A PRACTICAL
LOOK AT
ACTIVITY-BASED COSTING

How can finance make better contribution to the business?
Today, the answer is clear:
undertake activity analysis and implement activity-based costing and
new performance measures, thereby helping to refocus business activities on
the critical success factors.

by Paul Sharman

Cost leadership and cost effectiveness are key elements of business competitiveness. Management needs accurate information for decision making, and management accountants have traditionally been the providers of such information.

An approach known as activity-based costing (ABC) can provide management with relatively accurate overhead attribution to product cost and therefore a better understanding of profitability. ABC also provides cost information on major activities undertaken by the business as well as the cost of business processes, all of which can be captured in a comprehensive cost model relatively cheaply.

The value of this analysis is that it provides information on the economic structure of the business so profoundly different from all previous understanding that it will influence change. Just to satisfy the accountant in each of us, activity-based costing fully integrates into the existing financial reporting structure. It also indicates how we can substantially improve our reporting processes.

ABC systems use direct cost information in the same way as traditional systems, which are detailed engineered bills of material and labor. Where ABC systems differ is that they create a bill of activities for overhead. The concept of a bill of activities is a little taxing when you consider functions such as purchasing, business systems, materials planning, human resource functions, and accounting. The way to think about solving the problem, however, is with a simple rule: cost activities, not products.

In a pilot activity-based costing project, Northern Telecom used an approach and software developed from the research works of Bob Kaplan and Robin Cooper of Harvard University. The ABC system was based upon a two-stage indirect cost tracing procedure in which costs are first traced to activity centres and then to products.

Traditional cost systems assume all overhead activities are consumed equally by all products relative to volume produced. Further, all costs are allocated to products because the system assumes that current output drives current overhead costs. Consider how the costs of people involved in redesigning plant layout to accommodate new products or engineers working on updated drawings of limited-production, old products are charged to overhead. Overhead costs are allocated to products on the basis of the product's demand for some volume variable direct cost, usually labor hours, machine hours, or materials cost. But none of these bases individually represents the actual overhead incurred to make the product. Conventional thinking holds that the inaccuracy is not relevant because in total all costs are accounted for, and on average the relative distortion in margin reporting can not be significant.

Activity-based costing, by contrast, identifies what activities are performed by the overhead organization and calculates the cost incurred to perform each activity. Costs are traced to products on the basis of the individual product's demand for these activities throughout the process of converting raw materials, energy and human enterprise into the finished article. The allocation bases used in ABC, then, are the quantification of activities performed. These might include hours of set-up time or number of times handled. Such measures are described as cost drivers.

Cost drivers link activities together. Thus if a batch of 20,000 units of product to be manufactured required 100 different components, the size of the order and the number of components would drive the costs associated with purchasing, receiving, inspecting and handling the material among the many factors of production. The cost of each activity can be calculated by determining the amount of effort required by function and its associated cost to arrive at the total batch cost which, divided by the number of units, will yield a unit value.



Not only has the nature of the allocation base changed but so has the number. Instead of at most three allocation bases in the traditional system, an ABC system uses many. The initial Northern Telecom system identified 44 cost drivers. This makes the system data-intense and somewhat complicated, thus requiring a degree of effort to undertake and an investment in both time and resources. The payback, however, is relatively more accurate product costs. Exhibit 1 shows how substantial the change can be as costs are charged back to the product that caused them to be incurred.

Northern Telecom discovered that only 17% of products costed under ABC were within 20% of their original standard cost (material, labor and overhead). Five per cent of all products, notably high-volume items, cost less than standard, with three having been overcosted by as much as eight times. The remaining 78% of all products were calculated to cost more than 20% over standard. Exhibit 2 summarizes the distribution.

The magnitude of cost change at the Northern Telecom pilot was not unusual; in fact, the deviation curve has been observed in almost all ABCs. Keep in mind that the sole cause of change in product cost was a redistribution of overhead among the products. The pie of total overhead cost incurred remained the same.

Can it really be possible that a product could cost twice or, conversely, half the amount identified by our standard cost system? Analysis has shown that it often takes as much effort to buy a 1¢ article as it does a $2 one, and yet material-based loading rates do not make that distinction. Since activity-based costing takes all variables into consideration, it will by definition yield a relatively more accurate valuation.

What it all means

Typically, high volume and less complex products cost less than traditional standard cost systems would indicate. The reverse also hold true; relatively more complex and low volume products cost substantially more to produce. In most companies, 80% of sales are generated by 20% of products; clearly, they are set up primarily to produce high runner products economically. Hence, it makes sense that overhead distribution, premised today on an essentially arbitrary volume-based variable, is generally not representative - except on average.

Activity-based costing provides a relatively accurate product cost because it captures many dynamic variables. It will tell us that a product costs approximately $80-90 rather than precisely $41.6758 as a traditional system would. Activity-based costing systems, therefore, provide relative accuracy while traditional systems provide absolute inaccuracy, often calculated to four decimal places.

Obviously, if shifting from a traditional to an activity-based cost system changes our understanding of product cost, it follows that gross margin and hence our entire concept of product profitability is distorted. Often gross margins as a percentage of sales run in bands, say 10-40% or 30-80% dependent on the business using traditional costing, with high volume products at the low end of the scale. Activity-based costing totally contradicts this conventional wisdom and explodes gross margin bands, calculating bigger profits on the 20% or so of high volume products which the business is optimally configured to produce. ABC experience indicates that those 20% of high volume products produce, not surprisingly, 80% of total gross margin dollars; and that 60% produce 99% of total gross margin. These observations beg a lot of questions about the other 40% of products which produce 1% of the gross margin in total.

General managers sometimes comment that they know a special, low volume or customized product costs more than the cost system reports. Since there has been no way to find out how much more, however, the sale proceeds, often justified as strategic. The good news is, with activity-based costing, we can now determine with reasonable accuracy how much those special orders cost. The bad news is that they often cost twice or three times "estimated standard" because of all the sales activities which get lost in overhead. Maybe strategic reasons are OK but if the general manager knew the true cost, he or she might think twice about accepting the order or at least the price. Now imagine the implications of having 40% of all products falling into this category.

Robin Cooper identifies the major influences on product cost as size, volume and process diversity. He also points out that, if a business spends a lot of money on overhead and allocates it to product using traditional methods, it is highly likely the costing system is providing misleading, inaccurate

Northern Telecom's ABC pilot project Dissatisfied with existing overhead allocation methods after a shift from manual to automated manufacturing methods, Northern Telecom had been searching for a new approach to cost management information.

As a result, the company invited Professor Robin Cooper of Harvard University and me to tour three of its divisions which were interested in installing an activity-based costing system. We were looking for a natural management buy-in as well as the classic signs of suitability for an ABC system - lots of product, a mix of production methods, and overhead comprising a substantial proportion of product cost.

One division chose to proceed more rapidly than the rest. It was a medium-sized manufacturer of complete voice and communications systems consisting of more than 300 individually-priced product entities. The division had already instituted a formal manufacturing simplification program and was in the process of introducing an advanced automated flexible manufacturing line operating without direct labor.

A team from the division's finance department was trained in ABC concepts and set about implementing ABC single-handedly, starting in the late spring of 1988.

The process involved six phases and a dimensional shift:

* The team interviewed every divisional manager to obtain a detailed breakdown of all activities classified as overhead.

* Assisted by the accounting staff, it then analyzed all functional expenses to determine the cost of each activity.

* Driver identification was an iterative process, involving talking to functional management about the causes of activity and then seeking quantitative information about the basis of calculating consumption either by first-stage drivers or directly by products.

* Software was supplied by Strategic Cost Systems Inc. Extremely powerful, the software is a standalone database model which runs on an MS-DOS personal computer equipped with 336 processor and 40 meg. of hard drive. A key phase was associated with populating the model with data provided by activity analysis and costing, known as the software's resources module.

* Product information was manually loaded, although capability exists to download full bill of material and labor routing data. For simplification, the decision was made to load product data only at the summary level rather than fully exploded.

* Data validation analysis and reporting were the final phase of implementation, leading consequently to management presentations.

* The next dimension beyond analysis is interpretation and internalization. This is where insights provided by the analysis are examined for reliability and relevance, and the opportunity emerges to use the information to influence decisions.

A concern expressed by many about ABC is that a large proportion of overhead costs would not be traceable to product. Experience at the NT pilot was good, with only 8% of total overhead assigned as general support.

Among the specific findings: high-volume and relatively uncomplex products, representing 65% of the division's standard cost of goods sold, were overcosted under traditional methods by as much as 40% of overhead; low-volume products incur significantly more overhead than currently attributed; a few key, high-volume products were making substantially more standard gross profit than previously believed, while piece parts and an older product were making losses.

The Northern Telecom pilot project proved that the approach is both viable and capable of providing critically important new information, and that traditional costing systems can produce significant errors in overhead allocation. This example is not unique; similar results have been consistently reported by other companies. The question remains whether or not any use can be made of this information. The most notable problem with the model is that 44 drivers may be too big to be comprehendible, and that as the system is refined, there will be a need to simplify it. *

management information. That is, if you have many products, some high volume and the rest all over the quantity map, or you have a mix of automated and manual manufacturing processes, the chances are that your costing system is producing information which is misleading management.

What are the implications? Consider how many of the success stories we hear about Japanese companies start with their having attacked the market with low prices made possible by high volume but low variety production, and simplified, highly engineered, cost-optimized plants. Consider how the Japanese tactic has driven many Western manufacturers to become more specialized, going after the higher priced, low volume and customized markets. It is no mistake that the lower volume, specialized products have higher prices, while the high volume price is set by the cost leader in the market place; low volume products cost more to make.

Consider also the strategic and profit implications to those Western manufacturers who are being chased into the low volume market. The current thrust of Japanese manufacturers is flexible manufacturing, so that they can respond cost-effectively to customers' demands for variety, thereby placing more pressure on the already stressed domestic supplier.

Having more accurate product cost and profit information permits management to examine the business it is in, whether in fact it should be going after the low volume market, and if so how to make it profitable. Siemens Electric Motors Works discovered that it could make money in the custom motor business after implementing a simple ABC system, by being able to cost each opportunity accurately and then pricing accordingly. Siemens did not guess about pricing after ABC; it priced to make a profit or did not take the job - period.

For those of you in the system business, a product is defined as any item which individually has a price. If you're thinking, "This does not apply to me because I have only one or two products, system A or B," you may want to think again. Ask yourself how many different elements comprise a system, each with a separate price and a variety of quantities dependent on configuration specifications. When you are setting prices or involved in a competitive bid, you cost and price out a complete list. Imagine what your profit picture might look like with ABC.

If you have many products, some high volume and the rest all over the quantity map, or you have a mix of automated and manual manufacturing processes, chances are that your costing system is producing information which is misleading management.

The ABC model used by Northern Telecom create a three-dimensional cost model of the business. One dimension is product cost. The second is cost of activities, eg., how much is spent by accounts payable to process 20,000 invoices a year. The third is business process of driver costs, e.g., the overhead cost associated with planning, procuring and maintaining a raw materials portfolio of 4,000 components.

This information permits management to prioritize and focus its cost reduction efforts. Just as overhead variance analysis techniques are used by consulting companies to streamline staff functions, so is cost data use for manufacturing consulting. They all start out by looking for where the money is being spent, and using a pareto analysis to shoot at the big targets. Questions such as whether to undertake more automation, add another manual flow line or warehouse, or to implement just-in-time management methods can be sensibly addressed by the controller in terms of cost-benefit analysis. After all, isn't that our job?

Secret weapon

Activity-based costing permits the management accountant to act as an internal management consultant, using this new found analytical tool to probe the cost structure of the business in a way never before possible. The added benefit of using management accountants to provide relevant information to your own engineers and other staff is that the people employed by the business already know it inside out. What they need is decent information.

The information made available can be used to make strategic decisions and to integrate the manufacturing excellence thrust. Being as good as the competition in terms of manufacturing process is a competitive battle which many Western manufacturers have already lost. Catching up is clearly important, but there is no point in having world-class manufacturing if the products you are selling drive large overhead costs, causing the overall business to become inefficient.

ABC is an invention of the Western world which is still possesses. The Japanese have not caught on yet, but you can bet they surely will. Activity-based costing is the scalpel management accountants can provide to their companies to carve out strategy for the next competitive battle, that of economic survival. Once the strategy based on economic analysis has been formulated, then is it the right time to start talking about performance measurement. Otherwise we will not know what to measure.

I believe finance can add substantially more value for the business by producing correct information, developed with insight, that can be used proactively. Some argue in academic terms that our current method of overhead allocation is appropriate for inventory valuation, satisfying GAAP, and all the other rules. Yet what is all that worth if in the process we have misled management? How can one justify continued and deliberate distortion of critical information, particularly at such a crucial, competitive time. The existing rules have been developed for the specific purpose of providing a much-needed framework for accounting with, primarily, an external focus.

Management accountants also have a responsibility to look inwards. We have not had the capability until recently. But now we have it, let's use it - before our overseas competitors grab the opportunity.

Paul Sharman, ACMA, is a principal of Focused Management Information, a company established to help businesses implement new cost management techniques. He is a well-known conference speaker on strategic cost management and activity-based costing and conducts training in both through the Institute for International Research and other organizations.

CMA MAGAZINE