“Many companies have difficulty justifying fundamental work”

Types of Work Undertaken by R&D

Traditionally R&D has been described as passing through the phases of pure research, applied research, development, testing, engineering and so through to production. This model has been widely adhered to even though it is widely agreed to be wrong. Under the tenets of third generation R&D, work is classified according to its strategic significance.

• Incremental R&D is work intended to bring about a definable modest improvement in a particular product or a process
• Radical R&D is work which promises to produce a new product or process
• Fundamental R&D is work which is not intended to produce a specific business benefit, but is intended to produce scientific and technical information; this information might for example lead to a decision to try to develop a new line of business

In addition to these, there is:

• Compliance work, which is undertaken to meet environmental or other legislative requirements
• Customer support work, in which R&D staff work on the problems of existing products

A common pattern of progress is that a product line or process is improved by many incremental developments over the years, until improvements become harder to find. It often then happens that a radical development replaces that product or process by a new one, which then enters into its own series of incremental improvements. From time to time a fundamental change renders the whole product line or process type obsolete.

For incremental work, the probability of success is fairly high, say seventy per cent, and the rewards are modest but reasonably easy to estimate. For radical work the chances of success are low, say twenty per cent, and the possible rewards are high but less predictable. Many companies have difficulty justifying fundamental work; its saving grace is that often it not very expensive. Incremental work enables companies to compete on grounds of product performance, price and profit. When radical change occurs companies can win or lose a dominant position in an industry. When fundamental change occurs whole industries are likely to disappear or arise.

“The parallel with share options”

Financial Appraisal of R&D

In the recession years R&D directors were asked to justify the cost of their departments in financial terms such as return on investment. Many were not able to do this, and the result was closed or reduced R&D departments. The classification described above makes it easier to see how R&D can be assessed, by breaking the work up into its components.

• Compliance work is a necessary cost of continuing in business, so it is an overhead, and the only requirement is that compliance is achieved at as low a cost as possible.
• Customer support work is done to assist production or marketing. It is a cost of sales, even though the work is done by people employed in the R&D department, and their time should be booked to the appropriate function.
• Incremental work can be assessed in terms of the return on investment, modified by the risk involved. Since many incremental projects are undertaken, costs and benefits can be estimated in the light of past experience so long as the costs and outcomes of past projects have been recorded. Incremental work should be paid for by the product division that benefits, who should be a willing purchaser.
• Radical work may be radical to a product division or radical for the company as a whole and should be commissioned and paid for by the division or the company as appropriate.
• Fundamental work is paid for in most companies by the centre: the cost of gaining new scientific information which can be used in strategic appraisal and which may lead to new business project ideas, is an overhead of the business.

It can be seen that the total spend on R&D is not globally determined, but is the sum of the spends by the different types of customer for the different types of work.

A weakness of attempts to justify R&D financially is that the whole of the process of exploiting R&D output is not under the control of the R&D department. A perfectly good development may not proceed to manufacture because of a company decision to manufacture something else. An R&D investment has the character that if it is successful, ten or a hundred times as much money is likely to be needed for investment in production startup and launch, as was spent on the R&D. An R&D project will either fail, so that the cost must be written off, or present the company with the option of spending a great deal more. The parallel with share options has been noticed, and it is argued that what the company buys with its R&D spend is the option to be able to invest more later if it chooses. Newton and Pearson, [1994] have put option pricing of R&D on a mathematical basis.

“efficiency seems a less relevant concept”

Defining the R&D task

No general guidance can be given on managing R&D because the work to be done varies greatly with the type of company and its circumstances. Some of the work is reactive to events, but most is concerned with creating the future, and this must necessarily start with decisions about where the company wants to go and how much it wants to invest in trying to create technology that it will be able to own and use to the exclusion of others. Management in this field must therefore start with strategy, developed with input from R&D about what is likely to be achievable.

Traditional management thinking derives mainly from the management of manufacturing enterprises, where the objectives are cost saving, efficiency and quality. Attempts to apply these ideas to R&D have run into difficulties. To measure efficiency people have tried to find ways to measure R&D output, but none have been very convincing. There have been attempts to benchmark one company’s R&D performance against another, but unless the companies are very similar it proves to be difficult to make comparisons.

In the case of pharmaceutical companies, each major company is identified with one major drug. The drug hunt consists of synthesizing and screening thousands of possible molecules, selected by scientific hunch, to find one that will secure the company’s future. Glaxo-Wellcome has committed itself to the automation of this drug search. In the case of aero engines, Rolls Royce has had to decide which low pollution combustor possibility to commit its development to, and whether to commit itself wholly to computer-aided design and manufacture of its range of about twenty engine types. In these situations successful decision making and implementation are critical, and efficiency seems a less relevant concept.

In the case of car development or household goods development, the number of small improvements or new products is so large that the efficiency of the new product process becomes the dominating consideration.