An NPD process

All successful businesses have a myriad of business processes they follow, some almost religiously They range from human resource management (e.g., salary determination, promotion assessment, career moves, hiring) to accounting and invoicing to supply chain management. Three major categories of processes are:3

MANAGEMENT PROCESSES: those governing the operation of the business. These include corporate governance and strategic management.

OPERATIONAL PROCESSES: those constituting the core daily operation of the business. Examples include purchasing, manufacturing, marketing, and sales.

SUPPORT PROCESSES: those supporting the core daily operational processes.

Examples include accounting, recruitment, clerical, and technical support.

There are a host of subprocesses under each major category such as purchasing under Operation Processes and recruitment under Support Processes. The list of possible subprocesses can, of course, be quite long depending on the size and complexity of the business. One process that should be on this list is new product development. This is needed just as much as any other process any business follows. A new product development process is unique and cannot be subsumed under the three listed above. It may be guided by management, but it is not a management process per se. It is not operational process - which is concerned with daily activity, not long-term issues - and new product development is certainly a long-term issue. It is not support which is actually concerned, in part, with existing products through, say, fulfillment houses, call centers, repair centers, and warranty claim centers. So new product development is separate but as equally important as the other three.

Like all processes, a new product development NPD process is complex, with many interacting components, each of which is complex in its own right. A large number of books deal with NPD processes. See, for example, Cooper [ 19931, Cooper [20171, and Trott [2017] for some excellent discussions. The classic treatment is Urban and Hauser [1980]. Also see the blog article by Thomas Davenport4. This last is particularly important because he outlines five stages for an NPD process: ideation, business case, create, test, and launch. This is a good paradigm because it emphasizes the major components of a process. I disagree, however, with a few stages. The business case stage is not a separate one, but is an ongoing check at all stages of the NPD process as I mentioned above. It is a gate-keeping function that allows a new product to enter the pipeline and to pass from one stage to the next or be terminated. The launch is not the last stage since once a product is out the door, it must be tracked to see how well it performs and if it is meeting the financial and market targets. If not, then it must be withdrawn from the market. Some may argue that this tracking is not applicable to new products because once the product has been launched, it is no longer “new.” It has been “realized” and a new function or team within the business now has ownership of and responsibility for the product. The tracking process, however, is not just for determining how well the product is performing post-launch, but also for determining what is wrong with it or what else may be needed for the next “new” product. In short, a tracking function has two purposes:

  • 1. monitoring performance of the product post-launch; and
  • 2. identifying new opportunities for the next new product.

This second function is, in fact, an input into the beginning of the NPD process: ideation or concept formation. In short, the NPD process is not linear, progressing from one stage to the next in an almost mechanical fashion allowing, of course, for the business case gate keeping, but rather it is cyclical with the end feeding back to the beginning of the next cycle. In fact, even at a point midway through the process, the product could be “sent back to the drawing board” and redone. Testing, for example, could indicate major design flaws that could terminate the product and force it back into design mode.

My version of the NPD process consists of five stages, each one designed to address a logical question:

IDEATION: “ What do we do?”

DEVELOP: “How do we do it?”

TEST: “ Will it work and sell?”

LAUNCH: “ What is the right marketing mix?”

TRACK: “Did we succeed?”

with the third stage, Testing, and the last stage, Tracking, cycling back to the beginning Ideation stage. This iVPD process is illustrated in Figure 1.3.

The heart of the NPD process

The NPD process is a decision-making one with key decisions at each stage. These decisions must pass muster by going through various business case gates and overcoming hurdles as established by the executive management. Like all decisions, their quality depends on the input into that decision. That input is information based on data. At the heart of the NPD process is Deep Data Analytics (DDA), a paradigm

This illustrates the five stages of the NPD process

FIGURE 1.3 This illustrates the five stages of the NPD process. Although the figure appears to be linear, it is actually a cyclical process with the results of the Tracking and Testing stages feeding back into the Ideation stage as emphasized by the arrows.

Deep analytics, a paradigm for converting raw data into actionable, insightful, and useful information consists of three components as illustrated here. This model is based on Wong [2012]

FIGURE 1.4 Deep analytics, a paradigm for converting raw data into actionable, insightful, and useful information consists of three components as illustrated here. This model is based on Wong [2012].

for converting raw data into actionable, insightful, and useful information. Deep Data Analytics consists of three components illustrated in Figure 1.4:

  • • Data Management;
  • • Interdepartmental Collaboration; and an
  • • Analytical Bridge.

Data are often maintained or housed in different settings as simple as personal computer files and as complex as data stores, data warehouses, data lakes, and data marts. Since they are maintained in various types of settings, a method must be devised to manage them so that everyone involved in the NPD process can access and use them to extract information. So one important component of Deep Data Analytics is Data Management. Data Management involves the organization, cleansing, and distribution of data throughout the enterprise.

More often than not in large organizations, different groups have data that “belong” to them and so are maintained in their personal silos. This is unfortunate since the larger organization, that is the company, suffers from a suboptimal, inefficient use of data. The goals of the larger organization are better served if organizational units share and collaborate. Collaboration across multiple disciplines and functional units, especially when repeated and well orchestrated by a skilled management team, can reduce the costs of analytical work in new product development. As noted by The Economist magazine5:

The characteristics of information - be it software, text or even biotech research - make it an economically obvious thing to share. It is a “non-rival” good: i.e., your use of it does not intefere with my use. Better still, there are network effects: i.e., the more people who use it, the more useful it is to any individual user. Best of all, the existence of the internet means that the costs of sharing are remarkably low. The cost of distribution is negligible, and co-ordination is easy because people can easily find others with similar goals and can contribute when convenient.

A multicollaborative NPD process involves multiple organizations. Figure 1.5 illustrates some possibilities.

Data are not information. They are the raw building blocks for information. They are raw, unfiltered, disorganized pieces of material (i.e., “stuff”) that, like clay bricks, you can arrange and assemble in infinite ways reflecting your creativity and questions. Information is contained in the raw data bricks and so must be extracted. This information is not a binary concept meaning that you either have information or you do not. This is how most people view information. Information is better viewed as a continuum as illustrated in Figure 1.6. Rich Information is insight built and extracted from data by creatively manipulating data bricks while Poor Information, in the form of simple means and proportions, can also be extracted, but provides limited insight. You leave a lot of other information in the data when you extract Poor Information. The goal for data analysis should be to minimize the amount of information left in the data. The Analytical Bridge is the process of taking raw data bricks and extracting the Rich Information contained in the data by using different data views. Deep Data Analytics consists of the tools and methodologies

Numerous functional areas contribute to the NPD process. Part of this collaboration should be data sharing

FIGURE 1.5 Numerous functional areas contribute to the NPD process. Part of this collaboration should be data sharing.

Information is not binary, but continuous. At one end is Poor Information while Rich Information is at the other end. It is the Rich Information that provides insight for actionable decision making

FIGURE 1.6 Information is not binary, but continuous. At one end is Poor Information while Rich Information is at the other end. It is the Rich Information that provides insight for actionable decision making.

There are two major components to the Analytical Bridge for extracting information from raw data for new product development

FIGURE 1.7 There are two major components to the Analytical Bridge for extracting information from raw data for new product development: Market Research and Big Data Analytics. These components or functions can certainly be used for other purposes in a business so they are not restricted to just new product development.

of statistics, econometrics, machine learning, and text analysis that are designed to extract the Rich Information. Analytics that are not advanced, that are based on simple means and proportions, extract Poor Information. See Paczkowski [2016] and Paczkowski 12018 ] for discussions of some methodologies.

The Analytical Bridge component involves two subordinate components:

  • 1. Market Research; and
  • 2. Business Analytics.

These are illustrated in Figure 1.7. They should be considered as frameworks for Deep Analytics since neither is an analytical method per se. Subcomponents of each framework, however, are analytical methods, a few of which are listed in Figure 1.7. I briefly discuss these in the following subsections.

 
Source
< Prev   CONTENTS   Source   Next >