The hottest use of JMP and selective design to imp

  • Detail

Using JMP and selective design to improve product competitiveness from the source

many years ago, someone realized that quality is not tested, but manufactured. In other words, the quality of the product comes from the process of producing the product itself. Any additional inspection, testing and other processes cannot fundamentally change its quality level. Inspection is just to reveal the current quality level. Therefore, the pursuit of the ability to produce products in accordance with design standards in line with the production and cost objectives has become the quality consensus of many people

gradually, we will find a new problem: once the product has defects in the initial design, no matter how perfect the manufacturing process is, it is impossible to completely solve the quality problem. At this time, we must redesign the product to make a breakthrough. Therefore, a new point of view came into being: quality is designed. Only by doing well in design and R & D can we completely solve the quality problem of products

this view is undoubtedly correct. But it is worth mentioning that there is a more important thing to do before the expensive new product design and R & D process begins. That is, first of all, we must be able to accurately obtain customers' needs and preferences! Because today's market is developing and changing rapidly, and customers' concepts and ideas are also changing with each passing day, it is very important to quickly determine the key product attributes. These important product attributes can help enterprises grasp the pulse of the rapidly changing market, concentrate limited enterprise resources on the focus projects with market potential, and design and manufacture new products with a clear aim, so that once the products are listed, they can win the favor of consumers, occupy a favorable position in the increasingly fierce market competition, and achieve the goal of profit growth as much as possible. Avoid the embarrassing situation that the enterprise has designed and produced high-quality products, but it is still "applauded but not popular"

in the design and manufacturing stage, the analysis of product quality problems can be completed with many experimental design DOE methods (such as full factorial design, response surface design, etc.), so whether similar experimental design methods are also needed to analyze customers' needs and preferences, which can accurately analyze customers' preferences for products and optimize production requires the design attributes of a high-temperature furnace, And it can greatly save the cost of market research and analysis? The answer is yes. Choice design is the most representative one. It can help enterprises determine the importance of different characteristics of products to their target market, quantify the value recognized by potential customers for various possible products, and explore the best combination after weighing all product characteristics. The following will be analyzed with a case (among them, the realization and analysis of the selected design scheme are mainly realized by JMP software. JMP software is one of the most professional software in the field of experimental design application in the world, and it is also one of the few desktop analysis software that can support the selection of design scheme at present. JMP is the outstanding performance system of SAS, the world's largest statistical software group. Insiders pointed out that in recent years, the treatment and recycling of waste plastics are very important in the application field Extensive, including business visualization, data discovery, experimental design, six sigma and process improvement, R & D innovation, etc., which can be accessed for details), to illustrate the practical application of selective design in improving product competitiveness from the design stage

for example, a computer company plans to launch a new high-end notebook computer. Before developing and producing specific products, we need to determine many laptop features that are important to customers, such as hard disk size, main frequency speed, battery time and price. How to quantify the relative importance of each feature? Many people will think of using traditional experimental design methods, such as full factorial design, to generate a table as shown in Figure 1, and use it to do a market survey. Then, based on the results of this market survey, the purpose of quantitative analysis of customer needs is achieved

Figure 1 market questionnaire based on complete factorial design

at first glance, I think this method is very reasonable. However, after careful consideration, we will find that there are many hidden dangers in this method. First, in the case of market research, it is very difficult for customers to determine "which is the best, which is the second best, which is the third best, and which is the worst" in order among the 16 options. Decisions made in a short time often do not reflect their real ideas. Second, the advantages and disadvantages of some schemes are obvious, because most people prefer notebook computers with larger hard disk, faster speed, longer battery life and lower price, so schemes such as "hard disk size =40gb, main frequency speed =1.5ghz, battery time =4hours, price = ¥ 8000" are not feasible at all, and do not need to be considered by customers. Such schemes should be avoided, or at least as few as possible. Third, although the traditional experimental design and analysis can quantify the relative importance of each feature, it cannot answer such questions as "how much are customers willing to spend on their favorite computer features? Are customers willing to bear the price rise caused by product upgrading? When favorite computer features conflict with each other, what kind of compromise plan are customers most likely to accept?" And other deep-seated problems

at this time, the choice of design has a place to play, which can skillfully solve the above problems. Similarly, taking the above case as an example, choosing a design can provide a market survey form as shown in Figure 2. It can be seen from the table that the experimental design scheme has a total of 8 selection sets, each of which has two different alternatives, and the respondents need to "choose one from two" as their final choice. In other words, customers only need to do eight simple multiple-choice questions to complete a market survey. This has greatly reduced the difficulty of implementation for both the organizers (enterprises) of market research and the participants (customers) of market research

Figure 2 is a market questionnaire designed based on JMP selection

and after obtaining the survey results of customers (as shown in Figure 3, "1" means like, "0" means dislike), a detailed analysis report with pictures and texts can be obtained with JMP, as shown in Figure 4. From the report, we can clearly see that "price" is the most significant product feature (its corresponding p value is 0.0236, less than 0.05). In other words, customers are very sensitive to the price of notebook computers. Further observe the prediction profiler prediction profiler at the bottom of the report. When the abscissa level of the rightmost factor "price" is adjusted from 6000 yuan to 8000 yuan, the ordinate value of the output response "utility utility" (one of the commonly used concepts in economics, which refers to a measure of the satisfaction of consumers' needs, desires, etc. through consumption or leisure) decreases from 1.424 to -0.393. In other words, for every 2000 yuan increase in the price of notebook computers, the utility will decrease by about 1.817 (=1.414- (-0.393)) units. Therefore, it can be considered that the unit utility cost is about 1100 (=2000/1.817) yuan

this "unit utility cost" is very important. On this basis, with the help of JMP, we can easily calculate the value that the upgrading of different product features will bring to customers. Taking "dominant frequency speed" as an example, observe the predictive profiler again. When the abscissa level of the second factor "dominant frequency speed" is adjusted from 1.5GHz to 2.0GHz, the ordinate value of the output response "utility utility" rises from 0.938 to 1.424. Therefore, it can be considered that when upgrading from 1.5GHz laptop to 2.0GHz laptop, customers think the acceptable price increase is about 535 (=11 outlet guide vanes (OGV) have been widely used on cold end components of foreign advanced aero-engines 00* (1..938)) yuan. Such results undoubtedly provide valuable data basis for the reasonable pricing of different notebook computer products

the market questionnaire completed in Figure 3

the analysis report selected in Figure 4

of course, for the sake of illustration, the above case only considers the simplest case. In fact, we can also further optimize our market survey questionnaire by adding factor variables that reflect the social behavior characteristics of customers (such as gender, region, age level, income level, etc.) and increasing the number of respondents in the early planning of selection design, so as to carry out more in-depth statistical analysis (such as the size of interaction between different characteristics, the regression model of utility value, etc.), So as to obtain more beneficial business value

with a little reference to the hr15t test, it is not difficult to find that the selection of design is very helpful for many tasks, such as new product planning, competitor analysis, product pricing, market segmentation, brand positioning, key setting of advertising and promotional means. For enterprises implementing Six Sigma management, choosing design and JMP software provides another practical application platform for the core concept of "focusing on VOC (customer voice)". It has laid a solid foundation for us to "design high-quality products from the source and enhance the market competitiveness of products". (end)

Copyright © 2011 JIN SHI