeping customers. The most obvious thing to do was to improve the level of service to the customer. This directed the Society's eyes away from the central mainframe computer and out towards what could be done at branch level.

By this simple shift of emphasis away from the central computer serving the society and towards the society using computers to better serve the customers, a new approach to computing has been implemented. The branches have personal computers connected to a `backoffice' computer which is, in turn, linked to the society's central computers. The personal computers at branch level can be used for tasks such as word processing, for producing on-the-spot insurance quotations, and for processing mortgage applications. The Leeds is currently testing a system which deals with a mortgage application in a single visit, a task which used to take 15 days before branch computing was implemented.

Management cannot go it alone in defining computer systems. It still needs the help and active commitment of the computing experts within the company. The difference is that now management should feel perfectly within its rights to demand systems which do more than mimic previously manual tasks. According to KPMG Peat Marwick McLintock, personal computers cost businesses between three and four thousand pounds a year each. If they are connected, then the annual running cost rises to between six and eight thousand pounds.

Now count up the personal computers in your company and ask whether it's getting the payback on these machines that it ought to. Now that the limitations of the rigidly centralised computerisation of the DP department and the ad hoc decentralisation of the personal computer era have become clear, it is time for management to regain the initiative it lost so many years ago.