Activity-Based Costing, Business Intelligence, and Profitability
If your company sells a wide variety of products or services, it can be hard to tell which offerings are the most profitable and which aren’t pulling their weight. Accounting tries to address this by establishing costs for each product or service. But cost assignment is a notoriously difficult problem, because although it’s pretty straightforward to measure the amount of direct labor and materials that go into a pair of scissors, it’s not easy to establish how much overhead should be assigned to each one, if you also make staplers and shredders in the same factory.
The physical-plant workers who maintain the building, the cleaning supplies used by the janitorial staff, and the cost of carrying inventory are certainly components of the cost of making and marketing a product – but unless you can devise a meaningful way to assign these costs to each unit of each of the products you make, you may end up with a misleading idea of how profitable each product is.
This has troubled executives and cost accountants for centuries. Way back in 1832, no less an authority than Charles Babbage wrote that, “it is of great importance to know the precise expense of every process, as well as of the wear and tear of machinery which is due to it.[i] Babbage is best known as the inventor of a calculating machine generally considered to be the forerunner of the computer, but he was also an influential economic theorist, encouraging division of labor and standardization of tasks in manufacturing.
Traditionally, manufacturing overhead was spread evenly across all cost objects, a method known to accountants as “peanut butter costing.” This approach was used mostly because it’s a lot easier than figuring out how much of each indirect cost was really driven by the production of a given item. In Charles Babbage’s day, the lion’s share of manufacturing costs took the form of direct labor and materials; if indirect costs weren’t assigned all that precisely, it didn’t matter too much, because those costs were small. Today, however, when you have sophisticated, multimilllion-dollar machines being used to design and manufacture products, the indirect costs may greatly outweigh the direct labor and materials – so it’s much more important to get a handle on what’s driving those indirect costs, so you can assign them to the right products and get an accurate idea of the true cost of each unit of output.
The high distribution costs of small orders could be remedied by negotiating a minimum order size with customers, reducing the number of shipments.
Activity-Based Costing (ABC) is a way of accomplishing this, by providing a better breakdown of indirect costs and their relationship to goods (or services) produced. The basic idea is that overhead costs are broken down into “activity cost pools,” which are then proportionately assigned to the procedural “activities” that go into the making of each product/service. For example, product design, sales expenses, customer order processing, and post-sale customer service and support are activities that vary directly with the number of products sold – so they should, by rights, be counted as part of the cost of each product. But some products require much more of each than others – an order for 100 wedding invitations requires a lot more design and customer support than an order for 100 sheets of business stationery with envelopes, even though the direct labor and materials costs might be about the same. If you used peanut butter costing, your total design expense would be divided equally between the invitations and the business stationery – meaning that the former would be under-costed and the latter over-costed.
That’s a simple example, but in a big company, ABC can make a huge difference. It gets complicated, however, so experienced practitioners usually do it in small steps – and, often, only for certain departments or divisions where it’s important to assign costs precisely.
BI systems are increasingly being used in conjunction with ABC in manufacturing, retail, banking and financial services; all of these industries have substantial indirect costs associated with their products (both tangible off-the-shelf products and intangible products like bank accounts and advisory services). Getting these costs allocated properly is essential for profitability analysis. By surveying each activity area, you can establish the cost allocation rules – so that, for example, the wedding invitations are getting assigned 65% of the design and 80% of the customer support cost, with the business stationery receiving the other 35% and 20%, respectively. Once you have those proportions established, you can really get an accurate picture of how profitable each product is. You can identify the less-profitable products and take action – for instance, negotiating with a customer to increase their order size, reducing the number of shipments, or you can investigate why a product has a high return rate. A BI system that lets you do impact analysis gives immediate feedback on the effects of such decisions – before you make them.
A BI system can help develop optimize a firm’s mix of product offerings, and with impact analysis, you can test the effect of proposed changes.
Profitability modeling can be applied not just to products, but to customers, product SKUs (stock keeping units), time periods, sales channels, days of the week (or months, quarters, etc.), and any other dimensions your enterprise data keeps track of. As with any other BI endeavor, the key is to get the best possible input from the business when developing an activity-based model. Having accurate cost-driver information will pay off in higher profits – so it’s worth putting in the time up front to make sure your models accurately reflect the way your organization works.
By John Kafalas
This monthly column covers Business Intelligence and data analysis issues. If you have questions, comments, or topic suggestions, please contact the author.
Copyright ©2014 4Sight Technologies, Inc. All rights reserved.
 Babbage, Charles. On the Economy of Machinery and Manufacture, London, 1832, quoted in Alexander Hamilton Church, The Distribution of Expense Burden, New York, 1908.