www.hbr.org
T
OOL K
IT
Time-Driven Activity-
Based Costing by Robert S. Kaplan and Steven R. Anderson
•
Many companie s abandoned activity-based costing because it did not capt ure the compl exity of their operations, took too long to implem ent, and was too expensive to build and maintai n. Here’s a way ar ou nd th ose problems.
R
eprint
R0411J
T
OOL
K
IT
Time-Driven Activity-
Based Costing by Robert S. Kaplan and Steven R. Anderson harvard business review
• november 2004 pa ge 1
C
OPYRIGHT © 2004 HAR
V
A
R
D BUSINESS SCHO
OL PUBLISHING C
ORP
OR
A
TION. ALL RIGHTS RESER
VED.
Many comp anies abandoned acti vit y-based costing because it did not capture th e complexity of their operations, took too long to implement, and wa s too expensive to build and maintain. Here’s a way around th ose problems.
In the classroom, activity-based costing looks like a great way to manage a company’s lim- ited resources. But many managers who have tried to implement ABC in their organizations on any significant scale have abandoned the at tempt in the face of rising costs and em- ployee irritation. They should try again, be- c ause the new approach we lay out in the fol- lowing pages sidesteps the difficulties traditionally associated with large-scale ABC implementation by relying on informed man- agerial estimates rather than on employee sur- ve ys. It also provides managers with a far more flexible cost model to capture the complexity of their operations.
ABC Made Difficult
The roots of the problem with ABC lie in the way people traditionally construct ABC mod- els. Assume you are analyzing a customer ser- vice department that performs three activi- ties: processing orders, handling inquiries, and performing credit checks. The department’s total expenses (the cost of the personnel, man- agement, IT, telecommunications, and other fixed resources) amount to $560,000. The ac- tual (or estimated) quarterly quantities of work in the three activities are 49,000 orders,
1,400 inquiries, and 2,500 credit checks.
To
build a traditional ABC model for this de- partment, you would survey employees to esti- mate the percentage of time they spend (or ex- pect to spend) on the three activities and then assign the department’s resource expenses ac- cor ding to the average percentages you get from the survey. Let’s say employees report that they spend (or expect to spend) about 70% of their time on customer orders, 10% on in- quiries or complaints, and 20% on credit checks. This implies, under ABC, that each order consumes $8 of resource expense, each inquiry $40, and each credit check $44.80, as shown in the exhibit “Doing ABC the Tradi- tional Way.” Armed with these figures, known as the cost-driver rates, managers can assign the costs of the department’s resources to the customers and products that use its services.
This approach works well in the limited set-
Time-Driven Activity-Based Costing
•
•
•
T
OOL
K
IT
harvard business review
• november 2004 page 2
Robert S. Kaplan ( rkaplan@hbs.edu ) is the Marvin Bower Professor of Leader- ship Development at Harvard Business
School in Boston.
Steven R. Ander- son ( sanderson@acornsys.com ) is the f ounder and chairman of Acorn Sys- t ems, a software and consulting firm headquartered in Houston. Kaplan serves on the board of Acorn Systems. ting in which it was initially applied, typically a single department, plant, or location. Difficul- ties arise, however, when you try to roll this ap- proach out on a large scale for use on an ongo- ing basis. In one large bank’s brokerage operation, the ABC data-gathering process re- quired 70,000 employees at more than 100 fa- cilities to submit monthly reports of their time allocation. The company employed 14 people full-time just to manage the data collection, processing, and reporting.
The time and cost demands of creating and maintaining an ABC model on this scale is a major barrier to widespread adoption at most c ompanies. Since the systems that are put in place are updated infrequently (because of the c osts of reinterviewing and resurveying), the model’s estimates of process, product, and cus- tomer costs soon become inaccurate. What’s more, people waste their time arguing about the accuracy of cost-driver rates that are de- rived from individuals’ subjective beliefs rather than addressing the deficiencies the model re- v eals: inefficient processes, unprofitable prod- ucts and customers, and excess capacity.
Tr
aditional ABC models also often fail to c apture the complexity of actual operations.
C
onsider the activity “ship order to customer.”
Ra
ther than assume a constant cost per order shipped, a company may wish to recognize the c ost differences when an order is shipped in a full truck, in a less-than-truckload (LTL) ship- ment, using overnight express, or by a com- mercial carrier. In addition, the order may be entered into the system either manually or electronically, and it may be either a standard or an expedited transaction. To allow for the significant variation in resources required by the different shipping arrangements, new ac- tivities must be added to the model, thereby e xpanding its complexity.
As
the activity dictionary expands—either to reflect more detail about activities or to ex- pand the scope of the model to the entire en- terprise—the demands on the computer pro- grams used to store and process the data escalate. Suppose a company has 150 activities in its enterprise ABC model, applies the costs in these activities to some 600,000 cost objects
(products and customers), and runs the model monthly for two years. That would require data estimates, calculations, and storage for more than 2 billion items.
Such expansion has caused ABC systems to exc eed the capacity of generic spreadsheet tools, such as Microsoft Excel, and even many
ABC software packages. The systems could ta ke days to process one month’s worth of data. For example, the automated ABC model f or Hendee Enterprises, a $12 million fabricator of awnings, took three days to calculate costs f or its 40 departments, 150 activities, 10,000 or- ders, and 45,000 line items.
These problems have become obvious to most ABC implementers. But a subtle and more serious problem arises from the inter- view and survey process itself. When people es- timate how much time they spend on a list of activities handed to them, they invariably re- port percentages that add up to 100. Few indi- viduals report that a significant percentage of their time is idle or unused. Therefore, cost- driver rates are calculated assuming that re- sources are working at full capacity. But as we all know, operations often run at considerably less than their capacity. That means that the es- timated cost-driver rates are usually much too high. (Technically, they will be overstated by the reciprocal of the capacity utilization per- c entage: At 80% utilization, the rates are 25% too high; at 67% utilization, the rates are 50% too high.)
The New ABC
The solution to the problems with ABC is not to abandon the concept. ABC after all has helped many companies identify important cost- and profit-enhancement opportunities through the r epricing of unprofitable customer relation- ships, process improvements on the shop floor, lower-cost product designs, and rationalized product variety. Its potential on a larger scale r epresents a huge opportunity for companies.
For
tunately, simplification is now possible through an approach that we call time-driven
ABC, which we have successfully helped more than 100 client companies implement, includ- ing those described in this article. In the revised approach, managers directly estimate the re- source demands imposed by each transaction, product, or customer rather than assign re- source costs first to activities and then to prod- ucts or customers. For each group of resources, estimates of only two parameters are required: the cost per time unit of supplying resource ca- pacity and the unit times of consumption of re- source capacity by products, services, and cus- tomers. At the same time, the new approach
Time-Driven Activity-Based Costing
•
•
•
T
OOL
K
IT
harvard business review
• november 2004 page 3 provides more accurate cost-driver rates by al- lowing unit times to be estimated even for com- plex, specialized transactions.
Estimating the cost per time unit of capac- ity. Instead of surveying employees on how they spend their time, managers first directly estimate the practical capacity of the resources supplied as a percentage of the theoretical ca- pacity. There are various ways to do this. As a rule of thumb, you could simply assume that practical full capacity is 80% to 85% of theoret- ical full capacity. So if an employee or ma- chine is available to work 40 hours per week, its practical full capacity is 32 to 35 hours per week. Typically, managers would allot a lower ra te—say 80%—to people, allowing 20% of their time for breaks, arrival and departure, c ommunication, and training. For machines, managers might allot a 15% differential be- t ween theoretical and practical capacity to allow for downtime due to maintenance, re- pair, and scheduling fluctuations. A more sys- tematic approach, perhaps, is to review past activity levels and identify the month with the largest number of orders handled without ex- c essive delays, poor quality, overtime, or stressed employees. Whichever approach you prefer, it’s important not to be overly sensitive to small errors. The objective is to be approxi- mately right, say within 5% to 10% of the ac- tual number, rather than precise. If the esti- mate of practical capacity is grossly in error, the process of running the time-driven ABC s ystem will reveal the error over time.
R
eturning to our example, let’s assume that the customer service department employs 28 r eps to do the frontline work and that each puts in eight hours per day. In theory, there- fo re
,
each worker supplies about 10,560 min- utes per month or 31,680 minutes per quarter.
The practical capacity at about 80% of theoret- ical is therefore about 25,000 minutes per quarter per employee, or 700,000 minutes in total. Since we already know the cost of sup- plying capacity—the $560,000 in overhead c osts—we can now calculate the cost per minute of supplying capacity ($0.80).
The capacity of most resources is measured in terms of time availability, but the new ABC approach can also recognize resources whose c apacity is measured in other units. For exam- ple, the capacity of a warehouse or vehicle would be measured by space provided, while memory storage would be measured by mega- bytes supplied. In these situations, the man- ager would calculate the resource cost per unit based on the appropriate capacity mea- sure, such as cost per cubic meter or cost per megabyte. Estimating the unit times of activities.
Having calculated the cost per time unit of supplying resources to the business’s activities, managers next determine the time it takes to c arry out one unit of each kind of activity.
These numbers can be obtained through inter- views with employees or by direct observa- tion. There is no need to conduct surveys, al- though in large organizations, surveying employees may help. It is important to stress, though, that the question is not about the per- c entage of time an employee spends doing an activity (say, processing orders) but how long it takes to complete one unit of that activity
(the time required to process one order). Once again, precision is not critical; rough accu