Setting up a Company I.C.

Transcribed from the printed issue; not yet reviewed. Read it in the PDF of the issue, page 42

Definitions and Objectives of an Information Centre (IC)¶

A straw poll of the audience revealed that 10 out of 80 felt they knew what an Information Centre was and 12 out of 80 felt their view was based on the IBM concept!

Romilly Cocking began with a couple of rather cynical definitions of an IC, namely:-

“The best sales aid IBM has developed to date”

and

“The DP department pacifier for irate users” — to distract their attention from applications that are late, wrong or way above budget.

A more positive definition put forward the purpose of an IC as providing the hardware, software and support to enable all employees to satisfy their information needs by being able to:-

a) get at information

and

b) use it

without becoming DP professionals.

Romilly Cocking (RC) stressed the need to keep in view the benefits to the organisation of providing information, which may be one or more of the following:-

a) To help directly or indirectly to increase sales

b) To help directly or indirectly to reduce costs

c) To improve cash flow

d) To meet legal or corporate requirements

The IC’s contribution to these objectives would be in the form of:-

a) better decisions

b) better control

c) better planning

and in helping communicate these decisions to people and justifying them.

The IC may also contribute to greater efficiency by enabling fewer people to do the same work.

Justification for an IC¶

The biggest problem, when deciding whether to set up an IC is that of cost justifying it in numeric terms. It is impossible in many cases to demonstrate savings. RC proposed that it should be viewed as going into a separate business, with the associated risks and suggested taking a good look at the experience of other companies before doing so.

Once established, it should be operated as a business with a separate cost centre from the rest of the DP department, paying its way within the organisation as a bureau service.

IC Personnel¶

RC stressed that the IC would be a service business, with people as its key resource, not the hardware. The level of staffing is an area of contention and the audience were invited to make suggestions as to an effective ratio of support staff to users. Opinions were varied, ranging from 1:4 up to 1:70. IBM apparently recommends a ratio of between 1:30 and 1:100 which in RC’s opinion, would be very unsatisfactory if trying to balance desirability against practicality. A more realistic and effective ratio would be somewhere between 1:5 and 1:20. Anywhere below that would be merely providing access to hardware, rather than providing an Information Centre.

Of the personal skills required within an IC, he suggested that some people would need to be salesmen, others highly skilled technically and all should be good at handling people. Furthermore, experience in the form of a “veteran” from a bureau is almost a necessity.

Software¶

Assuming that the quality and level of staffing is right, the next most important factor is software. Great care must be taken in choosing software that is easy to use and allows users to get at corporate data and process it. It is difficult to change products once users have become familiar with them. They will be reluctant to re-train and therefore a long term view should be taken when selecting software.

On the role of APL in the IC, Romilly Cocking’s view on VS APL (as opposed to other APL’s) was that it was adequate but not ideal due mainly to its restricted file handling capabilities. APL on its own would not meet all of the information requirements and has proven not to be an end user language. The majority of end users don’t want to learn it and favour the “point and grunt” approach which ADRS and ADI have attempted to provide. In RC’s opinion, he doubts that APL is ideal for spreadsheet type applications.

He had little to say on other software products, but mentioned the “IBM approved” packages Wizard and Intellect as looking good and being purchased by several companies.

Hardware¶

RC emphasised the benefits of having a completely separate mainframe from that used to process the operational systems. He warned that CPU usage would grow rapidly and that there would inevitably be a battle over charges if a machine were shared. This would result in a clamp down on usage by the IC and the consequent need to turn away customers. Costing would be much easier on an independent machine.

On the question of size, he warned that anything smaller than, say, an IBM 4341/G2 would soon be inadequate.

VM would be preferable to TSO and CMS should support a larger user base.

As well as a mainframe, the IC should support a network of VDU’s and PC’s with provision for hooking up PC’s with the mainframe. Transferral of data between PC and M/F as yet presents a problem, with various partial solutions but no satisfactory universal one.

Operating the Information Centre¶

Education, consultancy and support are among the responsibilities of an IC and Romilly Cocking suggested issuing a newsletter, arranging seminars and setting up a help desk as just some ways of meeting them and making the IC more effective. IC staff should spend time with users, find out how they operate, since it may not be obvious to them how the IC can help. The IC may become involved in developing software with users, which RC says is a good thing. If users are encouraged to develop systems entirely independently, they will usually either fail completely of if they are successful, will be tempted to tackle larger systems and will inevitably meet and try to solve the same problems that DP met and overcame years ago. IBM maintain that the IC should not become involved with development, but one of the original reasons for setting up an IC was to help clear the backlog of systems waiting to be developed and implemented. The IC can and should take on some of the small developments and complete them more rapidly.

Summary¶

RC summed up by stating that the IC concept has worked and that is possible to identify specific projects which have been cost effective. It gives management a tighter grip on individual developments and ensures that direct savings can be achieved.

If the IC is set up properly, there will be rapid growth and it is essential to sort out the charging system right from the start. It is much easier for the IC to establish whether and how it is contributing if each project is costed out. Therefore, when sales revenue grows rapidly, there will be ample justification for expanding its resources.

The talk ended with the questions “Is the IC a revolution in computing and should it be staffed from outside DP?” RC maintained that there are often revolutions in mainstream DP and that the Information centre needs the continuing skills from it.