Has ABC become a bad word?

Following the strong initial uptake, rapid growth through the 1990’s, ABC peaked by 2000 and has been in decline since. There is consensus that prevailing costing information has shortcomings. Does this pave the way for ABC to be resurrected or will it decline further?

Below I share my thoughts on what has made ABC a bad word and how its relevance might be regained.

1) Not everyone relates to ABC easily

The conventional cost flow architecture of ABC equates costs to resources. The distinction, that cost refers to the “spend” and resources to what an organisation “gets” for the level of spend is not explicit. Resource consumption accounting (RCA) attempts to redress this, but its practice is not widespread.

Expense (cost) items are single dimensional, have a single metric, its amount of cost. Resources have two, quantity and cost.

I have found that differentiating cost incurred from resources in models 1) enables measurement of resource consumption, 2) improves the application of causal factors based on already measured metrics, 3) shifts management focus from cost pools to resource pools and 4) resonates better with non-financial managers.

In practice it is not necessary to quantify capacities for every resource pool. Careful consideration of the level of granularity resource pools are specified will help minimise cost distortions that arise from over averaging.

2) One size does not fit all

Thirty years after its conception, ABC is still veiled in misconceptions, even mystique, which discourages its adoption. I encounter managers who bemoan the lack of clarity of ABC models that depict cost flows in two steps, resources to activities to cost objects.

Cost flows are rarely linear: resources support other resources, not all activities performed are in support of other activities or cost objects; activities are also performed on behalf of resources. If this is implied in the the two-step ABC cost flow architecture, it isn't clear.

Driven either by convention or the limitations of existing software, frequently models are straight jacketed, ie., three layers: resources, activities and costs with no backward loop, making them incomprehensive to all but a few. I suspect that actual multi-tiered models exist, as illustrated to some degree in slide 4 of “Why Use ABC?” by Gary Cokins. Yet, why is that such designs seem not to have found their way into the public domain to offer guidance for would be adopters to apply these in specific industry sectors, without having to reinvent the wheel?

3) D-I-Y has not been a viable option

The implementation of ABC in the early years was monopolised by big accounting firms. The high cost of hiring external consultants to implement ABC has no doubt driven organisations to do it internally or not adopt ABC at all. To their credit, what the big firms charge in high fees is compensated by the best practice methods and collective experience they bring to the table.

The drawback of the D-I-Y approach is that often internal teams lack the requisite experience and expertise in ABC, which can lead to a failure to adhere to or even to violate the basic principles of ABC. What then passes off as ABC is not ABC at all. One of the first casualties has been causality! It comes as no surprise that when models provide little or no improved insight into profitability or performance, managers become disappointed and become deterred from persisting with ABC.

Implementing ABC is rarely easy, even when it is preceded by a prototype. The perceptions that ABC takes long to implement, costly to implement and maintain, is not as beneficial as claimed are not entirely unfounded and not easily dispelled.

I believe that the solution is to develop management accounting methods that are simpler, quicker, and less costly to implement and maintain than existing ones. This means giving priority in equal measure to revamping both existing methodologies and methods with the aim of empowering organisations to “plug and play” these with minimal risk.

4) ABC is also looking in the back mirror

One of the virtues claimed of ABC is that is it forward looking. By implication, that financial statements are merely reporting past events and of no value to making operational decisions.

There is little dispute that using historical data is the only means by which a comprehensive cost model for a business can be constructed. However, having done so it must be capable of being rapidly updated for future and immediate past periods, so it generates analytical constructs explaining just past and future outcomes in a timely manner.

It is not uncommon for CFO to lament that managers at both divisional and corporate levels are gravitating more and more towards self-service analytics produced by data analysts and away from management accounting reports. Presented in slick graphical format these are compelling and appealing. The lack of substance that is claimed of such analytics is compensated by rapid turnover, consistency, granularity, and their low cost. What gets said repeatedly has a propensity to be more readily believed.

Even if conceptually, ABC seems simple, modelling any organisation of size and scope of operations is no mean undertaking. Updating cost models is a time-consuming process. As a result, very few organisations have forward looking ABC models that are used for budgeting or forecasting. It would be interesting to know whether anyone has experienced differently.

For these reasons, ABC suffers from the same shortcomings that are levied against financial accounting information. The takeaway is the same: a dire need for easier, faster methods to design, build and maintain profitability models.

My own experimentations with trying alternative approaches to designing profitability models have been encouraging. I am certain that other members of PACE have also found innovative approaches to making new advanced management accounting methods more effective and easier to implement. How do we distil these into best practice and propagate them to a wider audience?

5) ABC is hiding from big data

The advent of “big data” has made large data sets commonplace, as are the tool sets to process them. Meanwhile, ABC practitioners are advocating for less detail in models as a means to making models more manageable.

In the context of ABC, big data refers to the large amounts of transactional, invoice and event level records captured in operational systems and other devices. Nowadays, such data is easy to access, but hardly anyone talks of costing and big data in the same breath.

What large data does for costing is that it eliminates the need for allocating costs sequentially, common in conventional modelling, from resources to activities to products and channels and then to customers. Instead, when cost objects are defined multi-dimensionally, for example “customer-product-transaction-channel-location,” resource usage by each of these elements can be simultaneously associated based on their independent causal factors. The improvements in costing accuracy and update speed are tangible. The ability to generate profitability analytics at any level of granularity is ground breaking.

I welcome comments, positive and negative as well as your own perspectives on the above. I can be reached on this forum or by email: sarath.amukotuwa@gmail.com