Improving the product development process

Hewlett-Packard Journal, June, 1991 by Spencer B. Graves, Carmichael William P., Douglas Daetz, Edith Wilson

MANAGERS IN MARKETING, manufacturing, and especially research and development at HP are becoming more aware that they jointly manage a cross-functional process. Their people define and design a product and develop processes to manufacture and market that product (see Fig. 1).

Many HP divisions are working to improve this process. Their improvement efforts rely on concepts such as break-even time (BET), post-introduction product reviews, in-process project retrospective reviews, and quality function deployment (QFD). This paper briefly describes these techniques. Improvement efforts in the R&D departments of other organizations are described in references 1, 2, and 3.

Break-Even Time

Break-even time, or BET, has been identified within HP as a primary metric for the development process. A simple definition of BET is that it is the time from the initiation of a project to the point when all the design and startup costs have been recovered and the project is generating a net profit. Illustrative calculations are presented in Fig. 2. BET is a reasonable measure because it includes market acceptance, cost of production and sales, selling price, time value of money, and startup costs. Partial support for using a concept like BET in other companies can be found in references 4 and 5.

Measures such as break-even time may ultimately improve management's ability to predict profitability as early as initial product concept. (*) The data necessary to attempt to improve forecasts of profitability would be similar to PIMS (Profit Impact of Marketing Strategy). [7] The PIMS research program was initiated in 1972 for the specific purpose of determining how key dimensions of strategy affect profitability and growth. Since that time, 450 corporations have contributed data. PIMS data supports part of the famous Deming [8] chain reaction: higher quality tends to improve market share, to enhance ability to charge a higher price, and to decrease costs, all of which contribute to higher profit. By combining BET data with the results of post-introduction product reviews (see below), managers should be able to identify things they need to do to build on previous experience (like PIMs) and improve their ability to forecast profitability.

Break-even time is only one measure, and is evaluated in combination with other measures such as the estimated net present value of a subproduct line under various future investment assumptions. BET alone could push managers to do the wrong thing in some cases because it does not consider leverage or product success beyond the initial payback period. BET will not properly evaluate a product that is expected to open up a new market and whose ultimate success depends on whether future products are able to leverage that entry for higher profitability. BET also does not consider the cannibalism of a successful product by a follow-on. In some cases, the organization can protect market share by introducing a follow-on product early. In other cases the early introduction of a follow-on product may

[TABULAR DATA OMITTED]

suggest a misdirection of R&D resources. No measure is perfect. For example, BET cannot be calculated if the project is canceled or if the product never recovers costs. To correct for this, some entities in HP compute a divisionwide BET that includes canceled and unprofitable development efforts. Throughout HP, BET is being used to improve the R&D process within a division. Some of the possible causes of poor break-even time are summarized in Fig. 3.

To complement the BET metric, tools are needed for evaluating the strengths and weaknesses of the management of a project. Two such diagnostic tools that are intended to help evaluate and improve project management are post-introduction product reviews and in-process project retrospective reviews.

Post-Introduction Product Reviews

Kaoru Ishikawa [9] said, "Whenever I am asked to help introduce a total quality control program to a company, I choose for my case study one of the company's new product development projects that is full of problems." Ishikawa found that his clients learned a lot by studying their own experience in a structured way -- and he provided the structure. The structure is essential to the learning. As Deming [10] explains: "Experience without theory teaches nothing. In fact, experience cannot even be recorded unless there is some theory, however crude, that leads to a hypothesis and a system by which to catalog observations."

Post-introduction product reviews and project retrospective reviews are vehicles for developing and improving the theoretical understanding that R&D managers use to guide their work. A study of this nature is being conducted at HP's Corporate Engineering. Ten factors have been identified that distinguish successful from unsuccessful projects (i.e., projects that met their initially stated objectives and those that didn't). Fig. 4 lists these ten factors and shows a comparison of six successful and six unsuccessful projects.