Case Study – Pelarsen Windows: Human v. Robots
Pelarsen Windows is in its third generation, founded initially in 1922 by Gunnar Pelarsen and now run by granddaughter and CEO Ingrid Pelarsen. The 1990s was an era of craftsmanship. One of the noted success factors for Pelarsen Windows at the time was its transition from craft to mass production. Pelarsen was a mover in streamlining the windows manufacturing process by standardizing the various components, allowing windows to be assembled in larger volumes and at remote locations.
Another one of its key success factors was its innovative products such as insulated glass, solar heating and cooling, and energy efficient windows. Due to its innovative capabilities, Pelarsen Windows had transformed glass from a commodity to a differentiating component of windows manufacturing. There are two main window types they produce; standard and architectural. Generally, sales offices decide the plants to assign orders to based on the available capacity, geographical proximity, and skill-level required. Pelarsen Windows now operates in 15 plants across North America, employs 10,000 employees and holds 34% of all windows manufacturing business in North America.
This case calls for an analysis of facts and recommendations on how the Texas plant can recover from the $6.34-million loss just experienced in 2007. What Plant Manager, Doug believes to be the cause of his loss is not necessary due to the lower volume of architectural window business. Another of his suggestions is to invest in more modern equipment. The analysis however shows that the higher costs of labor and an older facility also are also not the problem. A re-computation of cost data using ABC will reveal the flaws in Doug’s suggested recommendations.
The allocation of window business between the Oregon versus the Texas plant is the subject of concern for Texas Plant Manager, Doug Niedermeyer after incurring a $6.34 million loss in 2007. In his defense, Doug believes the Oregon plant receives preferential treatment and is given more of the expensive architectural business. In addition, the Oregon plant has state-of-the-art equipment that is highly automated while the Texas plant bears the high costs of labor and capital charge. He believes that the two plants are not on a level-playing-field. The issues to be analyzed are as follows:
Whether an increase in architectural windows business does in fact generate the profits Doug claims. What proportion of architectural versus standard windows production is optimal? The number of production lines to be exclusively dedicated to the architectural window business Whether there is sufficient capacity to support the planned output of architectural versus standard window. Whether increasing the architectural window production will increase costs of quality due to greater prevention and inspection costs across both product lines.
Some supplementary issues to analyze include:
Opportunity costs such as whether the decision to increase production of architectural windows impedes with any pre-existing orders or interferes with current customer accounts. If the Texas plant does take on more architectural window business, how efficient and effective is its performance with others in the industry. How the Texas plant’s performance compares with the arch-rival plant in Oregon after the proposed increase in architectural windows. Lastly, each of these two plants are concerned with their own profitability, but the entire well-being of Pelarsen Windows must be considered. In other words, whether increasing the architectural window business to the Texas plant support the overall goals of the company. As a start, the analysis will re-evaluate the Texas Plant’s cost data using ABC which provides clarity to its costs and operations and forms the basis for many of the conclusions. Analysis
Projected Income Statement for Texas Plant, Year-End 2008 Re-Computed using ABC The following is the projected income statement for the Texas Plant re-calculated using activity drivers rather than production volume as a cost driver. As will be seen below, the cost to manufacture a single unit of the architectural windows is much greater than the standard window (note the negative contribution margin of the architectural windows).
This is contrary to the volume-based method where standard windows were more expensive. Under ABC, It makes sense that the architectural window is more expensive as it is generally more labor-intensive, requires more direct material and consumes more fixed resources than its standard window counterpart. ABC is able to capture these costs by tracing the activities to the correct product. The following section will analyze the individual components of the income statement in further detail. What is realized is that cost information is severely distorted under volume-based allocation approach, leading Doug to make uninformed decisions. As it currently stands, architectural windows would be unable to bring the Texas plant into a profitable position.
Spoilage Rates Not Traced Properly
The projected 2008 income statement incorrectly combines all spoilage/waste into one pool, and distributes this cost based on production volume. The spoilage percentage for wood and glass is five times greater for the architectural windows than the standard window. It therefore does not make sense to allocate these costs using production volume as the standard windows would absorb the majority of these costs, simply because it has been produced the most. The partial income statement below, prepared using ABC shows wood and glass spoilage traced particularly to the product that incurs those costs. Both architectural and standard windows are appropriately showing their costs using their own prescribed spoilage rates. The assumption here is that spoilage is an inherent cost in the manufacturing process and therefore have been included in the per-unit costs. This is for simplification purposes only. In a different perspective, it may be worthwhile to keep these costs separate if they decide spoilage is an area of improvement and therefore should be visible as opposed to being buried into the direct material costs figure.
Direct Labor Costs are Inaccurate as it does not Consider Process Times The 2008 projected income statement does not consider the difference in process times between a standard window and architectural window. As a result, the extra amount of time required to produce an architectural window is overlooked, resulting in understated labor costs. On the contrary, the direct labor cost allocated to standard windows is overstated as it subsidizes the architectural windows. The consequence is that standard windows are picking up much of the labor costs that are actually incurred by the architectural windows. The following ABC calculation considers process times per window, level of skill required per process, and hourly labor costs. The assumption is that skilled workers are assigned the more complex architectural windows while helpers work on simpler windows.
The ABC calculation shows that architectural windows in fact consume 65% of all direct labor costs, compared to the previous calculation of 15.6% under the volume-based approach. This indicates that the direct labor costs for architectural windows were severely understated. By using ABC, it makes sense that labor costs for architectural windows are higher as it requires more expensive labor and time to manufacture. After recalculating labor costs using ABC, variable overhead can now also be computed accurately using the new labor hours.
Re-Work Time for Broken Glass Should Not be Allocated to Standard Windows The labor hours needed to replace broken glass for the architectural window line should not be absorbed by the standard window. The current calculation for direct labor does not distinguish between products that require re-work and those that do not. As a result, all costs of labor are combined into one pool and allocated based on production volume. The result is that standard windows bear the majority of these costs since it is produced at larger volumes, yet it in actuality it should not incur any. The table below shows that the replacement of glass is a separate cost, not to be absorbed by the standard window’s cost per minute figure.
Fixed Overhead Costs Improperly Calculated
Fixed overhead costs should also not be allocated based on production volume for the same reasons above. The architectural windows actually require more process activities and therefore should bear more costs. Fixed costs are allocated to each window type based on activity drivers on activity drivers below and proportions below:
In the next step, fixed costs per activity unit are then calculated:
In the final step, fixed costs are properly allocated to architectural and
standard windows based on activity usage per process, per window:
What can be seen here is that the architectural windows require more activity “units” to produce. It therefore absorbs its fair share of fixed costs as seen above.
Marketing and G&A Improperly Calculated
Same as the preceding reasons, Marketing and G&A should be recalculated using activity drivers rather than volume:
Summary: Volume-Based Versus Activity-Based Cost Allocation
As seen above, the current costing system assigns all costs using the number of windows produced as an allocation base. This essentially assigns higher costs to higher-volume products (standard windows) and lower costs to lower-volume products (architectural windows) without any regard to the increased costs incurred by the more complex /resource-intensive products. The simpler window was incurring higher costs in direct proportion with production volume, however volume does not in itself form a causal relationship to costs. In actuality, higher costs should not be associated with volume but rather the activities that in fact incur those costs. As a result, the simpler standard window is over-costed, and the complex architectural window is under-costed. By using ABC, activities are traced to specific products: architectural windows will be assigned their fair share of costs and standard windows will no longer subsidize the architectural window line by bearing the costs that they do not actually incur. Realization: Architectural Windows Not as Profitable As Believed Below is the computation of cost per unit of standard and architectural windows using ABC. Note that the architectural windows that Doug once believed yielded a $162 contribution margin actually yields a loss of -$102.80, 258% less than was calculated previously. Profit margin on the standard window is 43% higher than the volume-based approach at $73.98 per unit. Clearly, the standard windows had been subsidizing the architectural windows by absorbing many of its costs. Recalculating costs using activity-based drivers presents the proposed 2008 financial more realistically.
Production Capacity of Architectural and Standard Windows – Texas Plant It is useful to analyze the production capacity of the Texas plant in order to see if the proposed production of 270,000 standard windows and 50,000 architectural windows in 2008 is even physically feasible. As a starting point for analysis, the tables below analyze the Texas plant’s production capacity for standard windows alone and then architectural windows alone.
The grey-shaded cell is the number of units that can be produced per day on five production lines. The logic is that production can only be as fast as the slowest process which is: inserting glass for the standard window, and cutting and shaving for the architectural window. This analysis also assumes that there will always be units on hand from the previous day for each process. Thus, each processing area will have a total of 450 minutes of production time per day. This is important, as initially there is a bottle-neck for both architectural and standard window at the first two process steps: cutting & shaving and forming. For standard windows, the cutting & shaving consumes 56 minutes per batch and for forming 61 minutes per batch. The subsequent processes would need to wait for each of the two processes to finish first before it has a single unit to process, theoretically 56 min + 61 min= 1.91 hours later. Therefore it is assumed that units are available from the previous, days, weeks, years, etc. If the variables are changed to dedicate three lines to standard windows, and two to architectural windows (3:1), below is the result. The second table shows (4:1). Note: # of Windows Per day = number of windows per day in grey-shaded cells above * x/5 production lines) From the first table, if three lines are dedicated to standard windows and two to architectural, we have an excess capacity of 33,077 units for the architectural and a shortage of 7,013 for standard. In the second table, we shift to four standard lines and one architectural (4:1), with an excess capacity of 80,649 and shortage of 8,462 respectively. The conclusion from this analysis is that the proposed 270,000/50,000 product mix will not work as efficiently and effectively as planned. Whether two lines or one is dedicated to architectural windows, there will always be a shortage of one window over the other. Therefore the opportunity cost of lost business must be considered if this product mix is required. The lost could be potentially
greater if a single customer requires both standard and architectural windows, and will only place an order if they can get both at the same vendor. The Texas plant may also lose its ability to get specialized orders to the front-of-the-line faster and therefore lose its differentiating factor in the eyes of the sales team. Note that this analysis has so far assumed a 100% utilization rate, when in fact management prefers 85%-90%. In fact, it may be even less due to human factors such as sickness, low productivity, and new learning curves or plant-related such as new processes, break-downs, maintenance, etc. Realistically, this would mean that dedicating at least one line to architectural production would mean shortages and losses greater than 10-15% (100% minus 85%-90% utilization rate) than presented above. Production Capacity of Architectural and Standard Windows – Oregon Plant Below is computation of the production capacity of the Oregon plant, for both standard and architectural windows combined. The subsequent table shows that the Oregon plant can produce windows at 12% and 7% of the time it takes Texas to manufacture a standard, and architectural window respectively. However, Texas makes up for this slightly as it has five concurrent production lines while Oregon only has one.
This analysis sheds light again on Doug’s proposed product mix. The Oregon plant would be unable to meet its planned production of 310,000 units. It would in fact only produce 270,000 units. If it had to concur with this proposed product mix, the proposed financial statements would need to be adjusted down 40,000 units. Assume that they would choose to forego the lower margin standard window business.
Furthermore, if standard windows are decreased by 40,000 units, the following would be the revised income statement and effect on its accounting profit.
It is important to note that although not portrayed above, fixed costs for the standard window may decrease proportionately due to the 40,000 unit decline. It is also likely that fixed costs would stay the same if the cost driver is based on a measure that is stationary of units produced – i.e. supervision, headcount, etc. Regardless, the accounting profit for standard
windows is quite low compared to the architectural window business and the Oregon plant may find that it is not worthwhile to keep. As a conservative assumption, it is assumed that fixed costs are stationary and does not fluctuate based on volume of standard window produced. The standard windows would be no longer worth producing for the Oregon plant due to its fixed costs exceeding contribution margin. Oregon would then resist the production of standard windows as it is now a losing product line. This is likely not the result Doug intended as he himself hopes for less standard windows and more architectural. For simplicity, note that this analysis assumes a 100% utilization rate, when in fact losses would be more due to expected utilization rate of 85%-90%. The conclusion would still be the same – standard windows would be a losing product for Oregon at the proposed product mix.
Considering Quality Costs for the Texas Plant
The process data of 80 units of standard and architectural windows are graphed below.
From the graph, it is evident that the architectural windows deviate much more from the ideal specification point (yellow line). Many units also come close to exceeding the upper control and lower control limits. The process for standard windows is much more controlled, and likely requires less prevention and appraisal costs. For standard windows, it is assumed that internal failure is 3% and there is no external failure as there is no need to replace broken glass at customer site. On the contrary, prevention and appraisal costs are likely significant for the architectural window as many of its units come close to exceeding the upper and lower control limits. Internal failure is about 15% due to glass breakage. External failure can be estimated to be about 12% from the estimate of glass breakage at customer on-site. Producing more of the architectural window will therefore increase its quality costs – preventative, appraisal, internal and external. Standard windows may too be affected if employees are dedicating more time to quality checks for architectural window, and neglecting the standard window. Non-quantitative costs must also be considered such as Texas’ lost of reputation for high quality and low returns, potential customer
dissatisfaction, and potential loss of combined sales contracts affecting too the standard window line. Conclusions
As is seen at first glance from the analysis above, architectural windows should not be produced at all by Texas. It is much more expensive to produce and results in a negative contribution margin that further depletes the plant’s bottom-line. It is also not worth switching any of the five production lines to architectural windows as this would change the production capacity, causing production shortages and additional losses. Producing more architectural windows would also incur greater costs of quality due to the increase in spoilage and labor hours required to correct windows on-site. This would be contrary to Doug’s intentions as he prides himself on his reputation for quality within the Pelarsen plant network. If he intends to upkeep his reputation, it can be assumed that the costs of quality would increase significantly with the introduction of more architectural window business. This proposed allocation of standard to architectural windows would also cause the Oregon plant losses due to shortages in production capacity. As a result, the decrease is to the point where its fixed costs exceed its contribution margin for standard windows and there is no longer a financial incentive to continue producing them. Taking into account the financial position for both the Texas and Oregon plants, it can be concluded that the proposed product mix does not support the overall objectives of the entire company. Losses at the Texas plant, declining profits at Oregon, shortages in production capacity, customer dissatisfaction, potential hostility between plant managers, and misalignment of company goals are some of the consequences of Doug’s proposed allocation. The following may or may not be the solution, but it is at the very least better than Doug’s. Standard Windows for Texas; Architectural Windows for Oregon Below is what the financial statements would look like for Texas and Oregon if they are to just specialize in one window type:
From the above analysis, both plants will be profitable if they specialize only in one window type. The following assumptions have been made: The Texas plant can now employ only helpers, reducing the labor rate from $34 to $18
There are no changes in fixed production overhead, and marketing and G&A costs following the restructure of product mix (there would likely be further cost savings here but the following assumption has been made to remain conservative). There are no changes in capital charge following the restructure of product mix. The following are reasons why this proposed mix is better:
As the Oregon plant can produce an architectural window at a fraction of the time and resources, it is more economically efficient to allocate all architecture windows here. The Texas plant expels too much labor, time, and resources in general producing an architectural window to the point that there is no profit involved. Both plants will still have excess capacity available. This is good lee-way as it cannot be expected that capacity utilization be 100%. For the Texas plant, it may be beneficial also so that sales teams can continue to send specialized orders that need to be at the front of the line faster:
Texas Plant (windows per year)
Production Capacity: 438,250
Proposed Production of Windows: 400,000
Excess Capacity: 38,250
Capacity Utilization % 91.3%
Oregon Plant (windows per year)
Production Capacity: 270,000
Proposed Production of Windows: 230,000
Excess Capacity 40,000
Capacity Utilization % 85.2%
Both plants will turn over a profit. In fact, the Oregon plant will have even greater profit than what was projected based on Doug’s 2008 proposed income statement. The Texas plant will have a $17.8-million profit, a 1605% jump from his projected 2008 income statement. The results will support the overall objectives of the Pelarsen organization as a whole and the individual profit centers that make it up Company goals are aligned with individual goals
Some Further Considerations
Although many assumptions have already been stated throughout the analysis, there are some further considerations to note: Additional costs of distribution – if the plants are to produce one type of window each, there will be additional costs of distribution that have not been considered on the presented income statements. Geographical proximity – the sales team will no longer be able to use geographical proximity and available capacity as factors in determining where to allocate business. This may change their entire sales strategy and incur additional costs to lead this change. This also has not been considered on the income statements. Multiple product orders – plants may have clients who regularly purchase both window types at one plant. It may cause customer dissatisfaction if they are no longer able to do this conveniently or if they prefer to purchase local. This could cause further lost of revenue. Labor Mix – The earlier assumption was that skilled workers are assigned to architectural windows while helpers are assigned to standard windows. If in fact a window now requires 50% skilled workers and 50% helpers, Texas’ cost of labor could change as follows, resulting in a an overall profit on the income statement.
Quite possibly the plants may find that for various reasons it is more beneficial to be producing two window types. The reasons may be proximity, existing contracts to supply both window types, expected high costs of distribution, etc. If labor is split 50%/50% between window types, then the Texas plant does in fact turn a profit at the 270,000/50,000 product mix. The assumption has thus far assumed skilled workers to be assigned architectural windows and helpers to be assigned standard windows but if there is a deviation, then the conclusions will also deviate.
Cite this Pelarsen Windows – Humans v. Robots
Pelarsen Windows – Humans v. Robots. (2016, Oct 01). Retrieved from https://graduateway.com/pelarsen-windows-humans-v-robots/