Bulletin 105, Mai 99 

The acceptance of Operational Research and its success in an organisation depends on many factors. These factors include: successful internal marketing; a dynamic team leader; good Operational Research staff to support him; the selection of projects; positive relationships between the Operational Research team and the organisation's management; positive cost-benefits to the organisation from implemented results.

This article briefly explores a way of selecting projects. I shall assume that you are commencing an Operational Research Group. This can come about in one of two ways. Firstly, you are on the staff of an organisation and you have been asked to start an Operational Research Group by senior management (perhaps as a result of your own efforts). Secondly, you have been appointed to the position of Operational Research manager from outside.

In both situations there is one common element and that is that you have the support of senior management. This is a vital ingredient and should be nurtured all the time you are the manager.

So, how do you go about developing Operational Research in the organisation? You need to develop an awareness programme and then select key managers and talk to them. You have to convince them of the role that Operational Research can play in their Department and the benefits of using it. A good part of this is getting the managers to talk about their work and what they see as the hot issues that they have to resolve. Indeed, some of these issues may seem to you to be mundane and not fully worthy of being an Operational Research 'problem'.

Nevertheless, you need to get results and you should be able to demonstrate some positive benefits after a 'quick and dirty' study. Don't worry if the models are simple and can be done on the back of an envelope. With some positive results behind you the marketing becomes easier.

Don't try to start with a big interesting project - you can do those later. The senior management will not support you forever.

Once you have the confidence of all the managers you will find that they begin to make requests for your help. This is where you begin to make a large impact on the running of the organisation. Very quickly you will be in a position of having more potential work than you can handle. In this pleasant situation you have to select which projects you undertake. You won't be able to satisfy everybody so you need to develop a transparent way of prioritising the requests. You must not select on an arbitrary basis like 'this sounds interesting with lots of nice theory and model building and it might even benefit the organisation'.

Typically a structured procedure would include the following elements being agreed in plain and simple language for each request:

• The definition and aims of the project. What it is that will be resolved by the project and how it will be achieved.
• List the expected results. What are the likely results of the project?
• The timescales for the project. How long will it take to resolve the problem and what resources will be needed to accomplish it?
• List the expected benefits. What will be the expected benefits to the organisation if the results of the study were implemented?
• Cost of project. What is the cost of undertaking the project?
• Value of benefits in first year. What is the value to the organisation of the benefits in the first year after implementation? Only the tangible benefits should be included in the costing.
• The likely cost of implementation. What resources will be used during implementation and what would they cost the organisation.

To prioritise the projects choose a simple payback measure such as the ratio of first year's expected benefit to the sum of cost of project and cost of implementation. This measure can then assist in the selection process. One organisation received so many requests for Operational Research projects that they would only consider projects with a payback of over 100:1.

Of course, as you work on a project more information comes to light and you should re-assess the elements above on a regular basis. There may come a time when you have to make a difficult decision to discontinue a project. If you have maintained the information then it will be easier to make that decision because other future projects will have a significantly higher payback. This decision should only be taken with the agreement of senior management and should take into account the amount of work already expended on the project.

Finally, one of the evaluation points of the whole project will be one year after implementation. Agree with the manager concerned what benefits have actually been achieved and were they what were expected. Were the actual values of the benefits fully achieved? This will enable you to see how accurate the whole process has been and how you cam improve the estimating process.

At the end of the day, successful projects with demonstrable benefits can only benefit the organisation and of course will show the true value of Operational Research.

John Wilkes
Le Romandie Dom. Résidence
1884 Villars-sur-Ollon VD
Tel: +41 24 495 16 15
Fax: +41 24 495 16 15
wilkes.infor@com.mcnet.ch