A Small Company With Big Supply Chain Designs
Cascade Designs is a Seattle-based manufacturing company. Established in 1973 by two ex-Boeing engineers, we make and sell products that enhance the active lifestyles of our consumers. Our primary line of business is selling products that allow millions of people to enjoy the great outdoors. Our outdoor brands include Therm-a-Rest, Mountain Safety Research, SealLine, Platypus and Tracks. We also have leveraged our product expertise in the outdoor market to gain entry into the medical rehabilitation market through our Varilite brand, which provides seating and positioning systems for wheelchair users.
Laying the Technology Foundation
In 1998, Cascade Designs took a major step forward by purchasing the JD Edwards One World ERP system. At the time, this was considered extremely risky. Companies Cascades size at that time we had approximately $40 million in annual revenue simply didnt make that type of investment. We implemented financials, distribution and manufacturing in six months and launched. Most of the first year was spent learning how to run the operational business with the new ERP system. Our prior transaction-processing environment had two separate systems: one for financials, order entry A/R and A/P; the other for purchasing and MRP. The first benefit we got from One World was an obvious but valuable one: namely that we were able to achieve an extremely high level of visibility within each transaction and were able to follow those transaction strings from source to terminus. That may sound like a blinding grasp of the obvious, but when you lack that ability, its a revelation when you receive it. The other benefit we got was through a detailed view into our manufacturing accounting costs. The data captured by One World allowed us to do detailed analysis of our above-the-line costs and enabled us to make much better decisions with respect to costing and pricing.
Once we settled into using our new ERP system, we began a series of projects that were designed to lay a system foundation that would enable us to thrive as we grew. We implemented Microsofts OLAP tool (then called Plato) to begin analyzing the data we were capturing in JDE, and implemented in-house EDI processing capability and automated data collection devices in our warehouse.
We have continuously improved our use of these tools and it has helped enormously in our understanding of the business. The most dramatic change it facilitated was our move away from being the kind of company that makes decisions based on instinct or gut feel. While that approach certainly has its appeal, we prefer to make decisions on a solid base of analysis and information provided by these core systems. Although none of these projects were glamorous on their own, they were undertaken for the purpose of building the foundation from which we grew.
Building an Operations Base
In late 2001 we acquired a competitor, Mountain Safety Research (MSR), from one of our biggest customers, REI. This forced us to cease work on virtually all other IT projects to support the acquisition. The integration of MSR was our first real test of how well our build the base strategy was working. Once the acquisition was final, we set about migrating MSR onto our systems and integrating the companies business processes. MSR had been burned in the past with a series of failed, painful systems implementations and generally poor IT support. Their user base was skittish and skeptical when we told them we would be embarking on a four-month project, without consultants or new hires, to cut them over.
We produced, shipped and invoiced successfully from the first day. Prior to this event, our own management and users were not especially enthusiastic about our move to JDE. Serious ERP software requires serious investment to work well, whether youre a global multinational or a small manufacturing company in Seattle, and there was plenty of skepticism about the value of our decision to implement JDE. The speed and success of our MSR integration proved the value of that decision. The entire systems infrastructure issue, which can easily derail acquisitions of any size, was completed in short order and we were able to focus on other business issues.
Our next task was to think about how to migrate our European operations, based in Midleton, Ireland, onto our base systems architecture. It took 10 months. Again, we knew we had a winning implementation when we produced, shipped and invoiced successfully from the first day forward.
Radical Changes at the Plant Level Using Common Sense and Simple Tools
During the same period that IT was working hard on laying our systems foundation, our operations group was busy with radical changes inside our manufacturing plants. What made these changes successful was using common sense, simple tools and allowing each plant to capitalize on their unique strengths. Today we maintain three plants in Seattle. Our Yosemite plant produces Therm-a-Rest sleeping pads; our Chamonix plant produces SealLine dry bags, Platypus hydration systems, Packtowl sports and travel towels and SweetWater water filters; our Denali plant produces MSR stoves, snowshoes, cookware and water filters. Our Varilite medical products utilize capabilities from each of the three plants. Each plant has its own unique characteristics and challenges, and each plant made significant but different, improvements to its processes. Interestingly, none of their success was strongly linked to IT; however, in every case they were able to achieve their successes by leveraging the system foundations put in place before and during this period.
Yosemite Plant Its All About the People
In 2001 we realized that things had gotten complacent in our Yosemite plant. We installed a new plant manager and let him go to work. Our first order of business was to figure out what was wrong. We ran a series of production rate analysis exercises and quickly realized our biggest issue was gross labor inefficiencies. In additon, our crews lacked direction and motivation. In genearl, we realized that in order to optimize the production in this plant, we needed to focus on the people.
We made several organizational changes that had a dramatic impact on our bottom line. First, we articulated clear goals for each workstation coordinator. These folks became responsible for three things cost, quality and delivery. It sounds simple, but each element is measurable and improvable. By empowering the people who were in charge of each workstation for these elements, the benefits flow up from the point where they can be meaningfully impacted. In addition, we made process changes, which allowed for reductions in both indirect and direct labor. We reduced overall direct labor by 18 percent, but increased production from 7 to 30 percent (depending upon production line). We also began to take a serious look at our external supply lines and began working much more tightly with our two main raw materials suppliers to increase delivery frequency and raise product quality and consistency. By pushing quality back to the suppliers we were able to deliver raw materials directly to the line.
Our biggest change, however, was a philosophical shift. Prior to that time we were beholden to our annual sales forecast, which was consumed by an in-house production planning system, which in-turn fed our JDE master production scheduling algorithms. We were essentially a make-to-forecast production model operating in a highly seasonal sales environment. During our peak season, we could make lots of everything, but once things slowed down we were still building to plan and would often make lots and lots of products that very few customers wanted. Our Yosemite plant manager decided that enough was enough, and began examining other sources of data beyond the plan for input. He began to operate under the philosophy that if customers didnt want it, he wasnt going to make it. Sounds simple, but it initially wreaked havoc on our conventional costing metrics, which were constructed around building to the plan. It took us a while, but once the bottom line results started showing up, we decided to look at different metrics.
Chamonix Plant There Must Be a Better Way
In 2001, our Chamonix plant looked like all the plants they taught you about in college in 1990. We had a large raw materials store and shelves full of work in process (WIP), which we drew on to build large runs of products, communicated by work orders which were driven by our annual production plan. Our production lines were sprawling, space-intensive affairs. Around this time, the plant manager and one of the line supervisors began thinking that there must be a better way to do things. They began studying lean manufacturing principles and started making incremental improvements, primarily focused around optimizing space utilization. Each incremental improvement was built upon, and eventually we began to see some dramatic shifts. Our focus on eliminating wasted space led us to incorporate one-piece flow principles into our lines. We implemented kanban bins and started back-flushing raw materials. As we got better at this, we eliminated WIP and cut down the size of our raw materials stores. We saw dramatic increases in quality because we caught problems on the first piece, not after we had produced a batch of 1,000.
These space improvements allowed us to consider the idea of consolidating a smaller sewing plant into the Chamonix and Yosemite plants. The director of manufacturing, a lean thinker himself, brought together like-minded individuals and using cardboard cutouts figured out how to re-layout the plants and eliminate the need for 29,000 square feet of manufacturing space. By presenting such a radical change, using simple tools that everyone understands, people got on board because the consolidation made good business sense. This idea saved the company $12,400 a month money that now drops straight to the bottom line.
We had become efficient enough in this plant to begin experimenting with moving to a demand-driven model during slow sales periods, similar to what the plant manager was doing in Yosemite.
Denali Plant A More Conventional Lean Approach
Our Denali plant, obtained during the MSR acquisition, is a multifaceted manufacturing operation. Its core competencies include metal fabrication (stamping, cutting and machining), ceramic production and a variety of assembly operations. The output is primarily stoves, water filters and snowshoes. We began a lean journey in this plant and took deliberate steps to reduce waste and maximize production efficiencies. Along this journey our Denali plant manager asked, why do we need all these work orders floating around. They dont add value. This led us to establish new systems that created a paperless factory. Work orders were streamlined in JD Edwards One World using the kanban transaction feature to support supermarket kanbans set up on the shop floor. Rather than print all the work orders for a week, a daily one-page schedule was posted for the days requirement so people knew what to build. Once we implemented kanban bins for raw materials and WIP, the plant manager then went to work reducing the number of those bins as our processes became more reliable. Through the extensive use of kanban both on the shop floor and within the JDE software we have executed a textbook case of lean principles in this plant. We also spent a great deal of time reworking our plant layout, and then focused on individual production line layouts. We have implemented 5S, visual work instructions, conducted kaizen events and a myriad of other small improvements designed to eliminate wasted effort. In addition, we have crosstrained all of our direct labor force in this plant and reinforced this goal through incentives. Our skill-based system financially rewarded employees who learned new job skills. In 1997, a team member might have two to three skills. In 2005, we have team members with 10 to 15 skills. This flexibility of people allows the plant to be extremely flexible and responsive and creates the perfect platform for future initiatives. This ongoing focus on lean had a transformative effect on this plant. We reduced our floor space by 25 percent and reduced WIP by 70 percent.
Measuring Results
In each of the three plants, we had tangible and intangible rewards for our efforts. In many cases our employees felt empowered by their involvement in our plant transformations, and we have seen our morale improve as a result. In addition, as a U.S.-based manufacturing company, these improvements are a critical factor in enabling us to remain competitive relative to our offshore competitor, so much so, that we now believe domestic manufacturing is one of our strategic advantages. While many companies focus on reducing direct labor we see streamlining our indirect overhead functions as the largest opportunity. We have also seen a dramatic impact on the bottom line and the following improvements:
- Inventory turns improved by 40 percent;
- Output dollars/square foot improved 41 percent;
- Units produced/square foot improved 16 percent;
- Sales dollars/indirect operations heads improved by 17 percent; and
- Gross margin as a percent of revenue increased 16 percent.
Taking Some Risks, Extending the Footprint
Once our technology foundation was in place and our operations base financially secure, we began to think about interesting ways to leverage our knowledge and people. We decided on three main thrusts that we began work on in 2004 and will continue into 2006:
- Demand-driven supply chain linked with our ERP software package;
- Tighter integration with our customers; and
- Demand-driven links to our suppliers.
Our first initiative is moving our production and suppliers to a demand-driven supply chain. We believe we can best extend our technology and process improvements by redesigning our entire supply chain to focus on building what our customers want to buy. This will require some radical rethinking about how we forecast, schedule and execute. Our factories will be set up to manufacture any model any day allowing us to respond faster and more efficiently to the needs of the customer and the marketplace. This will also require tighter integration with customers and suppliers.
Currently we are running a pilot on our Varilite product line to fully integrate the three above concepts to create a demand-driven supply chain all the way from our customer to the supplier. We expect to see dramatic improvements in our responsiveness to customer orders from a lead time of two weeks to a 24-hour turnaround.
As we make this transition, we will apply all that we have learned through years of experience and leverage IT and operations to provide a sustainable competitive advantage.
