The Rest of the Industrial World Does Not Operate Like This
My first job after graduating from Purdue University was as a manufacturing engineer with the Japanese company, Aisin Seiki. Aisin U.S.A. was in the process of transplanting to North America to supply components to Toyota manufacturing operations in Fremont, California (NUMMI), Georgetown, Kentucky, and Cambridge, Ontario. The Georgetown assembly plant was not yet running production.
My acceptance of the Aisin position was not predicated on finding a Toyota Group Company or learning the Toyota Production System. Those things had absolutely no meaning to me then. I took the position because it was literally a startup from the ground up, and the work was machining and assembly, which was what I was interested in learning. So after several years at Aisin, I started looking for a process development opportunity. This search landed me in Milwaukee, Wisconsin in an advance manufacturing group for the Briggs & Stratton Corporation.
On my second day at Briggs, I had a highly disconcerting revelation. It dawned on me that the rest of the industrial world did not operate anything like Aisin and Toyota. And that was why I got this new job—because I was supposed to be one of the Toyota Production System gurus for Briggs & Stratton. That was the point where my deep learning of TPS actually began.
An Accountant Who Did the Right Thing
One of the divisional accountants, Jeff, was trying to keep pace with the massive change going on in the shop from the flow and lean implementation. While maintaining the usual cost data and budgets of the traditional product and manufacturing projects, he also had to incorporate the cost surge for the massive plant rearrangement including tracking and managing the extra expenses.
The complexity was compounded when information was in transition from the traditional departments to the new work cells and Focus Factories (our term for engine value streams). The change was rapid and over multi-years. Thankfully, Jeff was a smart, even-keeled person and managed his way through all the ensuing chaos.
Jeff began to notice that our standard costing system did not fit well with the changes the engineers were making on the plant floor. What had seemed reasonable in the context of the old departments did not make sense for the new cells in the Focus Factories.
Then the engineers developing, designing, and responsible for the cells received a dictate from corporate. They wanted us to calculate a return on investment (ROI) on the changes and moves from the existing departments to the new Focus Factory cells. As one of the engineers, I was not thrilled about the dictate—more tasks to do when we were already extremely busy with all the redesigns and moves while working 50-60 hours a week and often on weekends. I reluctantly complied.
As a result I begin digging into Briggs & Stratton’s standard costing system and data. With the help of Jeff and two shop foremen, Jim Anderson and Bill Schneider, I learned how to read and use the company’s “Cost per 100” forms, the complied cost data collected about transactions from the shop floor processes.
As I complied the actual cost of each crankshaft—with over forty separate part numbers for the engine model I was evaluating—I discovered the cost data was absolutely wrong. I knew Briggs & Stratton based nearly every major business decision on this cost information, and I was stunned and horrified. Driven by their lack of validity, I became quite astute at crunching through the forms, and I grew to realize my employer was making many faulty decisions based on this data. This was not good!
The linchpin of this discovery was one crankshaft part number, which was under three percent of the overall volume of nearly a million engines per year. It also had multiple extra and expensive processes because of its pinion gear-based power-take-off (PTO) end. This meant extra turning, grinding, grooving, hardening, and gear-shaping.
However, our standard cost system calculated that part number to be substantially less expensive to manufacture than another crankshaft part number which constituted nearly seventy percent of the overall annual volume but had only a simple straight-shaft PTO with a keyway. This meant that all of the extra processes required for the pinion gear PTO crankshaft were not required for the straight-shaft PTO, which, in turn, meant the straight-shaft PTO had to be less costly for us to manufacture. It also had the advantage of much higher volume, which improved the scale and efficiencies of producing it.
As a result of my discovery, I marched into Jeff’s office and proclaimed, “Jeff, these numbers are completely wrong, so why are we doing this and why are we using them!” Jeff calmly replied, “I understand, Jim, but corporate asked us to do the ROI so we need to get it done.” My even more agitated reply was, “But Jeff, these numbers are wrong!” Jeff reassured me that we needed to complete the ROI calculations as required. But rightfully and thankfully, Jeff had already begun to realize these same issues. He also knew we needed to pass along the requested information to keep corporate happy and out of our hair in order for us to pursue the changes we really needed.
This initial meeting turned into a series of discussions that Jeff and I had over several months. We discussed the standard cost system and what would make sense to present that would actually reflect the changes being made on the shop floor. Much of our discussion consisted of me asking him questions since I was an engineer and not an accountant. I needed to understand the correlations between cost accounting, physicality on the shop floor, and our implementation of flow.
One day, I was paraphrasing back to Jeff what I believed a costing system should do to accumulate the costs of each component in the cells. Jeff asked me to take a look at his computer screen and asked, “Is this is what you mean and are thinking?” What I saw is what, today, would be called value stream costing. I replied, “Yes, that sure looks like it! Can you explain it to me?” He did. Our technical term for value stream costing was “Jeff’s spreadsheet.”
Jeff had, on his own and from our discussions, decided he needed to make changes in accounting that would support what the engineers were doing in the shop. When the work cells were physically set into place and operated, several things became apparent. Equipment, operators, setup personnel, material, tooling, floor space, and much more were directly part of the cells. In essence, all these resources were being consumed directly and exclusively by the cells. The need to do all of the traditional transactional and allocation processes had vanished. Most costs were now aligned with a product family and could be viewed as direct costs. Jeff began to make changes to the cost tracking system as he implemented value stream costing, then known as Jeff’s spreadsheet.
At that point, we were operating yet to be named value streams and using yet to be named value steam costing.
Jeff’s changes resulted in some remarkable accounting gains. Under the traditional costing system, overhead, the costs from unknown sources, accounted for approximately 85 percent of the cost of the product. It was the “peanut butter spread” of costs that had to go somewhere.
After Jeff started using his breakthrough version of value stream costing, overhead accounted for only about 15 percent of product cost. And of this 15 percent, we knew its breakdown and sources. There were no “unknowns”. Overhead consisted mainly of corporate, divisional, and Focus Factory overhead. Frequent adjustments were unnecessary as those resource consumptions were reasonably stable and easily trackable. More importantly we could adjust them easily, even daily, if required. The new, detailed cost data was available at the engine and the component level. It was easier to change overhead rates as well and have confidence in the accuracy of the information. This significantly improved the division’s decision-making capability.
The Accountant as a Change Agent
The methods Jeff used are still not predominant in manufacturing today. Only a few companies are trying to implement and use these changes and methods.
What is most impressive is that this application took place in the early 1990s—over 20 years ago! At the time there were no books describing what to do. The terminology, training, and techniques of value streams and value stream mapping were not known until Womack and Jones’ book, Lean Thinking. Jeff accomplished what he did using a sharp, open, inquisitive mind while trying to do the right thing for his employer.
As an accountant, he was supporting necessary changes being made on the shop floor to improve the product. He recognized a need, understood that traditional accounting techniques were of limited use, and invented, on his own (perhaps with a little help from a crabby engineer), an alpha version of value stream costing.
Jeff truly became a change agent. He accomplished something quite remarkable, and did it humbly, as just part of his job as an accountant.