spacer spacer spacer Steve Colton's Accounting Insights spacer Management Accounting & Business Economics spacer

contact us   |   privacy policy   |   code of ethicsspacer


The Changing Role of the Controller:
A Model for Initiating and Guiding the Transition

by Stephen D. Colton

Executive Summary

Everyone seems to agree that the Controller’s role must change. Missing from the discussion is a conceptual model of the necessary changes for a Controller to proactively initiate. Using a process-based model of the accounting function, this article provides the missing guidance.


“You must be in tune with the times and prepared to break with tradition.” These words from William M. Agee, chairman of Bendix Corporation, capture the essence of changes occurring today within the management accounting profession — changes that enable the Controller to cast off unwanted stereotypes and finally contribute to the general management of an organization. The traditional Controller, the old Controller, was tolerated as a necessary evil, viewed as a bean counter and the corporate-cop. The New Controller is welcomed in the halls of management as a business partner — sought after for business acumen and the strategic perspectives she brings to the table.

How does one prepare to break with tradition? Many authors characterize these changes as a shift in the roles a controller is asked to play within the organization. Others focus on the new skills a controller must master to thrive in the changing environment. Both approaches provide valuable insights. Yet missing from all the literature is a conceptual model of the changes—a model that is necessary for a Controller to proactively initiate the changes and to shape changes as they occur. Using a process-based model of the controller function this paper discusses a practical approach by which a controller can manage the changes.

Process-Based Models

A process-based organizational model characterizes any organizational endeavor in process terms as follows

  • Suppliers provide inputs to an organizational process. Suppliers may reside within the organization or be external to it.
  • The process itself is a combination of people, technology, supplies, methods and environment that converts inputs into outputs.
  • Resources (such as people costs, technology costs and the costs of supplies) are consumed in the conversion process. The distinction between the consumption of resources and the conversion of inputs is important when managing a process.
  • Process outputs are either delivered to an internal customer becoming inputs to a downstream activity; or delivered to an external customer.

To paraphrase Stephen Covey, the key to process definition is to start with the end in mind. The key to building process-based organizational models is to first investigate and define process outputs.

Process outputs are quantifiable; they are defined in terms of measurable output units. This is a straightforward proposition in the production world. A process output is a widget of some sort and widgets are measured by counting them. It is more difficult in the service world. The output of a service process is frequently intangible and must be measured using a tangible surrogate. Consider, for example, investment advice. One could define investment advice as the output of an investment research process. But investment advice cannot be measured directly. It is intangible. A reasonable surrogate might be a completed research report or a documented buy/sell/hold recommendation. Why is this important? While some of the controller function processes are production-like, many more are of a service nature. Outputs of a service process are notoriously difficult to define.

A useful discipline when building process-based models is to characterize each organizational process using a verb/noun combination; the verb describes the process of conversion and the noun identifies the process output. Using this technique, the earlier example of a production process could be stated as “make widgets” and the service process as “provide investment advice.” An extension of this technique is to append the process inputs onto the verb/noun phrase using the word “from.” The examples become “make widgets from widget parts” and “provide investment advice from investment research.”

Provide Information from Data

The controller function varies from organization to organization and comprises diverse and occasionally contradictory roles and responsibilities. At a high level, however, there are two common activities:

  • The controller develops information regarding the financial impact of organizational actions, and
  • subsequently helps internal decision makers use that information to further the organization’s goals.

In process terms, the controller’s process is: “Develop and communicate financial information.” The inputs to this process are data of all sorts, primarily but not exclusively financial data. It follows that the controller’s process description becomes: “Develop and communicate financial information from data.”

Data without context is not information. Peter Drucker writes, “Information is data endowed with relevance and purpose.” The process-based definition of the controller function adds insight to this discussion by focusing attention on the conversion process. The controller converts financial data into financial information by endowing the financial data with relevance and purpose — by providing context. This is why financial data is composed entirely of numbers while financial information comprises numbers, words and phrases. The words and phrase are used to convey context.

Two Dated Paradigms about Financial Information

The New Controller must break two dated paradigms regarding the development and use of financial information, specifically cost information.

  • All users of cost and other financial information need the same kind of information.
  • Cost information is only derived from accounting transactions that are traced to a general ledger account.

Financial information is not homogeneous. Rather than a single kind of cost information, an implicit time perspective distinguishes three as summarized in Exhibit 1. One kind of cost information has a historical or results perspective. Its purpose is to record what has happened in the past, and it is used by people outside of the organization — bankers, stockholders, creditors, donors, bondholders, regulators, and taxing authorities. A second kind of cost information has a right-now time perspective. Internal workers making day-to-day, real-time decisions about the organization’s business processes need this kind of information. The third kind of cost information has a forward-looking, strategic perspective. Managers making pricing and other strategic decisions about the organization’s future need it.

The New Controller must also abandon the paradigm that cost information is derived exclusively from accounting transactions that are traced and tied to a general ledger account. There is no argument that one must “tic ’n tie” historical cost information for external users. However, that is not a characteristic of the cost information needed by internal users. The information needed by internal users, both real-time managers and strategic thinkers, must be actionable, timely, composed in their language, and indicative of the causes of the costs. It cannot contradict information from the general ledger, but it certainly does not have to tie to the general ledger.

Exhibit 1: The Three Kinds of Financial Information
Figure 1

The Unmistakable Conclusion

Making these changes — breaking these two paradigms — leads to one unmistakable conclusion. Since there are three different kinds of financial information, there also must be three distinct processes for developing and communicating that information. The process definition of the controller generated earlier is too general to guide the transition to the New Controller. It must expand to encompass a these three distinct financial information processes. Using the verb/noun combination technique, a more comprehensive description of the New Controller’s three financial information processes, is as follows:

  • Develop and communicate fiduciary financial information,
  • Develop and communicate operational financial information, and
  • Develop and communicate strategic financial information.

Traditionally, organizations relied solely on a fiduciary information process and only developed cost information with a historical perspective. This information was required by external parties and was given to internal information users almost as an after thought. The management objective was cost recovery, and historical information supported achievement of that objective. Today, the management objective has changed from cost recovery to cost management, and historical cost information does not support cost management. Historical cost information has rarely supported the needs of strategic decision-makers. The New Controller designs and implements new processes — separate and distinct from the fiduciary financial information process — to develop and communicate operational and strategic cost information. Nothing changes until the process changes.

A Model for Managing the Change

The direction is clear, yet practical obstacles remain. The old controller’s plate was already full. Where are the resources to develop and nurture new processes? How does one do more to support operational and strategic decision-makers without doing less in the fiduciary arena?

There is no question that in the short term, the workload increases as the controller function transitions to its new role. Any organizational change is like this. There is double-duty. New processes are implemented before old processes are abandoned. The changes are incremental. A new balance for the effort spent among fiduciary, operational and strategic information processes is not achieved instantaneously. The good news is that a new balance is eventually achieved. The challenge for the New Controller is to assure that the new balance is appropriate to satisfy the information needs of the organization.

Exhibit 2 shows a model characterizing the tasks and activities that generate financial information. The top axis lists three different controller function activities based on the reporting interval from Exhibit 1. Recurring processes are those tasks performed every month or every week or every day. They are normally concerned with transaction processing and routine reporting. Ad-hoc responses are those exercises that are “one-of-a-kind” or re-occur on only an occasional basis, such as when the Production Manager asks a specific question. A special project is a bundled set of ad-hoc responses that relate to one specific business issue. Special projects require the commitment of dedicated resources for a longer duration than a single ad-hoc response.

The left axis of the table depicts the three information processes of the New Controller. Each cell of this model represents a kind of information and a kind of task. Process tasks can fall anywhere within the two-dimensional model. For example, the bottom row shows that a recurring task, an ad-hoc response or a special project can generate fiduciary accounting information. The first column denotes that recurring tasks can be used to generate strategic information, operational information and fiduciary information.

Exhibit 2: Categorizing Financial Information Tasks
Figure 2

The Optimum Configuration

While process tasks can fall anywhere within the two-dimensional model shown in Exhibit 2, there is an optimum configuration. The large arrow in Exhibit 3 shows the configuration for which to strive. Optimally, the fiduciary information process comprises recurring tasks, the operational information process comprises tasks that generate ad-hoc responses and the strategic information process comprises tasks that support special projects. These combinations represent a direct correlation between process tasks and the characteristics of the resulting information.

Tasks falling within cells of the model that border the optimum configuration are labeled “Caution” in Exhibit 3. Such tasks may be legitimate but should be examined carefully; they represent a misalignment between process tasks and the resulting information. Unfortunately these kinds of tasks divert resources from the activities and tasks comprising the optimum configuration and prevent the resource-constrained New Controller from delivering the financial information an organization really needs. Consider the following examples:

1. Exhibit 3 shows that the optimum fiduciary information process consists primarily of recurring effort. It is quite production-like. Manage these fiduciary information tasks with the same ruthless drive for efficiency one would use on any assembly line. When fiduciary information is delivered through an ad-hoc information process (a cell labeled “Caution” in Exhibit 3), the opportunities for efficiency are foregone and resources are consumed unnecessarily.

2. The optimum operational information process consists of tasks designed to deliver ad-hoc responses. Sometimes an ad-hoc request for operational information is made more than once, often in consecutive periods. The mistake organizations make is to assume that this operational information will then always be required and that this information should be provided through a recurring process. Note that the cell in Exhibit 3 that represents a recurring task that provides operational information is labeled “Caution.” The issue is that the operational information may not be needed every period; in fact the information may never be needed again. That is the nature of operational information and the reason for a process of ad-hoc responses. The resources wasted by a recurring process that is not needed or used is obvious.

Exhibit 3: The Optimum Configuration
Figure 3

Tasks should seldom fall under the large X’s in Exhibit 3. These cells represent a major disconnect between the nature of the information provided and the frequency of process tasks. One common scenario is the temporary delivery of fiduciary information through a special project. This occurs when an organization implements a new enterprise accounting system and fiduciary information must be reported before the recurring tasks are implemented. The temporary process is abandoned after the new accounting system is installed.

Using the Model to Manage Change

As the New Controller designs and implements new financial information processes, the activities and tasks will fall into every cell of the model. The challenge for the New Controller is to make incremental changes in the activities and tasks of the new information processes until they attain the optimum configuration and mature. Exhibit 4 illustrates this concept.

Exhibit 4: Using the Model
Figure 4

Continually assess the configuration of financial information processes throughout the transition. Eliminate activities and tasks that fall under the large X’s. Adjust tasks and activities that fall into cells of the model labeled “Caution.” Nurture processes that fall within the optimum configuration as they stabilize and mature. Dedicate material time and effort to supporting internal decision-makers with the new processes. The arrows in Exhibit 4 show how tasks should be nudged into the optimum configuration.

Software Tools

Computer databases and other software tools play an important role in the implementation of the New Controller’s information processes. Software tools, however, play a supporting role only; the primary focus must always remain on the informational needs of the business, first, and the management process intended to generate the information, second. The fact that there are three different kinds of financial information dictates that the New Controller operates three different information processes. The same analysis dictates that the New Controller uses three different software tools to support his information processes.

Enterprise accounting data is only marginally valuable to the operational decision-maker because it does not relate cost to physical counts and time. The strategic decision-maker is not helped since enterprise accounting systems don’t forecast future performance. Enterprise systems only report financial data. They do not convert that data into information.

ABC software carries similar limitations. The mistake often made with ABC software is to try and use the data from a single ABC model to satisfy the informational needs of both operational and strategic decision-makers. It can’t be done. Operational and strategic decision-makers need information at different and incompatible levels of aggregation. The strategic decision-maker needs predictive information while the operational decision maker needs real-time, actual performance information. These two time perspectives of the requisite financial data are incompatible within a single model. The organizational effort needed to analyze, define and document internal processes is common to any activity-based initiative. To assure the resulting ABC models are of value, customize them for the kind of data they are to provide.


A process-based model of the accounting function provides guidance for initiating and guiding the transition of the controller function to its new status as a valued member of the executive management team. It characterizes the three kinds of financial information and demonstrates how the New Controller must design and implement new processes — separate and distinct from the fiduciary financial information process — to develop and communicate operational and strategic cost information. Finally, the process-based model relates the frequency of information generating tasks to the kind of information the processes provide. In all, the model guides the New Controller in designing new processes and managing the transformation of the controller function.

Top | PDF

spacer spacer
scai Footer