Double measure: when Scottish Courage Brewing tried to graft a new system for activity-based budgeting on to its existing one for activity-based costing, it ran into problems and was forced to suspend the project. But the company did gain invaluable insights in the process. (Finance Activity-Based Budgeting).
Scottish Courage Brewing, which is widely regarded as a pioneer of ABB in the UK, has now shelved the further development of its system, but its experience provides insights that address a number of practical issues that have not been reported.
The starting point of Scottish Courage Brewing's ABB project was the implementation of ABC, which was a response to pressure from its customers (mainly supermarkets) for price reductions. When ABC was piloted in the early 1990s it was successful in that it provided a whole new set of data that raised managers' confidence in the accuracy of beer product costs. This breakthrough encouraged senior managers to approve a full-scale implementation, which aimed to build a system that could support both ABC and ABB processes.
A steering committee was formed to set strategies for the ABB implementation and monitor the progress of the project. It comprised the production director, the finance director and an ABB co-ordinator. An implementation team was also formed. This consisted of two full-time members (the ABB co-ordinator and a management accountant) and part-time members representing each of the company's six breweries. In the early stages, the support of senior management was ensured by the presence of the directors on the steering committee. During the implementation, workshops and presentations were held for line managers at each brewery with the aim of gathering information and demonstrating the finished models.
The conceptual model was based on the notion of ABB as a reverse process of ABC, as codified by Robert Kaplan and Robin Cooper in Cost and Effect (Harvard Business School Press, 1998). Sales forecasts were used to determine the product mix by taking account of factors such as brands, packaging and markets. Activity and resource requirements were then generated, taking capacity constraints into account. Lastly, the cost influences to resources were determined and fed back to general ledger to produce a budget.
If this model could be implemented successfully, it would be able to conduct "what if?" scenarios to give sound information on which to base strategic decisions.
The enhancements to the ABC software included the ability to explore various volume/ mix scenarios, to allow the process to work from sales forecasts towards cost budgets and to determine revised total resource costs. Processes were mapped out at each brewery. A key requirement here was to identify activities and resources that varied with volumes and to determine the linear relationships between them. Another requirement was to identify the maximum capacity of a given resource so that a warning signal would be given in the "what if?" scenarios when this capacity was exceeded. The violation of this constraint would invalidate total cost predictions.
It was intended that the ABB model would calculate the total product cost of a particular type of beer, as well as the total production costs of individual breweries, based on the computed variable unit costs, volumes and related fixed costs. Final calculations were performed automatically in a customised computer software model called Cost Control.
But the new system implementation did not proceed as intended, significant problems occurred and errors were found in the model. While the system was able to produce the necessary ABC data, it couldn't perform the ABB calculation process reliably, which cast doubt on the quality of the resulting budget data. The implementation problems, together with the perceived unreliability of budgets, eventually led the company to shelve the ABB project.
The implementation problems included:
* Differences in model specifications between ABB and ABC systems. The ABB model was designed on the basis of ABC in reverse. This was done using a computer software model that was designed to perform both processes. While this principle was theoretically sound and straightforward, its application was unexpectedly complex. Owing to software constraints, two models--spreadsheets to handle the variable cost data and the customised Cost Control model to deal with the fixed cost data--were required to handle the combined processes via a database interface transfer system. While it was able to execute the data aggregation and transfer functions satisfactorily during an ABC process, the system was unable to perform the reverse calculation exercises for ABB. The data dissemination during the ABB calculation process resulted in a huge number of possible permutations, which decelerated the computations to an unacceptably slow speed.
* "Linear" versus "stepped" cost behaviour. The main assumption made by the implementation team and built into the ABB model was the existence of linear relationships between volume and products, activities and resources. This was actually invalid, because some of the costs--eg, overtime--had stepped relationships. During ABC calculations, costs of a stepped nature were accounted for manually before they were put into Cost Control. The ABB model assumed that, in a scenario where the volume of a product was reduced, overtime costs were also reduced linearly. When this linearly reduced overtime cost was applied to Cost Control, the resulting budget was lower than expected. The stepped nature of the overtime cost required that volumes should be reduced below a certain level before an overtime shift could be removed. In this case, a change in volume did not lead to a linear change of overtime cost in reality. The difference between these relationships is shown in figure 2 on the next page.
* Complexity in systems and data modelling. Scottish Courage Brewing's ABC and ABB systems consisted of 250 activities, 100 products and 500 types of resources. Because of the complexity of the ABC system, the possible permutations for a large number of variables--necessary in the ABB process--made the computations for the ABB process formidable. Ultimately, there were 12,500,000 possible permutations covering all scenarios. In addition, given the large amount of data that had to be processed and transferred between the spreadsheets and Cost Control, the time required to perform these operations turned out to be significantly longer than expected. Another facet of this complexity related to the variety of systems used throughout the company. Some of these systems were based on ideas that conflicted with the principles of the ABC methodology. For example, the financial accounting system was configured using cost codes for reporting financial figures in a traditional form. Data from this system could not be directly downloaded to Cost Control. The book value of a given piece of equipment in the nominal ledger depreciates gradually. Over time, the book value of that equipment will eventually fall to zero. Since the book value of a piece of equipment is taken into consideration for product costing in ABC and ABB, products that use equipment of zero book value in their production processes will naturally have lower costs compared with those that use newer equipment. The account codes at the nominal ledger, which were created to fulfil the traditional accounting purpose, are not as detailed as those required by ABC and ABB. To perform an ABB process, data from nominal ledger accounts needs to be disseminated and converted to an ABC format and then divided into variable and fixed cost data before it enters the Cost Control system. This is a labour-intensive and time-consuming process.
* A lack of standardisation. The ABC and ABB projects involved only Scottish Courage Brewing, which is the production unit of the Scottish & Newcastle Group. But sales forecasts were modelled by a separate beer-selling unit. Since no standard guidelines existed for sales forecasts, managers at breweries had their own ways of obtaining them. This created a perceived credibility and budget reliability problem, because each brewery tried to obtain favourable sales forecast data to maintain an optimal production level and keep its plant running. In addition, each brewery had its own classification of fixed and variable costs, and the format (product volume and mix) of input data varied from one brewery to another. These various sources of non-standardised data, when used in an unified ABB model based on one pilot site, created problems in data conversion and validation. As a result, budgets produced by the ABB system were seen by managers as unreliable and unsuitable for planning, making budget-related decisions and measuring performance.
* The effects of structural change. The ABC/ABB project coincided with a major acquisition that saw the company expand from three to six breweries. This brought inevitable organisational changes and meant a possible change of priority that could affect the level of support the senior managers could offer the project. Constant restructuring was taking place in the Scottish& Newcastle Group in response to the competitive market situation. Managers required systems that could let them perform "what if?" analyses and produce information promptly and accurately. This was not possible in the ABB system given the technical difficulties. There were also other mechanisms used to control costs. For example, the quarterly financial review was an update of departmental annual budget performance. This was based on benchmarking variances between monthly forecasts and actual sales and expenses. The annual budget became less important in terms of planning because the monthly forecasts were seen to be more realistic.
* Behavioural influences. The ABB system was developed and implemented alongside the ABC system, so human inertia may have been "inherited" from the introduction of ABC. For example, its use of the term "non-value-added" to define activities in some divisions--maintenance, for example--caused resentment among the line managers concerned. They were also disaffected by their lack of involvement in the model's development. Owing to time constraints, the implementation team was unable to consult all managers at individual plants. Training was provided for accountants and managers at each brewery, but, because of the complexity of the models and their lack of familiarity with activity-based principles the managers had to rely on the accountants to build budgets. Compared with conventional budgeting, where managers could participate with relative ease, they found it hard to get involved in the budget preparation and negotiation process under ABB.
From Scottish Courage Brewing's experience it is evident that ABC and ABB systems are linked--in that an ABB implementation is influenced by some of the issues that affect an ABC implementation--but they are distinctive processes serving two different purposes. An ABC system is for product costing and an ABB system is for management planning. A computer system designed to support an ABC process may not be well suited for ABB.
There are potential benefits to be gained by having separate systems for ABC and ABB. The ABB system may require detailed information for a broader range of activities to allow the suitable allocation of resources based on the predicted production volume, while an ABC system may aggregate activities for product costing. Functions for management planning and control purposes can be conducted in the ABB system without interrupting any operations within the ABC process, thus increasing the system's efficiency.
ABB is unlikely to require the same level of detail as ABC. Scottish Courage Brewing built a detailed ABC model that identified more than 200 activities and cost drivers. This amount of detail increased the level of complexity in the ABB process in a way that significantly affected its outputs. The problem of complexity will eventually outweigh the benefits offered, so a trade-off between efficiency, cost-effectiveness and accuracy is required.
Sufficient training must be provided to ensure that all users are familiar with the operation and are at ease with the ABB system. Constant technical support must also be provided to ensure the smooth operation of the system. An adequate "parallel run" of new and existing systems should be considered.
Organisations in volatile and competitive markets may experience greater demand from their managers for tools that can provide more appropriate and timely information. Any ABB system must meet these requirements if it is to succeed. Although ABB has been discontinued at Scottish Courage Brewing, the company learnt some valuable lessons from the experience. It did use the product costs obtained through the ABC and ABB projects as the standard cost at that time. Since then, updates of product costs have been used as an alternative approach. Lastly, the people involved in these projects have developed invaluable skills and knowledge, which the organisation would do well to retain.
Lana Liu is lecturer in accounting and finance at the University of Newcastle (firstname.lastname@example.org), John Martin is finance manager at Scottish Courage Brewing and John Robinson is principal lecturer at the University of Northumbria. The authors thank Professor Falconer Mitchell for his contribution to the research