From the Australian Navy to Universities. 

Reducing implementations from multiple years to multiple weeks. 

The perception that Activity-Based Costing (ABC) is overly complex, prone to failure, and largely abandoned was once widely held and, unfortunately, not without basis. However, the landscape of ABC has significantly transformed over the past two decades, marking substantial advancements in how we approach and execute its implementation.

Initially, constructing ABC models was a daunting and time-consuming task. For instance, the project undertaken by the Royal Australian Navy in the 1990s spanned an extensive five years from inception to completion. Fast forward to the present day, and the timeline for developing such models has dramatically decreased, with projects being finalized in as brief a period as six weeks.

In this blog post, I will guide you through a collection of case studies that showcase the practical application of ABC across various sectors, beginning with the Royal Australian Navy and concluding with implementations within Higher Education. These examples serve to illustrate not only the evolution of ABC but also its relevance and adaptability to diverse operational contexts.

The development of the RAN model was uniquely challenging due to the decision to create custom costing software instead of opting for a commercial-off-the-shelf solution. During the mid-1990s, such a choice was not particularly uncommon, as the enterprise software required to support a large-scale model like this could cost several million dollars to purchase.

This project marked my introduction to Activity-Based Costing (ABC), which was characterized by an extremely hands-on approach. The process involved team members traveling across Australia, armed with pen and paper for interviews—a significant undertaking given the country’s vast size, with around five hours of flight time from the East to the West Coast.

At the time, this method was the norm, and the project benefited from the oversight of a well-respected international consulting firm. However, this approach also illuminated why many early ABC models struggled to succeed; the construction and subsequent maintenance and updating of the models were excessively time-consuming.

Following this experience, I joined a then-nascent consulting firm that would later be known as Pilbara Group. Initially, we employed a commercial software package (Oros by ABC Technologies) for our consulting work, but it quickly became apparent that it was inadequate for large, collaborative models due to its limitations in handling multiple users simultaneously. An alternative was Peoplesoft’s own ABC solution, priced at around $2 million for a license, which led us to develop our proprietary modeling engine. This engine has since matured into what is now known as Pilbara Insights.

Following our project with the Australian Navy, the US Navy expressed interest in adopting a similar model, starting with the US Pacific Fleet in Pearl Harbor, Hawaii, before transitioning to the Atlantic Fleet, starting in Jacksonville, FL, and extending to Norfolk, VA.

The scope of the Atlantic Fleet model encompassed all Base Operating Support (BOS) units stretching from Maine to Guantanamo Bay, Cuba. Base Operating Support includes services such as Port Operations, Air Operations, Security, Fire Fighting, and Engineering, among others. The goal of the model was to accurately assess the total costs associated with providing Base Operating Support, enhance the data available for making outsourcing decisions and managing contracts, and develop a suite of BOS metrics, such as the cost per ship movement or per aircraft take-off/landing.

Following a successful initial seven-month implementation at Navy Region Southeast (Jacksonville, FL), we were tasked with extending the model to the rest of the Atlantic Fleet bases within a tight timeframe of only another seven months. This ambitious goal was achieved by leveraging substantial amounts of data from primary source systems and employing distributed teams, which were coordinated by a central management team. This approach allowed for rapid updates to a singular, centralized model. The project was completed both on time and within budget.

Lessons Learnt: The key takeaways from this experience were the importance of developing standards and templates, the effectiveness of centralized project management, the value of utilizing as much existing source data as possible, and the need to minimize reliance on interviews and manual data entry wherever feasible.

One of the most pressing challenges faced by academic leaders, including department chairs and deans, revolves around how to efficiently deliver curriculum offerings with limited resources. Chairs, faculty members, and deans increasingly find themselves faced with questions such as “What is the best way to deploy resources (people, funding, classrooms etc.) to provide the desired curriculum?”, “Could an alternative allocation of resources achieve better results with the same investment of time and money?”, ”Where can existing resources be better optimized?”.  To better address questions such as these, the University of California, Riverside launched the Activity-Based Costing initiative.

The project at UC Riverside stood out because of our collaboration with both Grant Thornton and Deloitte, and partial funding from the Bill and Melinda Gates Foundation. Although the entire project spanned a year, the construction of the model was accomplished in 14 weeks.

By the time we approached this initiative, we had already developed several models for Higher Education in Australia, which allowed us to apply a refined methodology. We had introduced the concept of the Scoping Study to examine data from various systems, including Finance, HR, Payroll, Facilities Management, Timetabling, and Student Management Systems. This crucial early step ensured the data was of sufficient quality and completeness for model building. It also served as a pivotal decision point for our clients, allowing them to opt out early if data issues were too significant, based on our two-week review and feedback on how these issues might affect the model and its objectives.

To expedite the implementation process, we focused on maximizing the use of drivers and reducing manual allocations within the model. This approach allowed us to allocate resources efficiently within complex hierarchical structures, with the modeling engine automating the allocation process based on specified drivers.

Lessons Learnt:  Key insights from this experience include the importance of dedicating time upfront to review all necessary data before starting the model build. Having already developed several Higher Education models, we were able to precisely identify the types of data needed from various systems. The Scoping Study emerged as a critical component of a successful model implementation. Additionally, maximizing the automation capabilities within the model and minimizing manual interventions proved essential for efficient and effective model allocations.

In 2018, CQU had six weeks to comply with the new Department of Education mandatory reporting on the cost of teaching and scholarships. The short-term goal was to comply with the Department of Education submission requirement, but the university also wanted a long-term solution to empower finance and academic users to understand the university cost model. This was achieved by starting with a simpler methodology and evolving the complexity over time.

Because we were on a very tight timeframe of six weeks, we focused on the essential elements of the model, to provide the data required for mandatory reporting and a baseline model that could be easily expanded.  By this stage we had a standard methodology, data templates, and standard drivers to enable a rapid implementation of the model.

The model was implemented on time and on budget.

Lessons Learnt: Key insights from this experience underscore the benefits of starting with a simple, standardized approach and progressively evolving the model over time. We found that a complex starting model is hard to understand, and if university users don’t understand the model, they don’t trust the model. If they don’t trust the model, they won’t use the model. Starting with a standard is important, because it’s much better to build something quickly and seek feedback, rather than starting with a blank piece of paper and spend up to 12 months developing detailed model specifications. The important thing to note, is that this is only the starting point, the model can grow and evolve to address every single specific business requirement of the institution.

Next Steps

The next evolutionary step in enhancing ABC implementations involves the development of the Rapid Insights model. This is a more rigorous scoping study, incorporating newly developed business rule functionality, which not only utilizes drivers but also streamlines manual allocation processes. Initially, it will focus on integrating essential data from two primary source systems—the financial system and the student management system. The goal is to develop a comprehensive rapid model within a two to four-week timeframe.

Conclusion

Activity-Based Costing did develop a reputation for being complex, costly, and too hard to sustain. ABC was used extensively in the 1980s and 1990s when interest rates and inflation were high, and therefore cost management was a major priority. When the economy turned around and money became cheaper, the corporate mindset seemed to shift to “growth at any cost”. However, as we have seen in the past couple of years, the pendulum is swinging back again, to more of a cost management, right-sizing, efficiency and profitability mindset. The great news is that we have learnt numerous lessons over the last couple of decades and ABC is significantly cheaper and easier to implement and more importantly, to maintain. Many of our clients are still updating their models 10-15 years after initial implementation. It may be time for an ABC renaissance!