The relationship between project management techniques and a matrix management structure is a close one, but not a necessary one. Project management requires that skilled staff are treated as resources, and as such any management system which breaks down traditional lines of command in favour of a more flexible system is likely to be more suitable for project work. "In order to manage projects you need a very different management structure. You will usually end up with matrix management," says Jim Young, who is a partner with consultancy PA.
At BREL the management structure is flatter than in most companies. The company is broken into three divisions, each with a director; beneath this a business manager receives direct reports from project managers, who effectively have access to all the resources in the company. It is not run along departmental lines, but is a matrix of functions and projects.
The role of computerised project management systems is pivotal. A project-based organisation requires that information regarding resources and schedules is always available to the project managers. Computers offer the opportunity to integrate and distribute data almost instantaneously, where previously it was scattered about in departments, each reporting information through the management chain.
According to Ray Palmer of Cheltonian, supplier of the Panorama system, information should be stored in one place, in a centralised system, and it should be stored only once. Earlier generations of project management systems ran on separate computers. But, says Palmer: "There is no reason why you should ever separate your project management system from the rest of the management information systems."
Despite the enthusiasm of the project-based engineering companies, the use of project management systems for business planning and control should not be exaggerated: project management software is selling well, but most of it is being used for specific project work, and is not being used at a high level or integrated with other management systems.
Most software packages sold run on personal computers and, according to Palmer, "with the best will in the world the packages will never form part of the corporate information system". Moreover, it is not clear if many of the systems being installed on PCs are then used properly.
According to Kevin Daley of the National Computing Centre, many companies like what they hear about project management and then buy packages with little understanding of the techniques. They are often under-used and critical information is frequently not entered into the computer.
Computerised project management systems could be likened to the manufacturing resource planning methods and software which have swept through industry. These are widely credited with dramatic achievements in terms of management control, quality and reducing waste and inventory. But there is also strong evidence to suggest that many companies have struggled with the burden of managing the computer systems and adopting the strict disciplines involved.
The role of the different types of project management systems needs to be fully understood. For example, the UK's market-leading product, Hoskyn's Project Manager Workbench, has been used for hundreds of simple projects and is especially good for managing software development. But no one would attempt to run their company with it.
The leading vendors of the high-powered minicomputer and mainframe-based project management software systems, however, fall into a different category. They claim to mimic the classical management hierarchy. At the operational level the computer will produce detailed project information regarding costs and resources; further up, tactical information for use by the project managers can be accessed and managed; and at the highest level strategic overviews are presented.
Using an integrated management information and project management system will not necessarily flatten the organisation, but it should help. "A system can relieve managers from weeks of work sitting at computers going through figures. It frees them from drudgery, increases immediacy and at the end of the month the system will tell you whether you are winning or losing," says Palmer.
The promise is great, but as with all new methods and systems, no sizeable organisation should be under illusions about what it could cost, both financially and in terms of management effort. Effective project management is easily understood but difficult to achieve. Introducing these systems is a large project in itself - which is probably one reason why it is the engineering companies which are leading the way.
(Andrew Lawrence is a freelance journalist.)