At a time when the chemical industry has made unprecedented investments in improving product and process development and similarly is beginning to improve product strategy and basic technology development, the need to measure the effect of such improvements has never been greater. There is hardly a chemical company I have visited that has not planned or installed a structured set of policies, protocols, or methodologies to improve product and process development. Some companies have done it more than once, not satisfied with initial efforts. Some are in an active mode of continuous improvement. Others are taking what they have learned from structuring product development and applying similar ideas and discipline to new business initiatives, applied research, or manufacturing process improvements. Still others, not happy with what they have done, are trying other ways of organizing their efforts (2, 3).
But how well are all these efforts going? How do individual companies measure up? And against what yardstick should they measure themselves? Without sensible indicators of "success" (or worse, no indicators at all), do they risk finding themselves in the position of the cimetidine researchers: working hard but incapable of knowing whether they are succeeding?
Measuring R&D productivity is notoriously difficult. Projects are hard to categorize reliably; there is a huge disparity in scale and complexity from the simplest to the most ambitious. Development of a major new chemical process or discovery of a useful new molecule and its properties can take years, sometimes even decades. Simple reformulation or applications development projects may take only weeks. There are so many differences in project objectives that it seems daunting to try to compare them. Good data are hard to acquire. Some companies, believing they can't get the numbers they need, don't even try.
The relationship between effort expended and results achieved seems impossible to sort out at times. So many intervening factors affect project success that one wonders how to interpret the data anyway. If R&D is not apparently "productive" by some measure, is it because of poor research, nearsighted business management, unlucky timing, unreliable funding, or any of a hundred other possible explanations or combinations of explanations?
Measurement and careful interpretation of the results may be difficult, but the effort is worthwhile. When companies compare their practices with good performance at other companies, they can see how they measure up: What is working and should be reinforced? What isn't working and needs improving? What targets should we set for the future? How much can we realistically expect to achieve, and by when?
Here is a list of useful performance measures. It is not exhaustive, but it touches on the major categories of concern: effectiveness, timeliness, productivity, selectivity, good decision making, and strategic fit. One measure is qualitative; the remainder are quantitative. In my opinion, these measures or their functional equivalents should be part of any self-assessment.
Process maturity is the one qualitative measure. Enough companies have implemented improvements to product and process development that a clear pattern has emerged: an implementation order of operations that dictates the sequence in which process improvements are best adopted by a company. One such model used in the chemical industry is the stages model used by DuPont and several other companies for internal assessments (4). If one knows the stage of the division or company, it can be compared with a set of companies at the same stage to see what its current performance ought to be. And knowing what the next stage's characteristic performance is, one can reasonably target the requisite process improvements. This is a useful start.
Among quantitative measures, several stand out as unusually helpful.
The R&D Effectiveness Index (RDEI) measures new product profit impact for every dollar of product development investment. When this measure is carefully applied, paying close attention to the product life cycle and its implications for the definition of what is a new product, one can learn a lot from the results. The measure appears to be increasing industry-wide over time. Even so, our 1995 Product Development Benchmarking Study showed that more than three-quarters of companies had return from new products running at a rate lower than the investment. Companies use RDEI to track improvements over time because it is such a good directional measure.
Development cycle time is an indispensable measure. In the chemical industry, being first to market is not as important as in other industries with very short product life cycles, but having the development process under control is every bit as important. There is no better measure of a development process under control than the ability to predict development cycle time accurately, and there is no better way to increase effective development capacity than by reducing the time it takes to develop products and processes.
Measuring how much development is "wasted" through late cancellation is a revealing quantitative measure. It addresses a component of R&D productivity, because companies that cancel projects early free up proportionately greater resources. It also provides a measure of the decisiveness with which choices are made to redeploy resources, a determinant of R&D success.
The measure of strategic balance is a newly emerging one. It addresses the balance of effort of the entire development portfolio, not simply individual project performance metrics. The key measure is the percentage of effort devoted to new platforms. Businesses are not simply asking their development organizations to develop new products; they're asking them to help renew their businesses with fundamental new advances, platforms that provide the basis for the development of entire product lines and that might lead to the opening of new markets and applications. A company without new platforms will decline no matter how it tries to shore up its position with acquisitions, restructuring, and cost reduction. A good predictive measure of readiness to renew is percentage of effort devoted to new platforms.
If these and similar performance measures are carefully collected and compared with appropriate populations of companies, a company can expect to determine how it measures up. And in the practical spirit of the carpenter's advice to "measure twice and cut once," companies will then be able to get on to the real task at hand: implementing the changes that drive key performance measures to greater and measurable heights.
TO SIDEBAR: Invitation to participate...
References
ChemCenter
Pubs Div. Home Page
Copyright © 1997 by the American Chemical Society.
