QUESTIONS

6-1    Nonfinancial measures such as customer satisfaction and customer loyalty are important in managing relationships with customers, but an excessive focus on improving customer performance with only these metrics can lead to deteriorating financial performance. To balance the pressure to meet and exceed customer expectations, companies should also be measuring the cost to serve each customer and the profits earned, customer by customer.

6-2    Examples of differences between customers who have high and low costs-to-serve may be drawn from the chapter’s Exhibit 6-1, part of which appears below.

High Cost-to-Serve Customers	Low Cost-to-Serve Customers
Order custom products	Order standard products
Small order quantities	Large order quantities
Customized delivery	Standard delivery
Manual processing; high order error rates	Electronic processing (EDI) with zero defects
Large amounts of pre-sales support (marketing, technical, and sales resources)	Little to no pre-sales support (standard pricing and ordering)
Large amounts of post-sales support (installation, training, warranty, field service)	No post-sales support
Pay slowly (have high accounts receivable from customer)	Pay on time (low accounts receivable)

 

6-3    Companies should not necessarily avoid high cost-to-serve customers. The high cost of serving such customers can be caused by their unpredictable order patterns, small order quantities for customized products, nonstandard logistics and delivery requirements, and large demands on technical and sales personnel. Activity-based pricing may be used to ensure that companies charge prices that are high enough to cover the high costs of serving such customers. Customers may, in response, change their behavior to become lower cost-to-serve customers. Companies may also improve the process used to produce, sell, deliver, and service customers in order to improve customer profitability.

6-4   The 80-20 rule as applied to sales revenues refers to the common finding that when companies rank products and customers from the highest sales volume to the lowest, they generally find that their top-selling 20% of products or customers generate about 80% of total sales.

6-5   The typical shape of a graph of cumulative profits versus percentage of customers ranked from most profitable to least profitable is a “whale curve,” in which the most profitable 20% of customers generate about 180% of total profits; this is the peak, or hump of the whale above sea level. The middle 60% of customers approximately break even, and the least profitable 20% of customers lose about 80% of total profits, leaving the company with its 100% of total profits (“sea level” in the whale curve represents the company’s actual reported profits). The hump (or maximum height) of a cumulative profitability curve generally hits 150% to 250% of total profits, and this height is usually achieved by the most profitable 20% to 40% of customers.

6-6   Service companies, even more so than manufacturing companies, must focus on customer costs and profitability sales because the variation in demand for organizational resources is much more customer driven than in manufacturing organizations. A manufacturing company producing standard products can calculate the cost of producing the products without regard to how their customers use them. In this sense, the manufacturing costs are customer independent. Of course, the costs of marketing, selling, order handling, delivery, and service of the products might be customer specific. For service companies, in contrast, customer behavior determines the quantity of demands for organizational resources that produce and deliver the service to customers.

6-7   Consider customers who maintain checking accounts at a bank. One customer may maintain a high cash balance in his checking account; make very few deposits, withdrawals, balance inquiries, or service requests; and use only electronic channels (i.e., automatic teller machines and the Internet). Another customer may manage her checking account balance very closely, keeping only the minimum amount on hand, and use her account heavily by making many small withdrawals and deposits via manual transactions with bank tellers. The consequence of the customers’ very different use of the bank’s resources is that the second customer is much more costly to serve than the first customer.

6-8   The four broad groups of actions that managers might use to transform unprofitable customers into profitable ones are:

• Improve the processes used to produce, sell, deliver, and service the customer.
• Deploy menu-based pricing to allow the customer to select the features and services it wishes to receive and pay for.
• Enhance the customer relationship to improve margins and lower the cost to serve that customer.
• Use more discipline in granting discounts and allowances.

6-9   A pricing waterfall chart depicts the multiple revenue leaks from list price caused by special allowances and discounts granted to obtain the order and build customer loyalty.

6-10 Salespersons’ incentives or compensation plans that set minimum quotas and commissions based on sales revenue, and tie bonuses and rewards to achieving sales revenues above a stretch target contribute to unprofitable customer relationships. Such arrangements encourage salespeople to close deals and generate revenues without regard to the cost of fulfilling the special arrangements negotiated in the deal and the impact of discounts and allowances granted to close the deals.

6-11  Life-cycle profitability analysis weighs the expected value of the stream of a customer’s net margins against the cost of acquiring the customer. Companies must understand variation of customer demands across multiple products and services and the expected length of each customer’s relationship with the company (given potentially costly retention efforts), in order to calculate each customer’s total life-cycle profitability. The analysis involves identifying characteristics of profitable customers. Companies can then direct their marketing efforts accordingly to specific segments that are most likely to yield and retain profitable customers.  Acquiring the information necessary for life-cycle profitability analysis and using this information to calculate life-cycle profitability thus provides insights about which customer groups are likely to be profitable over the expected life of a customer’s relationship with the company.

6-12  No. Experts now agree that it is a mistake for a company to use the satisfaction score as its only customer metric. A customer’s satisfaction is an attitude or belief stemming from a feeling that the product or service has generally delivered on the customer’s expectation of performance. But having attitudes and beliefs are not actions; a customer’s attitude toward a product or a company does not readily translate into the desired behavior of repeated and increased purchases of the product or service, or customer loyalty.  

6-13  Loyal customers are valuable for several reasons. Three reasons are required in this question; the chapter lists the following five reasons:

• Loyal customers have a greater likelihood to repurchase, and the costs to retain them are generally much lower than the cost to acquire an entirely new customer.
• Loyal customers can persuade others, through word of mouth, to become new customers; they can become references for potential future customers.
• Loyal customers are less likely to defect when a competitor offers a similar product at the same or slightly lower price.
• Loyal customers are often willing to pay a price premium to retain a known and trusted relationship with a key supplier.
• Loyal customers are willing to collaborate with the supplier to improve performance and develop new products.

6-14  Customer retention rate, though a traditional customer loyalty metric, is a poor indicator of a customer’s loyalty. This is because customers often remain with their current supplier because of inertia, high switching costs, or the current lack of an alternative supplier.

6-15  The five stages of a hierarchy for categorizing customer satisfaction and loyalty are:

• Satisfied customers, as measured by how well a customer’s expectations have been met or exceeded in an individual transaction or long-term relationship.
•  Loyal customers, as measured the customer devoting an increasing “share of wallet” for repeat purchases from the same supplier.
• Committed customers, those who not only purchase frequently from the supplier but also tell others about the supplier’s great products and service.
• Apostle customers, committed customers who have credibility and authority when they recommend the supplier to friends, neighbors, and colleagues.
• Customer “owners,” who take responsibility for the continuing success of the supplier’s product or service.

A company should strive to have more of its customers in categories 3, 4, and 5 above, since their willingness to recommend the company to others and to collaborate with it to continually improve product features and service makes them far more valuable, with a much higher customer lifetime value, than customers who are merely satisfied with the most recent transaction.

Companies that offer personalized services to customers, such as by a Nordstrom salesperson, or that offer rewards to loyal customers can generate high customer loyalty. These loyal customers then recommend the companies to others. Online companies such as amazon.com generate customer loyalty because provide a good selection of desired products, competitive prices, easy ordering, quick delivery, and reviewers’ comments on products. Loyal customers, in turn, provide comments on products; helpful reviews in conjunction with the other positive attributes of the company contribute to continuing and growing customer loyalty.

6-16  The net promoter score is computed based on responses to the question, “How likely is it that you would recommend [Company X] to a friend or colleague?” Customers respond on a scale from 1 (extremely unlikely) to 10 (extremely likely), with 5 representing a neutral point. The net promoter score is the percentage of customers who are “promoters” (score of 9 or 10) less the percentage who are “detractors” (scores of 1 through 6). The net promoter score is recommended based on research that finds that a customer’s willingness to recommend a company is strongly correlated with future growth and profits. In contrast, a customer retention rate, a traditional customer loyalty metric, can be a poor indicator of a customer’s loyalty. This is because customers may remain with a supplier because there no alternatives or simply because of inertia. Research suggests that “promoters” are the only truly loyal customers and “detractors” may harm the company’s reputation and brand value.

EXERCISES

6-17	(a)		Ashton	Brown
		Sales	$430,000	$350,000
		Cost of goods sold	$220,000	$155,000
		Gross margin	$210,000	$195,000
		Marketing, selling, distribution, and administrative expenses: 33% × sales	$141,900	$115,500
		Operating profit	$68,100	$79,500
		Operating profit/Sales	15.84%	22.71%
		-
	(b)	Sales	$430,000	$350,000
		Cost of goods sold	$220,000	$155,000
		Gross margin	$210,000	$195,000
		Marketing, selling, distribution, and administrative expenses
		Sales representative travel	$9,000	$42,000
		Service customers	15,000	110,000
		Handle customer orders	1,000	12,000
		Ship to customers	24,000	72,000
		Total activity expenses	$49,000	$236,000
		Operating profit	$161,000	–$41,000
		Operating profit/Sales	37.44%	–11.71%
	(c)	The activity-based costing method provides more accurate assignments of marketing, selling, distribution, and administrative expenses by identifying activities consumed by each customer and assigning costs to customers based on their activity usage. In this example, Brown places smaller orders, orders more frequently, and requires more after-sales support (travel and service support) than Ashton does.

6-18  (a)     $873,600/20 operators = $43,680 per operator

                   $43,680/1,560 productive hours/operator = $28.00 per hour

         (b)     (i)      0.1 + (10 × 0.02) = 0.3 hours

              0.3 hours × $28/hour = $8.40

                   (ii)     0.06 hours × $28/hour = $1.68

6-19  (a)     Students can refer to the In Practice box on textbook page 224. The following table shows the customer profits sorted from largest to smallest, the cumulative profit after adding each customer, and the cumulative profit percentages that were used to plot the whale curve shown below. In addition, the first row (rank 0) was included below so that the graph begins at 0.

Customer Number	Sorted Profit	Cumulative Profit	Cumulative Profit %	Cumulative Percent of Customers	Profit Rank
0	$            0	$               0	0%	0%	0
13	264,000	264,000	31%	4%	1
9	259,000	523,000	61%	8%	2
12	233,000	756,000	88%	12%	3
1	221,000	977,000	114%	16%	4
4	217,000	1,194,000	139%	20%	5
11	208,000	1,402,000	163%	24%	6
7	101,000	1,503,000	175%	28%	7
10	96,000	1,599,000	186%	32%	8
22	87,000	1,686,000	197%	36%	9
14	83,000	1,769,000	206%	40%	10
25	75,000	1,844,000	215%	44%	11
17	50,000	1,894,000	221%	48%	12
20	30,000	1,924,000	224%	52%	13
5	22,000	1,946,000	227%	56%	14
16	14,000	1,960,000	228%	60%	15
6	9,000	1,969,000	229%	64%	16
21	-  10,000	1,959,000	228%	68%	17
2	-  40,000	1,919,000	224%	72%	18
19	-  90,000	1,829,000	213%	76%	19
24	-100,000	1,729,000	202%	80%	20
3	-143,000	1,586,000	185%	84%	21
23	-158,000	1,428,000	166%	88%	22
15	-179,000	1,249,000	146%	92%	23
18	-191,000	1,058,000	123%	96%	24
8	-200,000	858,000	100%	100%	25

 

(b)     The most profitable 20% of the customers (i.e., the most profitable 5 out of 25 customers) generated 139% of the profit, as shown in the table in part (a).

(c)      The least profitable 20% of the customers (i.e., the least profitable 5 customers) lost (202% – 100%) = 102% of the profit.

6-20  (a)     A company might transform its breakeven or loss customers into profitable ones through process improvements that lower the costs of serving customers. For example, if most customers are migrating to smaller order sizes, companies should strive to reduce the costs of processes such as setup and order handling so that customer preferences can be accommodated without raising overall prices. One way to become more efficient in handling orders is to encourage customers to access a purchasing web page and place their orders over the Internet. This would substantially lower the cost of processing large quantities of small orders. If customers have a preference for suppliers offering high variety, manufacturing companies can try to customize their products at the latest possible stage, as well as use information technology to enhance the linkages from design to manufacturing so that greater variety and customization can be offered without cost penalties.

(b)     A company might also transform its breakeven or loss customers into profitable ones by activity-based (menu-based) pricing. This approach establishes a base price for producing and delivering a standard quantity for each standard product. In addition to this base price, the company provides a menu of options, with associated prices, for any special services requested by the customer. The prices for special services on the menu can be set simply to recover the activity-based cost to serve, allowing the customer to choose from the menu the features and services it wishes while also allowing the company to recover its cost of providing those features and services to that customer. Alternatively, the company may choose to earn a margin on special services by pricing such services above the costs of providing the service. Pricing surcharges could be imposed when designing and producing special variants for a customer’s particular needs. Discounts would be offered when a customer’s ordering pattern lowers the company’s cost of supplying it.

Activity-based pricing, therefore, prices orders, not products. When managers base prices on valid cost information, customers shift their ordering, shipping, and distribution patterns in ways that lower total supply chain costs to the benefit of both suppliers and customers

(c)      Managing customer relationships provides still another way for a company to transform its breakeven or loss customers into profitable ones. For example, companies can persuade their customers to use a greater scope of the company’s products and services. The margins from increased purchases contribute to covering customer-related costs that do not increase proportionately with volume, such as the cost of the salesperson assigned to the account. Companies can establish minimum order sizes from unprofitable customers, so that the margins from higher volumes more than cover the costs of processing an order and setting up a production run for the customer.

(d)     Finally, companies might transform their breakeven or loss customers into profitable ones by using their activity-based costing systems to trace all revenue deductions, as well promotional costs and allowances, to individual orders and customers in order to calculate actual, realized profit or loss, customer by customer. This approach should lead to disciplined discounts and allowances instead of a situation where companies fail to see all of the revenue leaks from list price because they record the discounts and allowances in different systems and make the revenue deductions at different times of the year. Without a disciplined approach to discounts and allowances, companies might find that various functional areas (e.g., salespeople, the finance group, and marketing) independently offer discounts or allowances during the year, leading to breakeven or loss customers because of the large total deductions from revenues.

6-21  (a)     Firms may fail to see all of the revenue leaks from list price on orders because they record the discounts and allowances in different systems and make the revenue deductions at different times of the year. For example, the prompt payment discount may be recorded by the finance department in an aggregate income statement account (sales deductions); the finance department may lump all freight costs into a general financial statement account labeled as transportation expenses. It does not link either the purchase discount or the freight expense back to a customer or an individual order. The volume discount may be refunded to the customer only once it has accumulated sufficient volume to qualify, and it is not linked back to the individual transactions that qualified for the volume discount. With discounts and allowances recorded into different accounts and at different times, no manager sees the complete picture for individual orders and consequently no one realizes how much revenue loss occurs with individual orders.

(b)     Once firms become aware of pricing waterfalls leading to undesirably large sales discounts, they can use their activity-based costing systems to trace all revenue deductions, as well promotional costs and allowances, to individual orders and customers in order to calculate realized profit or loss by order or by customer. Using this information, companies can periodically (e.g., every quarter) calculate an operating income statement for every customer. Furthermore, companies can use the activity-based information on MSDA costs to base salesperson incentives on order and customer profits, not just sales.

6-22  (a)     If Saunders reduced its sales discounts so that net revenues increased by 10%, the net revenue would increase to $220,000 and operating profit would increase by ($70,000 − $50,000)/$50,000 = 40%, as shown below. Thus, a 10% increase in net sales revenue would result in a 40% increase in operating profit.

		With 10%
	Initial	Revenue Increase
Net sales revenues	$200,000	$220,000
Variable costs	80,000	80,000
Contribution margin	120,000	140,000
Fixed costs	70,000	70,000
Operating profit	$50,000	$70,000
% change in operating profit		40%

(b)     If sales discounts increased by another 2%, the net revenue would decrease to $196,000 and operating profit would decrease by ($50,000 − $46,000)/$50,000 = 8%, as shown below. Thus, a 2% decrease in net sales revenue would result in an 8% decrease in operating profit.

		With 2%
	Initial	Additional Discount
Net sales revenues	$200,000	$196,000
Variable costs	80,000	80,000
Contribution margin	120,000	116,000
Fixed costs	70,000	70,000
Operating profit	$50,000	$46,000
% change in operating profit		8%

(c)      Let x% = the percentage change in net sales revenue. Under the circumstances in this problem,

Change in operating profit = (Initial revenue × x%)

Percentage change in operating profit =

(Initial revenue × x%)/(Initial operating profit) =

x% × (Initial revenue)/(Initial operating profit)

 

Thus, the higher the ratio of sales to operating profit, the larger the change in operating profit in response to a change in x. Equivalently, the lower the ratio of operating profit to sales, the larger the change in operating profit in response to a change in x. This implies that the higher the ratio of operating profit to sales, the smaller the change in operating profit in response to a change in x.

6-23  (a)     Based on the information given, Donner is more profitable and Carlson is unprofitable:  

	Carlson	Donner
Sales	$450,000	$400,000
Cost of goods sold	180,000	80,000
Gross margin	270,000	320,000
MDSA expenses	320,000	65,000
Operating profit	$−50,000	$255,000

(b)     The sum of the commissions is higher for the sales revenue scheme, and this scheme will encourage salespersons’ efforts to sell to Carlson, which is unprofitable. The profit scheme will encourage sales efforts to focus on Donner, which the company prefers because Donner is profitable and Carlson is not. Moreover, at the stated commission rates, the company will pay less in commissions with the profit scheme than under the sales revenue scheme.

	Carlson	Donner	Cost to Company
Sales	$450,000	$400,000
Commission on sales revenue	2%	2%
Total commissions on revenue	$9,000	$8,000	$17,000
Operating profit	$−50,000	$255,000
Commission on profit	4%	4%
Total commissions on profit	$0	$10,200	$10,200

6-24  The customer lifetime value, CLV, for Customer 421 is calculated by summing [(Mt − ct) × (rt)t − 1]/(1 + i)t, where i = the cost of capital, for t = 1, …, 6 and then subtracting the initial acquisition cost. In this problem, rt = 0.8 each year and i = 0.1. Calculations were performed in Excel and rounded, leading to the slight discrepancy in the total below.

t	Mt	ct	(rt)t − 1	(1 + i)t	[(Mt − ct) × (rt)t − 1]/(1 + i)t
1	$250	$ 60	1.0	1.1	$172.7273
2	300	50	0.8	1.21	165.2893
3	325	50	0.64	1.331	132.2314
4	350	50	0.512	1.4641	104.9109
5	375	40	0.4096	1.61051	85.2003
6	400	40	0.32768	1.771561	66.5881
					$726.9472

CLV = $726.95 − $600 = $126.95.

6-25  The net promoter score is the percentage of customers who are promoters (scores 9 or 10) less the percentage who are detractors (scores 1 through 6). Thus, the net promoter score = 12.82% + 25.30% − (2.18% + 0.84% + 1.26% + 7.14% + 9.86% + 10.96%) = 5.88%.

Score	Number of Responses	Percentage of Total
10	641	12.82%
9	1265	25.30%
8	1254	25.08%
7	228	4.56%
6	548	10.96%
5	493	9.86%
4	357	7.14%
3	63	1.26%
2	42	0.84%
1	109	2.18%
	5,000

PROBLEMS

6-26  This question is designed to generate discussion on what constitutes a desirable customer. Although a credit customer who charges a large dollar volume and pays the balance in full on time each month is probably a good credit risk, the customer is not the most profitable to the credit card issuer. (However, some credit card issuers may generate indirect additional revenue from such customers by working with advertisers to selectively add advertising inserts to these customers.) All credit card purchases generate merchant fees; the credit card issuer hopes to generate additional income through interest payments and late fees.

Type 6 is the least desirable type of customer because that type generates no revenue but causes the issuer to incur costs to send statements. Type 1 is preferable to Type 6, but less preferable than all the remaining types because of the short-term, low-interest arrangement. Type 3 is preferable to type 4 because of the late fees, and type 4 is preferred to type 5 because of the higher balances on which interest is paid. Type 2 is probably less preferred than types 3 and 4, and possibly type 5. The suggested complete ordering will depend on the relative dollar magnitudes assumed. A reasonable ordering, from most preferred to least preferred, is 3, 4, 2, 5, 1, 6.

 

6-27  (a)		Customer Type 1	Customer Type 2	Customer Type 3	Customer Type 4
	Sales	$1,000	$1,000	$2,500	$3,000
	Less returns	0	200	500	1,500
	Net sales	$1,000	$800	$2,000	$1,500
	Cost of goods sold, 75% of sales	750	600	1,500	1,125
	Processing mail orders, $5 per nonphone order	0	30	20	0
	Process phone orders, $80 per hour	20	0	0	80
	Process returns, $5 per item returned	0	20	10	120
	Process overnight delivery requests, $4 per request	4	0	0	48
	Maintain customer relations	50	50	50	50
	Profit	$176	$100	$420	$77
	Profit ¸ Sales	0.18	0.10	0.17	0.03

• Although customer type 4 has the highest sales, it has the highest dollar returns and the lowest profit. Customer type 3 is by far the most profitable, even though its sales are less than customer type 4’s (but customer type 3’s net sales exceed customer 4’s). Customer types 1 and 2 are more profitable than customer type 4 in total dollars and in percent of sales. Customer type 1 returns the highest profit as a percent of sales, slightly above customer type 3’s ratio. Cost of goods sold represents 75% of sales revenue, so the remaining costs as a percent of sales pertain to each customer’s interaction with the company. Customer type 4 is the most expensive to service because it orders frequently, places orders in a costly manner (one hour on the phone), returns many items, and requests overnight deliveries. Customer type 1 is fairly low-cost to serve in spite of ordering by phone and requesting overnight delivery because this customer type orders only once a year and does not return merchandise. Aside from returns, customer types 2 and 3 are fairly low-cost to serve because they order via mail and request regular delivery rather than overnight delivery.
•  
• Kronecker can seek to reduce the service activity usage or improve efficiency to reduce the cost of providing services. For example, Kronecker might ask customers the reason for returns, and follow up with ways to reduce problems that caused the returns. The company might also explore ways to make phone ordering more efficient, to reduce the time spent on the phone. Kronecker may also charge fees to handle overnight delivery requests.

 

6-28  Key points in the essay should include the items below. Exercise 6-19 provides a numerical example for developing a whale curve.

• Exhibit 6-2 illustrates an 80-20 graph for sales revenues, with cumulative percent of products or customers on the x-axis and cumulative percent of revenue on the y-axis. To prepare such a graph for n products, rank the products from highest to lowest revenue in spreadsheet column B. In column A, enter the ranks (integers from 1 to n). Compute the cumulative percent of products in column C by dividing each product rank by n and displaying the result in percent format. These percents will be plotted along the x-axis. Compute the cumulative revenues in column D and cumulative revenue percentages (divide each entry in column D by the total of all revenues in column D) in column E. Beginning with the highest-revenue product, plot the point indicating (Product 1’s percentage in column C, Product 1’s revenue percentage in column E). Continuing with the second-highest-revenue product, plot the point indicating (Product 2’s percentage in column C, cumulative percent of revenue for products 1 and 2 in column E). Continue to plot (cumulative percent of products, cumulative percent of revenue) in this manner until points representing all the products have been plotted, as in Exhibit 6-2. A graph for customers would be plotted similarly.
• Typically, companies find that their top-selling 20% of products or customers generate about 80% of total sales. The lowest volume 40% of products or customers generates only 1% of total sales.
• Although the 80–20 law applies well to sales revenues, it does not apply to profits.
• The “In Practice: Building a Whale Curve of Customer Profitability” describes in detail how to prepare a whale curve with a spreadsheet and Excel. (Also see Exercise 6-19, which describes adding a starting point of 0.) Briefly, rank the customers from most profitable to least profitable (or most unprofitable). Beginning with the most profitable customer, plot the point indicating (Customer 1’s percentage, Customer 1’s percent of total profit). Continuing with the next most profitable customer, plot the point indicating (Customer 2’s percentage, cumulative percent of total profit). Continue to plot (customer percentage, cumulative percent of profit in this manner until points representing all the customers have been plotted, as in Exhibit 6-3.
• Exhibit 6-3 provides a typical whale curve of cumulative customer profitability. The most profitable 20% of customers generated about 180% of total profits; this is the peak, or hump of the whale above sea level. The middle 60% of customers approximately break even, and the least profitable 20% of customers lose 80% of total profits, leaving the company with its 100% of total profits (“sea level” in the whale curve represents the company’s actual reported profits). The hump (or maximum height) of a cumulative profitability curve generally hits 150% to 250% of total profits, and this height is usually achieved by the most profitable 20% to 40% of customers.

6-29  (a)		Customer 1	Customer 2
	Volume discount if 20 or more units are ordered	2%	2%
	Pay in full in 15 days	3%
	Cooperative advertising allowance for featuring the company’s products in its advertisements	4%	4%
	Take a large shipment before the end of the quarter in advance of an expected seasonal increase in demand	5%
	Online ordering discount	2%	2%
	Rebate on sales during specific promotional periods	2%
	Free freight	3%
	Total	21%	8%

• Randolph Company’s management may have been unaware of the potentially large discounts offered to its customers because the discounts and allowances arise from different sources and are recorded in different systems. In addition, the revenue deductions may occur at various times of the year. For example, the prompt payment discount may be recorded by the finance department in an aggregate income statement account (sales deductions); the finance department may lumps all freight costs into a general financial statement account labeled as transportation expenses. The marketing department may initiate the cooperative advertising allowance, volume discount allowance, and rebates. With discounts and allowances recorded into different accounts and at different times, no manager sees the complete picture for individual orders and consequently no one realizes how much revenue loss occurs with individual orders.

 

(c)      Once firms become aware of pricing waterfalls leading to undesirably large sales discounts, they can use their activity-based costing systems to trace all revenue deductions, as well promotional costs and allowances, to individual orders and customers in order to calculate realized profit or loss by order or by customer. Using this information, companies can periodically (e.g., every quarter) calculate an operating income statement for every customer. Furthermore, companies can use the activity-based information on MSDA costs to base salesperson incentives on order and customer profits, not just sales. Randolph Company can also evaluate whether it wants to continue offering free freight to Customer 1.

6-30  (a)		Customer 1	Customer 2	Customer 3	Customer 4
	CLV	$42.47	−$253.94	−$129.14	$27.73

Supporting calculations appear below. The customer lifetime value, CLV, is calculated by summing [(Mt − ct) × (rt)t − 1]/(1 + i)t across each of the t years retained, where i = the cost of capital, and then subtracting the initial acquisition cost. In this problem, i = 0.1. Calculations were performed in Excel and rounded, leading to the slight discrepancy in the total.

          Customer 1:

t	Mt	ct	(rt)t − 1	(1 + i)t	[(Mt − ct) × (rt)t − 1]/(1 + i)t
1	$275	$0	1	1.1	$   250.00
2	275	0	1	1.21	227.27
3	275	0	1	1.331	206.61
4	275	0	1	1.4641	187.83
5	275	0	1	1.6105	170.75
					$1,042.47

CLV = $1,042.47− $1,000 = $42.47

          Customer 2:

t	Mt	ct	(rt)t − 1	(1 + i)t	[(Mt − ct) × (rt)t − 1]/(1 + i)t
1	$300	$0	1	1.1	$272.73
2	300	0	1	1.21	247.93
3	300	0	1	1.331	225.39
					$746.06

CLV = $746.06 − $1,000 = −$253.94

          Customer 3:

t	Mt	ct	(rt)t − 1	(1 + i)t	[(Mt − ct) × (rt)t − 1]/(1 + i)t
1	$275	$0	1.0	1.1	$250.00
2	275	0	0.9	1.21	204.55
3	275	0	0.81	1.331	167.36
4	275	0	0.729	1.4641	136.93
5	275	0	0.6561	1.6105	112.03
					$870.86

CLV = $870.86 − $1,000 = −$129.14

Customer 4:

t	Mt	ct	(rt)t − 1	(1 + i)t	[(Mt − ct) × (rt)t − 1]/(1 + i)t
1	$275	$50	1	1.1	$  204.55
2	275	25	1	1.21	206.61
3	300	0	1	1.331	225.39
4	300	0	1	1.4641	204.90
5	300	0	1	1.6105	186.28
					$1,027.73

CLV = $1,027.73 − $1,000 = $27.73

(b)     Comparing Customer 1 to Customer 2, even though Customer 1’s net margins (margins minus costs to serve) per year are smaller, the longer time period more than compensates for the smaller net margins, yielding a larger CLV. Contrasting Customer 1 and Customer 3, we see the importance of retention. A slight decrease in the retention rate changes the CLV from positive to negative. In comparing Customer 4 to Customers 1 and 3, note that Customer 4 has same total (M − c) as Customers 1 and 3. The additional cost to serve and retain Customer 4 in years 1 and 2 provides net benefits in the form of higher margins in later periods, yielding a positive CLV that is only slightly lower than Customer 1’s. Contrasting Customer 3 and Customer 4 again illustrates the importance of the retention rate. The additional cost to serve and retain Customer 4 in years 1 and 2 provides net benefits and helps create the difference between a positive CLV for Customer 4 and a negative CLV for Customer 3.

(c)      Assuming that n is very large and the numbers in the table remain about the same each year, the infinite horizon CLV is (M − c)/(i + [1 − r]) minus the acquisition cost.

		Acquisition
Customer	(M − c)/(i + [1 − r])	Cost	CLV
1	$2,750	$1,000	$1,750
2	3,000	1,000	2,000
3	1,375	1,000	375

(d)     Companies can use the information gathered to calculate customers’ estimated lifetime value to target customers with high expected lifetime value. The critical parameters for calculating lifetime customer value are

• Initial acquisition cost
• Profits or losses earned year
• Any additional costs incurred to retain the customer each year
• The duration of the relationship

Some companies have highly sophisticated analytic systems that allow them to estimate these parameters based on the demographic characteristics of a potential or newly-acquired customer. The analytics help guide the companies’ promotion strategies and campaigns to attract customers with the highest expected lifetime value. For example, RBC Financial Group in Canada uses an analytic model of a customer’s future profitability based on age, tenure with the bank, number of products and services already used at the bank, and the customer’s potential to purchase additional products and services, grow account balances, and generate fee-based income. The bank assigns a personal account representative to its estimated high lifetime value customers, ensures that their phone calls get picked up quickly, and provides them with ready access to credit at attractive terms.

6-31  The net promoter score is likely to have the greatest predictive power for repeat purchases and growth in business-to-customer settings where customers have frequent interactions with companies. The score is likely to have the least predictive power in business-to-business settings where purchasing decisions are made by highly sophisticated professionals. In this case, it is better to ask, “How likely is it that you will continue to purchases products or services from Company X?”

CASES

6-32  The responses below are based on “Survival Strategies: After Cost Cutting, Companies Turn Toward Price Increases,” by Timothy Aeppel, The Wall Street Journal (September 18, 2002, p. A1).

(a)     Jergens’ president based the price on what he determined to be the cost of producing the order of 10 odd-sized fasteners from scratch. The cost included setup for the odd size and overtime labor. The company actually produced the odd-sized fasteners by producing full-size fasteners and then shortening 10. This method was less costly than setting up the equipment to run a small batch of the required odd size.

(b)     Goodyear had been rewarding its sales force based on volume, providing an incentive for the sales force to deeply discount prices to large distributors. The discounts were so substantial that the large distributors could resell the tires to smaller distributors (even with transportation costs to other regions), reducing Goodyear’s sales at higher prices to smaller distributors. Goodyear responded by cutting the discounts to large distributors, removing discount approval authority from the sales force and transferring it to a “tactical pricing group” that determines whether Goodyear can profitably match a competitor’s prices. Goodyear also modified its sales force bonus scheme to include a “revenue per tire” metric.

(c)      Emerson discovered that customers were willing to pay about 20% more than Emerson’s initially proposed cost-based price of $2,650 for a new compact sensor. Emerson priced the sensor at $3,150. Note that the article does not provide information on how Emerson determined product costs that it used as a basis for its markups. A traditional cost system is more likely to undercost a low-volume or customized product because it allocates manufacturing support costs to products based on unit-level drivers. An activity-based costing system more accurately assigns costs based on resource usage.

(d)     Wildeck, “a maker of metal guard rails, mezzanines and material lifts for factories and warehouses,” promoted packages that included installing its products. The installations bring higher profit than parts catalog sales. Wildeck responded to a competitor’s lower-priced storage-rack protector by developing its own “lite” version and pricing it much lower than the competitor’s price. When customers called about purchasing the lite version, they were informed of the benefits of the original version, and most of these customers bought the original version. An accurate costing system, such as a good activity-based costing system that includes both manufacturing and nonmanufacturing costs of providing goods and services to customers, provides reasonably precise information to managers for making decisions about the mix of products and services to offer to customers and prices to charge in order to generate the desired level of profitability.

(e)      Union Pacific introduced a minimum price that was higher than a third of its customers paid. The company was not concerned if it lost these customers because customers who were paying higher prices would fill up the newly free space. Dropping unprofitable customers will not lead to an immediate increase in profit if the associated capacity-related costs are committed costs and the resources cannot be put to other profitable use.

6-33  Midwest Office Products (HBS Case 9-104-073)

Midwest is a revised version of Dakota Office Products (HBS Case No. 102-021), designed to introduce students to time-driven activity-based costing. Midwest is a constructed example, but one based on the actual experiences of several medical distribution and office products companies. It simplifies many of the complex processes that would arise in an actual application, but the simplification enables students to construct a time-driven ABC model for Midwest in a relatively short amount of time. Instructors can discuss with the class how to extend the model to capture more complex and realistic situations. The costing of five representative orders gives students insights about the power and management implications of the approach.

 

The case is simple enough that it can be used as a first activity-based costing case. Given that the time-driven approach is both simpler and more powerful than original ABC, I now prefer to introduce students directly to this approach, rather than introduce ABC in its original form, and then migrate them to the time-driven approach.

Teaching Objectives

Introduce students to time-driven activity-based costing through a simplified numerical exercise based on actual company situations. Enable students to propose actions that would transform unprofitable orders and customers into profitable ones.

Assignment (for the standalone case)

Read

“Time-Driven Activity-Based Costing”   Harvard Business Review reprint #R0411J

“Activity-Based Costing and Capacity”   HBS No. 105-059
Optional: ABC Pen Factory Tutorial       HBS No. 103-704

The Kaplan–Anderson Harvard Business Review article on Time-Driven ABC (#R0411) should be sufficient to introduce students to ABC even if they have not previously studied the subject. Should instructors want a more comprehensive introduction to “original” ABC, they can consider: R.S. Kaplan, “Introduction to Activity-Based Costing,” HBS Note No. 197-076 (Boston: Harvard Business School Publishing, 1997).

The technical note on ABC and capacity can be optional or background reading. It helps students understand why ABC cost driver rates should be based on practical capacity, not actual utilization. This may shorten discussion that otherwise arises in class about what happens to cost driver rates after order mix changes and process efficiencies reduce the demand on organizational resources. The ABC Pen Factory Tutorial provides an on-line demonstration of why companies can benefit from ABC. It shows how ABC captures cost behavior better than traditional, volume-based costing.

Textbook Assignment Questions (adapted from the original case questions):

• Based on the interviews and data in the case, estimate:
• The cost of processing cartons through the facility
• The cost of entering electronic and manual customer orders
• The cost of shipping cartons on commercial carriers
• The cost per hour for desktop deliveries

• Using this capacity cost rate information, calculate the cost and profitability of the five orders in Exhibit 6-10. What explains the variation in profitability across the five orders?
• On the basis of your analysis, what actions should John Malone take to improve Midwest’s profitability? Include suggestions for managing customer profitability.
• Suppose that currently, Midwest processes 40,000 manual orders per year, with a total of 200,000 line items to enter, and processes 30,000 electronic orders.
• How much unused practical capacity does the company have?
• If the company’s efforts to encourage customers who order manually to change to electronic ordering results in 20,000 manual orders per year, 100,000 line items to enter, and 50,000 electronic orders, how many order entry operators will the company require?  If order entry resource costs can be reduced in proportion to the number of employees, what will be the cost savings from the changes? 
• Returning to the original information in (d), if the company’s process improvement efforts result in a 20% reduction in time to perform each of the three order entry activities, how many order entry operators will the company require?  If order entry resource costs can be reduced in proportion to the number of employees, what will be the cost savings from the process improvements?
• Discussion

Introductory Remarks

 

If this is the students or executives’ first exposure to activity-based costing, I often start with the data in TN Exhibit 6-1, taken from a company I consulted with in 1995. While flying to the company, I thought about how to motivate the discussion with the senior executive team about whether to adopt activity-based costing. In reading the company’s annual report, I saw the 10-year summary, which I have excerpted in TN Exhibit 6-1. During the past ten years, the company was fortunate to have had a sales increase of nearly 135%, as sales had increased from $2.08 billion to $4.89 billion. Conventional cost accounting thinking focuses on distinguishing between fixed and variable costs. If this company had any significant fixed costs during a period when sales increased by 135%, the company’s main problem would be the frequency with which it would have to deposit cash into its bank accounts. But as the study of cost of goods sold and gross margin revealed, the CGS—rather than decreasing as a percentage of sales, as it would have to if any fixed costs existed—actually increased as a percentage of sales. So not only were CGS not fixed, they were not even variable (that would have produced a constant gross margin percentage). In fact, the CGS were super-variable costs, rising faster than sales.

A similar result was found in “below the gross margin line” costs—selling, general, and administrative costs. Despite many accountants treating these as “fixed costs,” these SG&A costs were also super-variable, increasing faster than sales. These finding set the stage for the ABC cost model, trying to understand why companies, operating with a traditional volume-based cost system, typically increase their sales by offering more variety, features, and service, leading to costs rising faster than sales. An ABC model helps a company identify when it has not recovered all the incremental costs associated with an order that has special features, such as small order quantities, expedited ordering, and customized delivery options.

 

Analysis

Q: What are the operating processes at Midwest Office Products?
I try to get the students help me get the simple process flow model below on the top of the board:

Process

Students may also comment on the differential lengths of time that customers take to pay.

I then describe how to build an ABC model for the three processes performed by Midwest:
Step 1:        Assign costs to processes
Step 2:        Calculate practical capacity of processes
Step 3:        Calculate cost rate of supplying capacity
Step 4:        Assign costs to orders or products based on their use of capacity

I start with the warehouse process to receive and store a carton:
(a) (1)
Q: What does it cost to process a carton through the distribution center?
The cost of the warehouse personnel is calculated by subtracting the cost of the delivery personnel (not involved with warehouse operations) from the total distribution center personnel cost. To this cost must be added the cost of operating the warehouse.

• Cost of process:                        $2,320,000 + 2,000,000 = $4,320,000
• Practical capacity                                                        ¸ 80,000 cartons
• Cost rate for supplying capacity                           = $54 per carton

You should ask what assumption is made by this calculation. The answer, of course, is that all cartons place the same demands on warehouse resources. The follow up question, therefore, is when would this assumption be violated? The hoped-for responses are:

• Cartons can be of varying size, and therefore occupy differential amounts of space
• Cartons can stay in storage for varying amounts of time, consuming differential quantities of days of cubic storage space
• Cartons can require different quantities of time by warehouse personnel to process; as a specific example, a customer order may ask for a less than carton quantity to be shipped. In this case, a warehouse person has to break open the carton, pick the desired quantity of materials, and then package them up to be shipped to the customer. Such breakpack quantities are far more expensive to process per unit than carton-sized quantities.

All of these complexities can be handled by a slightly more complex ABC model. The analyst could treat the warehousing function as consisting of two sub-processes, one for supplying space (cost per cubic meter per day) and one for supplying warehouse personnel time (cost per hour). In this way, products that take up much space and stay in storage longer will have higher costs assigned to them than products packaged in small cartons that turn over quickly in the warehouse. Similarly, breakpack quantities will have more costs assigned to them than standard, carton-sized quantities. To introduce such complexity into the case would lead to students (and the instructor) spending more time building the model, which is why I did not incorporate them. But the instructor should point out how the cost model could be easily expanded to incorporate these more realistic situations, as long as the company had data on carton size, product-specific turnover rates, and actual quantities ordered—as it almost always would.

Q:  What else is missing from the calculation of warehousing cost?
We do not have enough information to assign some general and administrative costs related to warehouse personnel. General and selling (G&S) costs are reported as a single line item in the income statement, and the project team did not collect information about how these might be traced to order handling costs. This was a deliberate omission in the case. Some students will allocate these costs as a percentage markup over the $54 per carton warehouse processing costs. I then question the rationale for such a procedure, asking them what assumption this allocation method makes. The answer is an assumption that G&S costs are proportional to warehouse processing costs, which they almost surely are not. I try to get the class to understand that using percentage markups to allocate indirect and support costs is indefensible. The percentages bear no causal relationship back to the cost incurrence. If instructors follow Midwest Office Products with a case on customer profitability, they can point out that some G&S costs relate to customers not to processing orders, and that this cost assignment will be discussed in a subsequent case.

The instructor has the option to give a little lecturette at this point, pointing out that a more complete costing model would perform an ABC analysis for each component of G&S expenses—such as human resources, finance, and information technology costs—so that the costs of these departments can be more accurately driven directly to processes based on the demands that the processes make on each of the corporate support departments. The instructor can go back to an example reported by Tom Johnson of how Weyerhaeuser assigned its corporate support departments to its business units using cost drivers, not percentages.

(a) (2)

Customer Order Entry

Having explored approaches for driving warehousing costs to orders, the instructor can now turn to estimating order entry costs. Again the four-part numerical sequence should be straightforward to elicit from the class:

Calculate cost per hour for the resources doing the order entry

• Cost of resources performing order entry                   $840,000
• Capacity of order entry resources: 16 ´ 1,500 hours =          24,000

 

• Cost of order entry process           $35/hour (or $0.583/minute)

• Assign costs based on use of resources

Entering a manual customer order: 0.150 ´ 35 = $5.25/manual order
Entering a line item on order  0.075 ´ 35 =   2.625/line item
Validating an EDI order         0.100 ´ 35 =   3.50/EDI order

As with the previous discussion (on warehousing costs), the instructor can ask the class about what assumptions are being made by the above calculation. Again, G&S costs are excluded because we don’t have sufficient information to assign them analytically. The core assumption is that all electronic orders take the same amount of time to validate or verify, and all manual orders take the same amount of time to enter the basic information, independent of the complexity of the order (other than the number of line items in a manual order). More realistically, some orders may be more complex to handle: expedited orders, orders requiring special packaging or delivery instructions, and, in general (beyond the Midwest case example), international orders, orders of fragile items, etc. This is a good time to review the concept of time equations, discussed in Chapter 5 and the Kaplan-Anderson HBR article, as an easy and effective tool for handling order complexity. TN Exhibit 6-2 shows a slide that can be used to illustrate time equations. Emphasize with the class that this type of complexity was suppressed in the Midwest case to reduce class preparation time. The other important point to be made about time equations is that the information on order characteristics is generally available from the company’s ERP or automated order entry system, so that each order can be individually costed based on its specific characteristics. This feature enables time-driven ABC to incorporate far more process complexity and diversity than original ABC, which would require separate activity pools to be set up to handle each major variation in order diversity and complexity.

If time permits the instructor can also work with the class to learn how to update the time-driven ABC model: This material is optional to cover, especially in a 75-minute class.

Q:  What causes a cost rate ($/hour) or unit time estimate to change?
Suppose through a business process improvement (such as TQM, or 6 sigma), Midwest improves the process so that manual customer orders can be handled 20% faster. It can then just reduce the unit time estimate by 20% putting 0.15 × (1 −.20) = 0.12 into the ABC program or Excel spreadsheet.

If Midwest grants its order entry clerks a 6% raise, modify the resource cost and re-estimate the cost per hour (from $35/hour to $37.10/hour). Similarly, if Midwest grants them an extra holiday or vacation day, reduce the practical capacity of these resources, and recalculate the cost per available hour. Students should see how simple it is to keep a time-driven ABC model current, by modifying hourly cost and unit time estimates based on known shifts in the process.
The class can now turn to costing the third major process, distributing orders to customers. The instructor can ask:

Q: What does it cost to ship a carton using commercial delivery?
The four-step calculation is simple to perform:

• Cost of resources for commercial shipments              $450,000
• Quantity of cartons shipped                                      ¸ 75,000
•  3. and 4.  Calculate shipping cost per carton                  = $ 6/carton

Midwest has outsourced its standard shipping arrangements (rather than performing this process with internal resources). Therefore, the unit cost does not involve calculating a cost of supplying capacity to perform this process.

The case example was simplified to assume an identical cost for each carton shipped. In general you could direct charge the actual shipping expense to each order without much difficulty by linking the shipping invoice to the order. If you had different shipping arrangements—expedite, prepare for overseas shipment, special handling—you would again estimate specific times using time equations.

Having calculated the unit cost for commercial shipments, the class can now perform the final cost estimate, for distributing cartons directly to the customer’s location.

(a) (4)
Q: How much does a desktop delivery cost?

• Cost of desktop delivery resources: $200,000 + 250,000  =           $450,000
• Number of hours available for DD = 4 × 1,500                           ¸ 6,000
• Cost rate for desktop delivery                                               = $75/hour

Students may propose two alternative methods for calculating the cost of desktop deliveries. One option, based on an original ABC calculation, assumes an identical cost per carton for desktop delivery:

$450,000/2,000 DD’s = $225 per DD

The error in this calculation is that it doesn’t allow for variation in desktop deliveries, some may be close others may be far away. Also, some cartons may be part of a large order delivered to a customer location, while others may be for a sole delivery of a carton. The actual delivery cost varies by number of deliveries, not the number of cartons in a delivery. Thus the original ABC calculation above contains two sources of error.
The other method estimates a standard cost per delivery. This allows number of deliveries rather than number of cartons to be the cost driver. With 2,000 deliveries, the standard cost is $225 per delivery. While this calculation avoids the second problem above, it still assumes that all deliveries are equally costly, thereby suppressing the different costs associated with delivering to nearby versus distant customers. The preferred procedure, which is both accurate and simple to implement, is to use the $75 per hour costing rate, and calculate the time required for each delivery. You can point out that today’s technology and information systems (such as GPS to track trucks in real time and hand-held computers that truck drivers can use to log in upon arrival and departure at a customer’s location) enable Midwest to use cost drivers that are more specific and accurate for individual customers, rather than averages! In general, Midwest could even track the time used at the customer’s site if it had to deliver to multiple locations.

At Kemp’s, the dairy company described in the sidebar of the assigned HBR article, drivers have computer entry device in their trucks. They record when leaving a company’s warehouse, when they arrive at the next site, and when they leave that site. The data are automatically uploaded to Kemp’s ERP and costing system.

For desktop deliveries, orders are costed only for the capacity they use. Suppose the same $450,000 of resources for desktop delivery were supplied during the year but only 1,500 desktop deliveries were made. Then the traditional cost per DD would have been calculated at $300 per direct delivery, not the $225 previously calculated. So the cost assigned fluctuates with capacity utilization, which is not a good idea. With the time-driven approach, the cost rate automatically stays at $75/hour and orders are charged only for the hours they use. The cost of unused capacity is reported each month—for management attention. This is subject of the technical note, which can be assigned as optional reading for the class.

At this point, the instructor can step back from the calculation details to get the class focused on the big picture, about when and why companies benefit from having a more accurate costing system.

Q: When is it worth tracking this much detail to customer orders (the cost of processes and the unit times used on the process by various orders or deliveries)?

• The Willie Sutton rule: Look for areas with large expenses in indirect and support resources, especially where such expenses have been growing over time. Operations where almost all expenses are direct labor and direct materials, which can already be directly traced to individual products by traditional costing systems, may have less need for ABC systems.
• High Diversity rule: Look for a situation in which large variety exists in products, customers, or processes. For example, consider a facility that produces mature and newly introduced products, standard and custom products, and high-volume and low-volume products. For marketing and selling expenses, companies may have a mixture of customers who order high-volume, standard products with few special demands as well as customers who order in small volumes, special products, and require large quantities of pre-sales and post-sales technical support.

We are finally ready to cost the five orders. I recommend having a student estimate the cost for Order 1 and Order 2 and then hand out TN Exhibit 6-3, which contains the proposed solution, based on the model cost estimates

Order 1 and Order 2 seem identical, except that the price for Order 2 is 4% higher because it involves direct delivery. Thus Order 2’s gross margin is substantially higher than for Order 1. Companies operating with a traditional or contribution margin costing system will celebrate the success of their direct delivery option. But the ABC analysis reveals that Order 2 requires far more resources to fulfill, involving $300 of delivery resources, $4.38 more in order entry resources (because of manual order entry), and nearly $20 more in finance charges because of its long collection period. So rather than being the more profitable order, Order 2 has a profit margin of minus 40%, compared to the attractive positive 6.6% net margin from Order 1.

One can go into more detail by showing that the customer has willingly paid $24 extra for the direct delivery option, but Midwest has incurred $294 in higher delivery costs to offer this option. Direct delivery is part of Midwest’s differentiating strategy, to develop more sales and higher customer loyalty by offering additional features and service that the customer values (and is willing to pay more for). But a differentiation strategy is only successful if the value created by the differentiation (the $24 price premium) exceeds the company’s cost ($300) for providing the differentiating service. In this case, the differentiation strategy is a quite spectacular failure, at least for single carton deliveries.

Orders 3 and 4 are the same as Orders 1 and 2 except that they are scaled by a factor of 10. Order 4 now makes a positive profit contribution and it is breaking even on its direct delivery service; the incremental revenues of $24 per carton now generate $240 of additional revenues, which equals the incremental costs of $240 ($300 – $60) for direct versus commercial delivery of the cartons. Also, by paying in 120 days, rather than 30 days, its financing costs are nearly $200 higher than for Order 3. In addition, manual ordering costs decrease profits. So even with the higher scale of operations, with a gross margin $240 higher than for Order 3, Order 4 has less than half the profit margin of Order 3.

Order 5 illustrates the impact of the hidden costs associated with manual order entry and long payment terms. It is the same as Order 3, except for manual versus electronic ordering and payment in 120 rather than 30 days. These seemingly innocuous changes reduce the profit margin by nearly 50%, a very important issue for companies with small operating margins.

Q: What do you suggest Midwest should do, based on its new cost estimates?
I recommend grouping the recommendations under the three columns, shown below, without necessarily labeling the columns as you accept suggestions from the class.

Process Improve	Pricing	Order Acceptance/Modify Customer Relationships
Route optimization	Specific charges for DD based on number of drop points, distance (time) traveled	Reject small orders; establish minimum order size
Migrate customers to more efficient channels (EDI)	Discounts for EDI orders	Limit distances for DD
Improve efficiency of warehouse operations, and order entry process	Menu-based pricing	Standard delivery, packaging
Customer picks up at warehouse; price quoted FOB

Q: Suppose customers all switch to EDI orders; no price discounts, get same quantity and mix of orders? How does this affect Midwest’s profits next year?
Certainly, the cost of processing orders goes down, and many more orders will be profitable. But Midwest still loses the same $80,000 in the year. How can this be?

Process efficiencies or shifting customers to lower cost channels, reduces the demands made on the organization’s resources, in this case – the order handling people. But until action is taken to redeploy or reduce the supply of these resources, Midwest keeps paying for them! The actions in the table above will create UNUSED CAPACITY. Management must act to exploit this unused capacity.

Q: A similar question to the one above is, “What if Midwest eliminates 30K of Type 2 orders because they are unprofitable, but doesn’t get any new orders to replace them”
The naïve model, based on actual capacity utilization would re-calculate the per carton warehouse processing cost as

Warehouse processing cost = $4,320,000/50K = $86.40/order

The risk to Midwest is it now starts to drop Type 4 and Type 5 orders leading to a death spiral.  This is why resources should be costed at practical capacity, not actual utilization. It is better to keep the warehousing cost rate at $54 per carton and report 30K × $54 = $1,620K as unused capacity.

You can close the class by pointing out the next steps in extending the ABC analysis beyond the accurate costing of individual orders:

• Analyze and assign general and selling costs
• Drive selling and order costs to customer; calculate cost-to-serve individual customers 
• Customer profitability: aggregate all orders by individual customers
• (d)
• Using the order entry times stated in part (a) (2), if Midwest processes 40,000 manual orders per year, with a total of 200,000 line items to enter, and processes 30,000 electronic orders, then Midwest requires (40,000 × 0.15) + (200,000 × 0.075) + (30,000 × 0.1) = 24,000 hours per year, which equals the current practical capacity. Therefore, the company needs all 16 operators and there is no unused capacity.
•  
• Midwest will require (20,000 × 0.15) + (100,000 × 0.075) + (50,000 × 0.1) = 15,500 hours per year, which requires 10.33 operators. The cost per operator is $840,000/16 = $52,500. If order entry costs can be reduced in proportion to the number of employees, the cost savings will be $52,500 × (16 −10.33) = $297,500. If Midwest only hires full-time employees, it will need 11 operators and the cost savings will be $52,500 × (16 −11) = $262,500.

• Midwest will require (40,000 × 0.12) + (200,000 × 0.06) + (30,000 × 0.08) = 19,200 hours per year, which requires 12.8 operators. If order entry costs can be reduced in proportion to the number of employees, the cost savings will be $52,500 × (16 − 12.8) = $168,000. If Midwest only hires full-time employees, it will need 13 operators and the cost savings will be $52,500 × (16 −13) = $157,500.

TN Exhibit 6-1 Super-Variable Costs, Profitability Declines with Growth
Growth Company Sales and Profit Summary, 1985–1994

	1994	1993	1992	1991	1990	1989	1988	1987	1986	1985
Sales	$4,889	$4,388	$4,193	$3,915	$3,498	$3,122	$2,878	$2,483	$2,234	$2,083
Gross profit	$950	$870	$818	$727	$689	$626	$584	$508	$457	$432
Margin	19.4%	19.8%	19.5%	18.6%	19.7%	20.1%	20.3%	20.5%	20.5%	20.7%
S & A expense	$491	$454	$412	$364	$327	$293	$271	$229	$206	$196
S & A %	10.0%	10.3%	9.8%	9.3%	9.3%	9.4%	9.4%	9.2%	9.2%	9.4%
Sales (1989 = 100)	157	141	134	125	112	100	92	80	72	67
S&A expense (1989=100)	168	155	141	124	112	100	92	78	70	67

Source:  Robert S. Kaplan

TN Exhibit 6-2 Time-Driven ABC Incorporates Variety and Complexity by Use of Time Equations

Customer Service Time =  6 minutes

+ 3 minutes for special orders

+ 5 minutes if credit memo required

– 2 minutes if electronic order

+ 4 minutes if Customer xyz

– 1.5 minutes if Customer pqr

…

…

Data about specific order characteristics for time equations

come automatically from company’s ERP and CRM systems.

Source:  Robert S. Kaplan

TN Exhibit 6-3 Five Order Profitability

	1	2	3	4	5
Sales	$610.00	$634.00	$6,100.00	$6,340.00	$6,100.00
Cost of items purchased	500.00	500.00	5,000.00	5,000.00	5,000.00
Gross margin	$110.00	$134.00	$1,100.00	$1,340.00	$1,100.00
Processing cartons	54.00	54.00	540.00	540.00	540.00
Shipping cartons, commercial	6.00	0.00	60.00	0.00	60.00
Desktop deliveries	0.00	300.00	0.00	300.00	0.00
Process manual order	0.00	5.25	0.00	5.25	5.25
Process line items, manual orders	0.00	2.63	0.00	26.25	26.25
Validate EDI order	3.50	0.00	3.50	0.00	0.00
Interest on receivable	6.10	25.36	61.00	253.60	244.00
Total processing costs	$ 69.60	$387.24	$  664.50	$1,125.10	$  875.50
Order profitability	$ 40.40	($253.24)	$  435.50	$   214.90	$  224.50
Return on sales	6.6%	-39.9%	7.1%	3.4%	3.7%

Source:       Robert S. Kaplan