Gotcha! Activity Costing

Did you know that:

Overambitious efforts to gather data on activities in your operations will suffer diminishing returns

Some activities can be easily metered, such as the automatically clocked time-per-call measures of a call center. But many need to be estimated, either by surveying employees on how much time they think they spend on different tasks or by conducting more scientific time-and-motion studies. While Roadway works hard to get the numbers right, for example, “we do accept that there are approximations in almost everything we do,” says Chief Information Officer Robert Obee.

One big challenge is figuring out which activities are truly important and just how precise measurement should be.

Electronic components manufacturer Reptron has some customers whose products require the placing of components by hand, rather than by machines. They are charged for the labor but not the overhead of machinery they don’t require.

But determining how much labor cost for Reptron’s engineering staff should be assigned to each customer is fairly informal—based on periodic estimates from the engineers, rather than time cards. “Sometimes the cost of collecting data and measuring it outweighs the benefit you might derive,” warns Tom Mackedon, controller for manufacturing operations at Reptron.

Replicating department boundaries or general ledger categories misses the point

You don’t want your activity analysis to tell you how much the shipping department spent on maintenance.

Instead, you want to get at activities like “repair conveyor belt”—specific actions from specific parts of your business. You should be able to look at activities that might cross customer service and distribution to get at activities like “fulfill customer order.”

That takes cooperation. Gene E. Obrock, vice president of operations at Henkel Consumer Adhesives, admits his company failed on the first try. Formerly known as Manco, Henkel is the maker of Duck Tape brand duct tape.

Managers who were stuck on “the way we’ve always done it” expected all the change to happen in other departments. The project started over again after about a month. By convincing participants to focus on activities that affected the entire organization, Henkel successfully identified the cost of serving different retailers.

The data and the way it is analyzed are likely to be politically charged

Gary Cokins, director of industry relations for SAS Performance Management (formerly ABC Technologies), recalls being thrown out of a large commercial bakery when he was working as a consultant for EDS. He ran afoul of the sweet-rolls manager.

The bakery was assigning overhead costs based on the direct hours of labor for producing each product. But the activity-based costing (ABC) analysis changed that picture, particularly in sweet rolls. “We discovered his product was causing more machine breakdowns,” Cokins says. “Going in, his product was the most profitable, but by us being able to distinguish which ones were responsible for the most maintenance, the overhead costs were being piled up more and more in his area.” The sweet-rolls manager responded by convincing his boss to dump the ABC project and replace it with a quality-management program that would be less of a threat.

Projects that drag on too long are more likely to fail

Here, the danger is that benefits may not be obvious until managers see how their numbers look. “The death knell is when management changes,” says Barbara Roudebush, a product strategy manager at PeopleSoft.

That’s when projects that are taking too long to identify activities and their costs tend to be killed off, she says.

Raj Aggarwal, a Henkel board member, says vendors and consultants initially pushed plans that were too expensive and too ambitious. “It took a while for us to sit down with them and say, ‘Look, we don’t want a gold-plated system,'” he says.

Henkel made activity-costing part of a move to a broader enterprise planning system. That held down the technology investment and produced some early results that could be built on through further refinements to the costing model.