Which customers are most profitable?

Some customers are more profitable than others. Most executives understand that, but simply acknowledging the truth of that statement is not enough. If companies are to maintain profitability over the long term, managers must have an understanding of the profitability of each of their customers and their products. With that knowledge, they can segment customers according to their value and maximize the mutual benefits of their relationships. As Tom Blackstock, former vice president of supply chain operations for Coca-Cola North America, noted in a 2005 speech:

There used to be an old adage in many companies that we treat all of our customers the same. That just does not work any more. Trying to treat all customers the same will only bring everything down to the lowest common denominator, low enough so that everybody can be the same. ... By segmenting customers based on their value to you over time, you can increase the value you deliver to them in return and thus obtain their increased loyalty by providing customized products and services in a highly predictable fashion.¹

Blackstock went on to explain that companies can accomplish customer segmentation by evaluating individual customers' profitability. The way to do that, he said, is by deducting avoidable costs from revenues. He also noted the importance of considering various customers' strategic value in addition to their profitability.

Article Figures

[Figure 1] Assigning costs to customers and customer segmentsEnlarge this image

[Figure 2] Profitability by type of account: A contribution approach ($000)Enlarge this image

[Figure 3] Profitability by type of drug store account: A contribution approach ($000)Enlarge this image

[Figure 4] Profitability by type of account: A full-cost approach ($000)Enlarge this image

[Figure 5] Customer profitability vs. product profitabilityEnlarge this image

[Figure 6] Profitability of low-volume products compared to total product line ($000)Enlarge this image

[Figure 7] The contribution approach using flexible budgets ($000)Enlarge this image

[Figure 8] Source documents, management reports, and profitability reportsEnlarge this image

[Figure 9] Information required for segment profitability reportsEnlarge this image

Unfortunately, few companies are able to follow Blackstock's advice because many of them don't have the ability to measure customer profitability in a meaningful way. As a result, management does not know what an individual customer contributes to earnings per share. This deficiency—caused in large part by bad metrics and the lack of a holistic view of their organizations—leads managers to make decisions that are not good for customers or for shareholders.

Here is one example: When manufacturing plant managers are required to minimize per-unit costs, their facilities often produce more than customers require. This is because the manufacturer allocates fixed costs to units and treats them like they are variable costs. As a result, producing more units appears to make the perunit cost go down.

Eventually, companies drop prices to encourage customers to buy up the excess. Witness the experience of General Motors in the summer of 2005, when the company introduced the "Employee Discount for Everyone" program in hopes of unloading bloated inventories. The program achieved the desired inventory reduction, but the automaker's third-quarter loss was US $1.7 billion, and sales declined markedly in September and October.

Even parts of the organization that are supposed to be customer-focused—sales and marketing—may not do what is best for the customer or for shareholders. The problem is that employees in those functional areas are paid based on the amount of revenue they generate. Consequently, there is no incentive for them to consider the profitability of the customer to their firm or the impact of their own actions on the customer's profitability. It is all about selling.

Is there any way to prevent internal organizations from making decisions that undermine customer and corporate profitability? The solution is to align management's efforts by tying functional performance to profitability. It sounds easy, but it is not. Making that connection requires the creation of customer contribution reports by deducting avoidable costs from revenues. In this article, we will explain how to produce this type of report and why it provides more accurate and useful data than traditional accounting analyses.

Customer profitability by type of account
The first step toward identifying which costs should be deducted from revenue when developing customer profitability reports is to determine whether the cost is dependent on unit volume (see Figure 1). If the answer is "yes," then it is appropriate to charge the cost to the applicable customer or customer segment. If the answer is "no," then it is necessary to determine whether the resource that incurs that cost is dedicated to a specific customer or customer segment. For example, if Procter & Gamble were to develop a profitability report on Wal-Mart, this is where P&G would deduct the costs of the customer team that is dedicated to serving Wal-Mart. Those costs would include salaries, benefits, travel budgets, and so forth. The goal is to deduct from revenue all costs that would disappear if the revenue from that customer disappeared.

If the answers to both questions in Figure 1 are "no," then the cost is placed in the contribution pool. This pool represents joint costs that are not traceable to an individual customer or segment of customers. While no customer is required to cover a specific portion of these costs, all customers together must generate a net segment margin that covers these costs in total and provides the desired earnings per share. The size of the net segment margin will determine the relative contribution of each customer/customer group or product/product group from the standpoint of financial performance. This information, combined with estimates of future growth for each customer or segment of customers, will enable management to develop strategies that maximize long-term profitability. As noted above, the overriding rule is to only include in segment reports those costs that would disappear if the revenues of the segment were lost.

To see how managers can make better decisions using reports developed in this manner, let us consider the example of one division of a large corporation. The division, which sells its products through four types of retail customers, reported annual sales of US $42.5 million. Traditional accounting data showed a net profit of $2.5 million before taxes. (Note: All figures in this article are in U.S. dollars.)

Management believed that this profit was not adequate, but traditional accounting information provided few clues regarding how to improve it. In the absence of fact, managers from the various business functions within the division put forth conflicting ideas about how to improve profitability. The marketing manager wanted to increase the advertising budget in order to increase sales. The representative from finance argued that the company was spending twice as much as it should on advertising. The sales and marketing managers wanted to introduce more new products, but managers from manufacturing and logistics argued that there already were too many stockkeeping units (SKUs) and that the addition of more would make it impossible to achieve the necessary efficiencies in operations. Salespeople wanted lower prices, while the finance manager argued that prices were too low and should be raised by 5 percent to increase profitability by $2 million. All of them made suggestions for improving profitability that were based on their own experience, and all of them were convinced that they were right.

Without accurate information, it would be difficult, if not impossible, to determine which—if any—of the suggested courses of action would achieve the best results. For that reason, the division adopted a contribution approach to profitability analysis by customer type so that it could identify with precision where performance was inadequate.

As shown in Figure 2, drugstores, with 45 percent of net sales, were the largest of the four customer types. But at 15.7 percent, the drugstore segment's controllable margin-to-sales ratio was the lowest of the four. It was less than one-half that of the second most-profitable segment, discount stores (31.7 percent), and only 37 percent as large as the most profitable segment, department stores (42.1 percent).

However, at $3.1 million, the drugstore segment's controllable margin was substantial, and simply discontinuing sales to that segment was not an option. Instead, managers looked for the causes of the drugstores' low margin-to-sales ratio. An analysis of product profitability by customer segment showed that product mix was not the problem. What other factors could be responsible, then?

While considering that question, one manager noted that the division's drugstore customers were not all the same. They included national chains, regional chains, and small, independent pharmacies. Further segmenting the drugstores into these three groups revealed that national drugstore chains had a controllable margin-tosales ratio of 34.9 percent, which was almost as large as that of the grocery chains (36.9 percent) and better than that of the discount stores (31.7 percent). Regional drugstore chains (30.9 percent) were almost as profitable as discount stores. The most important discovery was that the division was losing $85,000 per year on sales to independent pharmacies (see Figure 3).

Further investigation revealed that the cost of serving the independents—including costs associated with small orders, overnight package service, third-party warehousing, carrying inventory, slow payment of invoices, and a high rate of bad debts—contributed to that loss. This information enabled management to estimate the effect on corporate profitability of costcutting efforts like scheduled deliveries and alternative means of serving the independent pharmacies, such as through drug wholesalers. Managers chose the alternative that would lead to the greatest improvement in long-term profitability—in this case, shifting the business to wholesalers.

Drawbacks of fixed-cost allocations
Rather than using contribution reports, the accounting systems in most companies allocate fixed costs to individual customer segments. This traditional accounting approach provides incorrect information because costs that are common to multiple segments are allocated on the basis of arbitrary measures of activity, such as sales volume. Consequently, the traditional system does not capture vital information about the controllability and "behavior" of costs. For example, if a segment of customers or products is found to be unprofitable and therefore is discontinued, the fixed costs allocated to that segment will simply be reallocated to the remaining customers.

Figure 4 shows how the customer profitability analysis in Figure 2 would change if calculated using typical methods of cost allocation. Drugstores would show a profit of more than $1 million, which is significantly larger than the profit for the other customer groups. At 5.6 percent, the profit-to-sales ratio for drugstore customers would compare favorably with the other customer groups. Under this accounting method, the ratio would be 82 percent of the profitto- sales ratio for grocery chains, whereas under the contribution approach, the segment controllable margin-to-sales ratio of the drugstores (15.7 percent) was just 43 percent of that earned by the grocery stores.

The two methods of accounting can result in even greater differences in the profitability of products because manufacturing, marketing, and logistics costs often vary more across products than they do across customers. If the drugstore customers in Figure 4 were analyzed by type of drugstore, the profit-to-sales ratios for the three groups of customers would be approximately equal because average costs would have been used to determine customer profitability.

Similarly, revenue minus avoidable costs can be used to measure the profit contribution of products. Once product contribution reports are available, managers can begin to answer strategic questions, such as whether there are products that should be eliminated; whether some product prices should be raised (typically, prices for products where contribution is low and demand is not sensitive to price increases); or whether some product prices should be reduced (for example, prices for high-volume products where a price reduction will increase demand and profitability). More attention can be directed to those products that are most profitable.

Managers in companies that have implemented segment profitability reports have been able to identify products and customers that were either unprofitable or did not meet corporate financial objectives. Ironically, many of these products or customers previously were thought to be profitable, due either to their large sales volumes or their manufacturing margins.

Figure 5 (the "challenge quadrant") shows how management can combine customer profitability and product profitability to identify customers to drop or prices to increase. This type of analysis also identifies where management should focus its attention and ensure the highest levels of quality and service: products that are highly profitable and are being purchased by the most profitable customers.

Figure 6 shows that, in many cases, only a few key costs can be used to identify underperforming products at the SKU level. In this example, Product Group A comprised 340 SKUs with total inventory of $6,309,800. Annual unit sales were 148,527, and inventory turned four times yearly. It was decided to consider every SKU with fewer than two turns to be a slow-moving item; 135 of the 340 SKUs in the product line fell into that category. Average inventory turnover for slow-moving products was 0.6, and they represented $1,823,900 (28.9 percent) of total inventory. The total product contribution generated by all 340 SKUs was $19,947,200, but the contribution from the 135 slow-moving items was just $149,500, or 0.7 percent of the total. The logical next step was to determine which customers were purchasing these products, and then conduct an analysis similar to that in Figure 5 to identify SKUs for elimination or for price increases.

In summary, cost allocations can distort profitability reports for customers and products. The consequences of such inaccuracies are great, as Robert S. Kaplan explained in a 1988 article in Harvard Business Review: "Seriously distorted product costs can lead managers to choose a losing competitive strategy by de-emphasizing and overpricing products that are highly profitable and by expanding commitments to complex, unprofitable lines. The company persists in the losing strategy because executives have no alternative sources of information to signal when product costs are distorted."²

Limitations of profitability reports
Research has shown that the segment profitability reports used by managers often have serious shortcomings because most of them are based on average cost allocations rather than on the direct assignment of costs at the time a transaction occurs.³ Period costs are allocated to customers and products using arbitrary bases such as direct labor hours, sales revenue, or cost of sales. Opportunity costs related to investments in inventories and accounts receivable are not included. Moreover, key marketing and logistics costs frequently are ignored.

Many of the problems encountered by manufacturing companies are the result of using a "full cost" approach, whereby indirect costs (such as overhead and general administrative expenses) are allocated to each customer or product. As a result, many managers use control mechanisms that focus on the wrong targets: direct manufacturing labor or sales volume. Reward systems based on these control mechanisms drive behavior toward either simplistic goals that represent only a small fraction of total cost (labor) or single- minded sales efforts (volume). They ignore more effective ways to compete, such as product quality, on-time delivery, short lead times, rapid product innovations, flexible manufacturing and distribution, and efficient deployment of scarce capital.

Because of the following factors, many managers do not know the true cost of their companies' products or services, the most effective way to reduce expenses, or how to direct resources to the most profitable customers:⁴

• Accounting systems are designed to report the aggregate effects of a firm's operations to stockholders, creditors, and government agencies.
• Accounting costs are computed to provide historical records of a company's operations. Since costs that are common to multiple segments are allocated, the process is necessarily subjective and arbitrary.
• Accounting systems typically record marketing and logistics costs in aggregated accounts. Seldom is there an attempt to attach those costs to individual products or customers. In addition, accounting systems typically are not capable of identifying these costs at the time the transactions are recorded. When the detail is captured, it frequently is lost because only aggregated data are carried forward to subsequent accounting periods. Once data have been aggregated, the only way to disaggregate them is by allocation methods that distort the profitability.
• Profitability reports do not show a segment's contribution to overall corporate profitability but rather include fixed costs, joint product/service costs, and corporate overhead cost allocations. Often top management encourages this approach, fearing that sharing variable cost information with the sales force would lead to lower selling prices. (However, prices should be determined by the marketplace and should not be based on costs.)
• In many standard cost systems, fixed costs are treated as variable costs, which masks the true behavior of the fixed costs.

Importance of accurate cost data
Accurate cost data are critical for the development of the product and customer profitability reports that management needs in order to direct corporate resources to the products and customers that will provide the greatest return on those resources. The accounting system must be capable of providing information to answer questions such as the following:

• How do marketing, operations, logistics, and research and development costs affect contribution by product, by territory, by customer, and by salesperson?
• How should resources be used to support customers and customer segments?
• What are the costs associated with providing additional levels of customer service?
• What is the optimal amount of inventory?
• What mix of transport modes and carriers should be used?
• How many field warehouses should be used, and where should they be located?
• How many production setups are required?
• Which plants should be used to produce each product?
• What are the optimum manufacturing plant capacities based on alternative product mixes and volumes?
• What product packaging alternatives should be used?
• To what extent should information systems be automated?
• What marketing channels should be used?
• What products should be dropped from the product line?

To answer such questions, management must know what costs and revenues will change as a result of the decision that is being made. That is, the determination of a product's (or customer's) contribution should be based on how corporate revenues, expenses, and hence profitability will change if the product (or customer) were to be dropped. Any costs or revenues that will be unaffected by this decision are irrelevant. For example, relevant costs might include sales commissions, promotional allowances, and transportation costs associated with a product's sales. Examples of irrelevant costs include the overhead associated with the corporate head office and security guards at the plant.

Segment profitability reports become more useful as a management tool when they are developed on a pro forma basis and actual results are compared to a budget, as shown in Figure 7. This analysis shows a level of aggregation that would be of interest to someone at a firm's top executive level. This report would allow an executive to see at a glance why targeted net income has not been reached. There is a $1.5 million variance due to ineffectiveness, which is a measure of the net income the company has forgone because of its inability to meet its budgeted level of sales. There is also an inefficiency factor of $0.7 million. The difference between $9 million and the actual outcome of $8.3 million is $0.7 million, due to inefficiency within the marketing and logistics functions. This analysis can be performed for segments such as products, customers, geographic areas, or divisions.

The key to successful implementation of a flexible budget is the analysis of cost-behavior patterns. To properly conduct that type of analysis, it is necessary to determine the fixed and variable components of costs. For example, regression analysis can be used to determine a variable rate per unit of activity and a total fixed-cost component. Once that information has been determined, the flexible budget for control becomes a reality. One caution: Cost estimates that are based on past cost-behavior patterns will contain inefficiencies. The predicted measure of cost may not be a measure of what the activity should cost but rather is an estimate of what it will cost, based on the results of previous periods.

Information system for measuring profitability
While substantial savings can be generated when management is able to compare its actual costs to a set of predetermined standards or budgets, there are even greater opportunities for profit improvement in the area of decision making. If management is to make informed decisions, it must have accurate data. For example, the addition or deletion of territories, salespeople, products, or customers requires knowledge of how well existing segments are performing and how revenues and costs will change with the alternatives under consideration. For this purpose, management needs a database that is capable of aggregating data so that it can obtain routine information on individual segments, such as customers, salespeople, products, territories, or channels of distribution. The system also must be able to store data by both fixed and variable components, so that management can identify the incremental revenues and costs associated with alternative strategies.

Much of the required data will come from the source documents generated by individual business transactions. These documents include customer orders, bills of lading, sales invoices to customers, and invoices from suppliers or vendors, to name a few. Source documents for internal transactions and activities also contain useful data; examples include a private fleet's trip reports and carrier bills of lading. Other relevant costs may be identified by means of standard cost systems, engineering time studies, or statistical estimating.

For this type of effort to be successful, the source documents must be computerized. One of the most promising computerized database systems for generating profit contribution reports is a central storage system in which source documents are fed into the database in coded form. Inputs can be coded according to function, subfunction, customer, territory, product, class of trade, transportation mode, carrier, revenue or expense, or a host of other possibilities. When combined with standard cost information, the database is capable of generating both functional cost reports and segment contribution reports.

The system works by charging functions, such as warehousing and transportation, with actual costs; these costs are then compared to predetermined standards. Individual segments, such as customers or products, are credited with segment revenues and charged the standard cost plus controllable variances. The system is capable of filing large amounts of data and allows rapid aggregation and retrieval of various modules of information for decision making or external reporting. Figure 8 provides examples of source documents as well as some of the management reports and profitability reports that can be generated from this type of database.

The database must be capable of collecting all of the revenues and costs for every transaction and aggregating them by functional activity (for example, selling, advertising and promotion, transportation, warehousing, and order processing). This technique is commonly referred to as responsibility accounting and is primarily used to develop annual budgets and monthly variance reports by major categories and subcategories of corporate activities, such as manufacturing, research and development, marketing and sales, and logistics. In order to implement this type of database, cost data must be recorded at the time of the transaction with enough detail to identify fixed/variable and direct/indirect components. The data must be sufficiently defined to permit the formulation of meaningful modules, such as those shown in Figure 9.

The profit contribution approach has been described in the literature for more than 40 years. Yet surveys of corporate practices indicate that there are relatively few integrated operating systems that report segment profits on a timely and accurate basis. Why does this situation exist?

For one thing, many managers mistakenly think that they should employ the same accounting practices that are used to value inventories and report results to the government tax agencies—that is, allocation of all costs—to generate reports for managing their businesses. For another, they may believe that using only variable and direct fixed costs might encourage suboptimal pricing by their salespeople.

Moreover, accountants frequently oppose a separate management accounting system, and top management and accountants may feel more comfortable if they can tie the cumulative results of the various segments to the company's total profit-and-loss data. Yet another reason is that managers often fail to recognize the behavioral differences of fixed and variable costs as well as the distinction between direct and indirect expenses. As a result, they fail to understand the usefulness and purpose of contribution reports. Finally, data-processing personnel often discourage the development of such reports by citing the difficulties in creating the databases and operating systems required to assign direct costs to specific product and market segments. In today's information technology environment, these reasons are no longer valid.

The knowledge to compete
In many firms, managers don't know which customers and products are profitable and which are not, because they lack the accurate data that are required to make that judgment. This is a serious problem. It is difficult to compete even when firms have good financial information; it is almost impossible to compete effectively with bad information. In other words, successful management of a business depends on a full knowledge of the costs and revenues associated with customers and products. This includes the costs associated with marketing, sales, operations, and logistics, all of which can have a significant impact on the profitability of a company's customers and products.

Without that data, the managers from various business functions will hold conflicting ideas about how to improve customer or product profitability. A contribution approach to profitability analysis by customer type and product can prevent that from happening, and it can identify with precision where performance is inadequate.

The importance of this knowledge cannot be overstated. As Michael Dell has said, "Until you look inside and understand what's going on by business, by customer, by geography, you don't know anything."⁵

Endnotes:
1. Keynote speech by Thomas Blackstock, vice president, supply chain operations of Coca-Cola North America, to the International Association of Food Industry Suppliers in San Francisco, California, USA, March 11, 2005.

2. Robert S. Kaplan, "One Cost System Isn't Enough," Harvard Business Review, Vol. 66, No. 1 (1988), pp. 61-66.

3. Douglas M. Lambert and Jay U. Sterling, "What Types of Profitability Reports Do Marketing Managers Receive?" Industrial Marketing Management, Vol. 16, No. 4 (1987), pp. 295-303.

4. Thomas S. Dudick, "Why SG&A Doesn't Always Work," Harvard Business Review, Vol. 65, No. 1 (1987), pp. 30-35; Robert S. Kaplan, "How Cost Accounting Distorts Product Costs," Management Account, April 1988, pp. 20-27; John J. Wheatley, "The Allocation Controversy in Marketing Cost Analysis," University of Washington Business Review, Vol. 30, No. 4, (1971), pp. 61-70; Ford S. Worthy, "Accounting Bores You? Wake Up," Fortune, October 12, 1987, pp. 43-50; and Douglas M. Lambert and Jay U. Sterling, "What Types of Profitability Reports Do Marketing Managers Receive?" Industrial Marketing Management, Vol. 16, No. 4 (1987), pp. 295-303.

5. Joan Magretta, "The Power of Virtual Integration: An Interview with Dell Computer's Michael Dell," Harvard Business Review, Vol. 76, No. 2 (1998), p. 77.

This article is adapted from Douglas M. Lambert, editor, Supply Chain Management: Processes, Partnerships, Performance, third edition, Sarasota, Florida: Supply Chain Management Institute, 2008, pp. 41-52. Copyright 2008, Supply Chain Management Institute. Used with permission. For more information see: www.scm-institute.org.