Although Sherlock Holmes is not available for decision support consulting, he would have had no trouble finding engagements with manufacturing companies. In fact, if the great detective were to expound on why the challenges can be great, he would have answered a set of key questions, answered in the following piece. These questions include:
- What segments exist in manufacturing and how do they differ?
- How has the use of technology evolved for manufacturers?
- What are some key kinds of decision support applications for this sector?
- What challenges do manufacturers face integrating best practices decision support?
Manufacturing: Commercial and Consumer
In the consumer supply chain, manufacturing plays two roles. First, commercial manufacturing must extract resources and refine them. They sell these refined ingredients to consumer manufacturers. Next, consumer manufacturing must mix refined ingredients into consumer products and market them. This split between commercial and consumer manufacturers is important. Although these two segments will have similar business models, their products, customers and technologies will be quite different. Of course, some companies will have integrated both commercial and consumer manufacturing processes under one roof.
Commercial manufacturers will not only own some means of production, but some natural resources as well. Whether this is ore, timber, rubber, soy or livestock, these companies will have a number of processes dedicated to extraction of the raw materials. Once extracted, the raw materials will enter production cycles at manufacturing plants. The end products will be passed on through a sales force to other companies--the consumer goods companies. Having viewed the operational cycle for commercial manufacturers, there are very clear business processes for which data will be collected. These processes are extraction, manufacturing and sales. It is these areas that decision support applications will enhance.
Consumer manufacturing is different. Since most companies do not own their raw ingredients, the manufacturing process begins with managing a set of vendors from which ingredients are purchased. Once the ingredients are bought, manufacturing plants drive production. On the other side of the business, brand marketers are working to create compelling images of their products for the consumer. The consumer manufacturer has two customers: their retail channels of distribution and their product consumers. Although most consumer manufacturers are speaking about large retail chains when they use the word "customer," they cannot ignore that their success or failure is driven by human demand. The key areas for business intelligence applications are manufacturing, marketing, sales and feedback.
Evolution of Technology in Manufacturing
Often one can understand how an industry might integrate decision support by looking at the industry's tendency to work with technology as a whole. Unlike retailers, manufacturing companies were forced to adopt push technology very early in order to drive core processes. For these early adopters, the following traits are generally true:
- Manufacturers are not technology averse.
- Early adoption of the right technology is seen as a plus.
- Technical evaluations will be longer as they tend to delve deeper into the "nuts and bolts" of considered products.
- They are likely to have managerial staff that are technically proficient.
The first large-scale technical migration in both commercial and consumer manufacturing companies was to automate manufacturing processes. This tactical exercise allowed these companies to compete with overseas and domestic rivals by improving operations at the plant. Today, manufacturers continue to view operational technology initiatives as vital to the health of the organization. As a result, if technology is considered operational, there are large budgets and staff available for the integration. On the other hand, if the technology is not operational, its value is questioned in the organization.
The next migrations in manufacturing came with the advent of business integration software, such as SAP and Baan. The DBMS paradigm allowed these technologies to recognize similarities across process improvement in manufacturing and exploit them. During this time, the manufacturing companies were experiencing a schism, so the technology was welcomed. The manufacturing plants were becoming more remote--geographically, technically and psychologically--to the rest of the business. The front office of marketing and sales was not communicating tightly with the plants. As a result, the integration technology was seen as a way to solve a problem as well as to meet a challenge.
Today, most manufacturing companies have their enterprise integration technology in place. They may be completing the final step of extending this into the sales force and the supply chain. As supply chain management solutions and sales force automation solutions are finished, the second major migration will be complete. Now, manufacturing companies are watching their competition to see how they are differentiating along the technology axis. And what they are seeing is decision support.
Today and Tomorrow
The early adopters of decision support have followed the traits mentioned herein. They have looked for operational uses of the technology, and they have pushed the technical standard. Before considering specific challenges to manufacturing, one should review some actual applications that are in production today by early adopters.
Manufacturing: This application focuses on analyzing production data either toward inventory, quality or fulfillment. The inventory side answers questions such as "In the last year, what are the top ten most-used ingredients whose non-availability has slowed a production run?" Quality focuses on data collected by the plant around quality offenders, their frequency and the types of problems that arise and are solved. The quality module answers "Which part of the line pulls the most quality offending product for Product A, and how has this changed in the past year?" The fulfillment portions would cover topics such as "Show me lost revenue for customer cancelled orders when the items causing our backorder were to be produced from Plants X, Y or Z."
Marketing: This application suite covers promotions, category management and sales force. Consumer manufacturing companies have promotions budgets which need to be managed operationally and fine-tuned for enhancing marketing ROI. This would answer "For the top five promotion types for new product lines, ranked by frequency of execution, show me the retail chains showing most upside and best new geo-market penetration." The category management applications would answer "For all of our subcategories whose scorecards show ACV (All Commodity Volume) in the top decile, list average lineal feet by major retail chain in the LA and NY markets." Finally, sales force applications could answer "Which of my most profitable products does my key customer buy the least of, and what is the percent of revenue spent below each of my customer's major competitors?"
Consumers: This application type focuses on the end consumer. The types include consumer intimacy, line cycle times and e-commerce. Today's manufactured packaged goods all have a "1-800" number to call with feedback. The resultant data is of use to marketers and plant managers. The applications might answer "What product was most frequently called about for calling type 'Quality Praise' and how do those callers correlate with television advertising markets we play in?" Few would argue that in today's fast-paced world, the later the customer can take delivery, the less they are satisfied. As a result, watching the lagging line items on an order is important. Driving down line cycle times drives satisfaction up. This application answers "Which of today's orders include items reported by the plants to be 'Backordered with no substitute.' What percent of these orders are from my Gold Customers?" E-commerce might answer questions such as "For those consumers buying high-margin items, what are the top five most frequent methods by which they were referred to my site; and for each of these, show the marketing budget this season."
The ability to answer these questions seems to be tied to the success of the modern manufacturing company. Even so, only a handful of early adopters have production systems of these types up and running today. Why? The reason is twofold. First, as mentioned, the manufacturing IS shops have been busy integrating mission-critical enterprise integration systems and supply chain management. Second, decision support offers some critical challenges that those ready to tackle it must understand.
Challenges for DSS Integration
The IS challenges for decision support in manufacturing all lead back to the operational truths of the manufacturing industry. The first key challenge is business complexity. Early decision support data models would include three dimensions: product, location and time. This classic retail business model did not begin to address the complexity in a manufacturer's model. Commercial manufacturing is especially plagued by this challenge. Imagine a chemical manufacturer's need to model a synthetic process. The technology must scale to recording and analyzing data such as temperature, sub-second time segments and cellular changes. This data model may have 50 business dimensions, hundreds of characteristics and other fields which do not fit neatly into any category. Further, the facts collected may be complex equations or simple counts, but never sums. Commercial manufacturers possess data models that break all but the most sophisticated decision support engines. The rule of thumb here is not to simplify the business model for the tool. Never wrap your business around the technology; always wrap the technology around the business.
Once you've found an engine that can handle the complexity of modeling the business, the next challenge will be distribution. Manufacturers maintain large, remote sales forces that need access to fast, correct information. Some consumer manufacturers will even supply these forces with their own one-tier, local data marts. Imagine trying to maintain one hundred separate, individual, remote decision support applications, each producing category management scorecards and presentations tailored to thousands of retail stores. Many manufacturers will say maintaining just one application is enough. The distribution challenge may be managed, as long as the implementation is phased and the technology is best-of-breed. A key technology driving ease of distribution is the Internet. And as security, extranets and core technologies improve, Web-distribution will only become safer and more prevalent. New technologies can even send personalized analyses from a data warehouse via mobile phone, e-mail, pager or fax, eliminating the need to use personal computers. Remember to solve the technical challenge of distribution in phase 1, but scope the actual distribution implementation and change management into phase 2.
Decision support is likely to find some of its toughest customers in the commercial and consumer manufacturing space. These operationally focused businesses perform complex operations through a geographically dispersed network of both plants and sales staff. As such, geographic distribution of technology and data will be challenging at best. The decision support applications must address complex business models as well. For those manufacturers dedicated to solving these challenges, use of decision support will bring added value to their relationships with customers, tighten their line cycles, improve manufacturing quality and increase corporate valuation. Mystery solved.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access