Getting Your ABCs Right

Tom Peters

Let’s say your tooling center racks up costs of $1000 in an accounting period; its work is divided equally ($500 each) between two products. Suppose Product A sells 200 units during the reporting period; Product B, 50 units. When it comes time to figure product profitability, tooling costs are clear: Product A (200 units) gets tagged with its $500 in tooling, which works out to $2.50 per unit. Product B (50 units) absorbs the other $500, or $10 per unit.

But that “obvious” answer is probably not the one you’ll get from your accounting system. Sad to say, most traditional systems would compute “250 total units, $1000 in tooling charges, $4 per unit.” Then when you assess product performance, these costing schemes end up assigning $800 in tooling costs to Product A (200 units times $4 per unit), and $200 to Product B (50 units, $4 per unit)!

This silliness is one of the hundreds of cases consultant-practitioner Peter Turney examines in his new book, Common Cents: ABC Performance Breakthrough. The “ABC,” now at work in many firms, is activity-based costing. The premise: Assign costs to “stuff” (activities) according to what is actually done.

The pressing need for ABC arises partly from accounting systems’ tendency to spread all indirect costs across products based on the direct-labor hours the product requires. That was satisfactory in a bygone era, when direct labor was a firm’s biggest expense. But now, direct labor seldom amounts to more than 15 percent of total costs. Typically undifferentiated overhead—order entry, design, marketing, inspection, tooling—comprises the bulk of expenses.

Consider another case: Each batch (100 units) of Product C requires four direct-labor hours and has to be inspected once; a batch of Product D (also 100 units) uses two direct-labor hours, but must be inspected twice. Suppose the cost of one inspection is $50. Product C clearly should get tagged with an inspection cost of $50 per batch of 100 units; Product D, with two inspections, has an inspection cost of $100 for every 100-unit batch. But don’t bet on it!

Suppose you produce one batch of each product. You incur $150 in inspection costs and consume six direct labor hours in all. Traditional systems would usually calculate inspection costs at $25 per direct-labor hour used, and would end up assigning inspection costs of $100 to Product C (four direct-labor hours times $25 per hour), $50 to Product D (two hours times $25). That is, Product C’s unit cost of inspection, assigned on the basis of direct-labor hours, will be double, not half, Product D’s. Reality be damned.

Turney leads us through true tales much more horrific than this. One situation involved a 25,000 percent costing error: A product with an “accounting cost” of $2 had a true cost of $500! Such miscalculations can lead to big strategic blunders. Take a routine instance of standard vs. customized product costing. Customers for standard products buy large quantities and usually make few demands on the producer’s system. Customers for non-standard products demand a lot. But normal accounting systems routinely assign the same unit cost to both types of product; they lump together all overhead costs, then distribute them evenly among all units, standard or non-standard. Since we charge more for the customized product, its margin looks great, and the marketing department ends up pushing it. In the real world, Turney points out, those non-standard products require “long setups, intense engineering support, heavy numerical-control programming support, substantial sales support, more order (entry) activity, more scrapped units, more inspection, and higher inventory.” But these additional costs are not ordinarily assigned to the special product. Do your sums correctly, says Turney, and you may find you’re losing a bundle on those customized products, despite their higher prices.

Finally, the saga of “a simple little brass strip.” Depending on which product model a company makes, a brass strip used for connecting wires to a grounding circuit requires one, two, or three holes, and is shorter or longer accordingly. Engineers propose using a single strip with three holes, which could serve all models. Turney says the idea probably makes sense, because only one part would have to be ordered, received, stored, and released to the shop floor. Just one part would have to be changed in the future. And it would be necessary to forecast volume for only one part; less inventory would be needed, etc.

But that’s not the way it turned out in a real-life case. Conventional costing concludes that material for a three-hole strip is more expensive than material for a one- or two-holer and punching three holes takes more effort than punching one hole. Case closed. All the “other” costs associated with producing three parts rather than one are spread equally across all units or are handled as a single administrative expense. Make three strips, the standard system mandates.

To be sure, many hot issues such as speed and quality are jostling for the top spot on the “keys to competitiveness” list. But Turney is not alone in contending that a correct costing scheme is the essential ingredient of competitiveness. He’s got a point.