The Process of Developing an ROI Model


Michael D. Atherton


For an example ROI model, please contact me at 703-486-8497 or send an email to Mike Atherton.


This paper was first presented to the Digital Equipment Corporation's User Group (DECUS).



Many managers, when faced with the task of preparing a Return On Investment (ROI) analysis, focus on the technical aspects of the exercise. They dust off old text books and review concepts like Internal Rate of Return, Payback Period, Net Present Value and Cash Flow analysis. These accounting principles are certainly critical to preparing an ROI analysis, but there is another step of equal importance that is often neglected. This step is the process of determining what a project's functional criteria are as they relate to overall corporate strategy, objectives of related functions (i.e. Manufacturing, Marketing, Distribution, Finance, Information Systems).

This paper will discuss both the process of developing an ROI model with a top down approach based on strategic objectives and technical aspects of ROI modeling. A sample model, based on criteria often considered when comparing physical distribution control systems, is included and referred to throughout the paper.

The specific criteria used in various models will be different depending on the functional area where investment options are being analyzed. The functional objectives of investing in robotic welders, opening a plant somewhere in the southwest, supplying field representatives with notebook computers for remote order entry, or implementing a distribution control system are without a doubt different. However, the process of developing project objectives and then incorporating them into a model that utilizes sound ROI techniques is the same.

ROI Modeling - First Things First

Why do we go to the trouble of doing ROI analysis? It seems they are often conducted by one level of management to persuade another level of management to spend money. For this reason ROI analysis is often fundamentally sound - the methods and formulas are correct - but the underlying elements are flawed. The most important component of ROI analysis is incorporating the objectives for the organization initially into the ROI model and ultimately into the project's management. This is done not through the model itself, but through the process by which it is developed. By incorporating organizational objectives, the model becomes an accurate reflection of the expected functional outcomes of the project and therefore the expected financial outcomes.

It is important to understand the limitations of ROI analysis. Text books provide elaborate descriptions of the various financial ROI techniques - they will be summarized later. But what does a model really tell you? In the end ROI analysis can help you determine:

  • whether to make an investment
  • which investment option to take
  • expected financial outcomes

These outcomes are only valid given four conditions:

  • the expected functional benefits are attainable
  • the model itself is technically sound
  • the investment costs are accurate
  • the options analyzed are compatible with corporate and functional trategic plans

Managers often use ROI analysis like a divining rod. They use it as a nice black and white decision tool to decide where to dig. Unfortunately they often find water when they really needed oil. You have to know what it is you're looking for, what it will do for you, and how it relates to the rest of your organization before you can assess which of several options will best enable you to get it.

A model can be technically sound. Its assumptions and functional benefits can be perfectly accurate. The cost, savings, and revenue estimates can be correct. But, if the motivations for making the investment are not thoroughly understood and if the means for achieving the functional benefits are not compatible with the organization's strategic direction, the model is effectively worthless.

Why Are We Doing This?

Strategic issues from the top down must be considered in defining the motivation for an investment. The starting point for any major project should be to define the objectives that are to be achieved by the project at the corporate, divisional, and functional levels. It is amazing how often projects are undertaken for truly arbitrary reasons - like saving money.

ROI Diagram
Figure 1 - This chart depicts the top down process of determining ROI model functional criteria from the corporate strategy through functional objectives and finally project objectives.

When the project's strategic objectives have not been clearly defined as part of the project team's task, the project team must do this ground work themselves, before developing their ROI model. The effort will not be in vane as it contributes significantly to identifying objectives up front. Early identification of objectives is truly valuable in that it serves to focus the project team early on and provides direction for the development and implementation stages of the project life cycle.

It is important to keep in mind that an ROI model does not answer the question why. If you do not know why you are initiating an effort in the first place, ROI analysis will be of no value.

ROI Functional Criteria

Functional benefits are the means by which a project will achieve its intended objectives. Figure 1 shows how a project's ROI functional criteria are determined from the top (corporate strategy) down (functional and related functional objectives to project objectives). For each functional criteria, one or more quantifiable measures are determined and incorporated in the ROI model. Some of these measures, especially those that directly reduce costs, are easily quantifiable. Others, like the nebulous "improved customer service" or "improved quality" are more difficult to quantify but are often the most strategic in nature. A difficult criteria like "improved customer service" may be quantified in terms of measures like improved sales through better customer loyalty, greater market share, and improved customer retention.

The ROI model provided in Appendix A is for a business application system that addresses physical distribution. Each item in the Functional Data section requires an entry for an area related to operations that may be improved by implementing a distribution control system. Each of the items listed in this section addresses an opportunity for improvement related to physical distribution.

Related Functional Objectives

Investments are not made in a vacuum. Managers often focus only on the relationship of a project's functional benefits as they relate to the customer and costs. They concentrate on what they are trying to achieve without giving enough consideration to how the objectives will be achieved and how they relate to the parts of the organization that may be required to support it. By examining the status and direction of related functions, a manager can avoid even considering an investment alternative that is not compatible with the expertise or direction of a support or related function.

Consider a simple example of the lead administrator who was tasked to select a new suite of office management software on which the company's administrative function would standardize. The administrative function was using a word processor based on the Macintosh PC so the administrator narrowed the choices to Macintosh based offerings and ultimately selected the one with the functional benefits that would enable the staff to be most effective given their type of work. Imagine the shock of the lead administrator when on the day the new software was delivered, the MIS department announced that after a careful review that had lasted several months they had decided to standardize the company on the Intel based PC. A multi-million dollar contract with a PC manufacturer had been consummated and each person in the company would receive their own PC within a year.

Parts of an ROI Model

Once we have an understanding of what we are trying to accomplish through the project and how it fits into overall corporate and functional strategies, we can begin to assemble the ROI model. The most convenient tool with which to build the model is the computer spreadsheet. It enables the development of a standard model that can then be adapted for specific projects and project options. The spreadsheet also enables quick what-if analysis in which one or more variables can be changed and their impact on the model seen immediately.

The model provided in appendix A (please call me at (703-486-8497 to get a copy of this model. Currently it is not available on the World Wide Web) is divided into the following sections:

Required Data

In this section general information about the business that will be used throughout the model in various calculations will be entered. Examples include annual sales, transportation costs, wage rates for specific functions, cost of capital, estimated inflation rates.

Functional Data

In this section information pertaining to a specific functional benefit is entered. The information is entered based on either the current situation (example: Lost Sales Due to Poor Order Fill) or a planned improvement (example: Inventory Reductions Resulting from Improved Inventory Accuracy).

Summary of Annual Cost Reductions

The Required Data along with the Functional Data is used in formulas in this section to determine actual annual savings. Each line of this section corresponds to the same letter in the Functional Data Section.

Analysis of Cash Flows and Returns on Investment

This section takes the system costs, income (savings), depreciation, taxes and determines a net and cumulative cash flow. The net cash flow line is used to determine the internal rate of return and the cumulative cash flow line is used to determine the payback period.

The Functional Data section provides the biggest challenge in the formulation of any model. This is the heart of the model because it requires the modeler to specify the objectives of the project in a quantifiable manner. If you do not have a handle on why you are considering making an investment and how it will fit into to your overall strategic mix, you will find it exceedingly difficult to develop a meaningful Functional Data section.

ROI Techniques

There are several common approaches to the financial aspects of Return on Investment analysis. Several of the the more common techniques are described below.

Internal Rate of Return (IRR)

The internal rate of return is the interest rate that equates the present value of an income stream (the Net Cash Flow line for years 1 through 5 in the example model) with the cost of the investment (the Net Cash Flow line for year 0 in the example model). There is no specific formula that can be used to calculate the IRR. It must be found by interpolation. However, most spreadsheets provide an IRR function and will do it for you.

The advantage to IRR analysis is that it enables the comparison of rates of return on alternative investment options. Given two investment alternatives and assuming that both fit the strategic objectives of the organization, the investment with the higher internal rate of return should be selected. Conceptually it is the easiest method to understand.

Net Present Value (NPV)

The net present value returns a nominal amount. It is the amount in today's dollars (present value) by which the projected income (the Net Cash Flow line for years 1 through 5 in the example model) of an investment exceeds its cost (the Net Cash Flow line for year 0 in the example model). The calculation is based on the cost of capital which in the model is entered in the Required Data section.

Given two investment alternatives and assuming that both fit the strategic objectives of the organization, the investment with the higher net present value should be selected. Again, most spreadsheets provide a Net Present Value (NPV) function to make this calculation.

Payback Period

The payback period is the amount of time it takes for the cumulative cash flows to equal the initial investment. The example model provides an approximation of the payback period on the Cumulative Cash Flow line. The investment will have paid for itself in the year where the cumulative cash flow is positive. Within the payback year a larger number indicates the the payback occurs towards the beginning of the year.

Other Considerations

The savings (income) produced by an investment is only a part of the actual economic impact. The actual cash flow must also be adjusted for depreciation, taxes, and business growth. These topics are discussed in the following paragraphs.


Because we pay taxes on our earnings, the savings derived from an investment is equivalent to income on which we have to pay taxes. It is important to reduce the expected cash flows with the correct real tax rate for that firm. If a firm has been consistently profitable, it is likely it will have a standard corporate tax rate. This rate is typically between 30 and 45 percent and is entered in the in Required Data Section of the example model. If, however, the firm has recently had a large loss or write downs due to an extraordinary item, the firm may have a tax loss carry forward. In this case its actual tax rate may be significantly lower than the norm.


Depreciation is a non-cash expense item. Therefore, cash flows in the model need to reflect this. If our investment includes acquisition of depreciable items such as machines, computer equipment, software, patents, trademarks, plant expansion, or other similar items, we must configure our model to account for them. The initial capital expense of the items will represent the cash outflow. However, we do not write checks for depreciation even though we account for it as an expense on the income statement. For this reason we should add back into the cash flow stream the amount of depreciation for depreciable items acquire as part of the investment. The example model uses the accelerated method called sum of the years digits (SYD).

Growth Rates and Inflation

It is important to incorporate expected growth rates into an ROI model. As an example, if you expect to save on labor costs, and you expect labor costs to grow as a function of inflation over the period of cash flow analysis covered by your model, the savings attributed to the investment should be adjusted accordingly.

Likewise, if you expect to save a specific amount per transaction as a result of an investment, and you expect the number of transactions to grow as a function of normal business growth, you should increase savings accordingly over the model's time horizon.


Return on Investment Analysis is a numerical technique for predicting the expected financial outcomes of an investment. A model may be developed which enables the comparison of several options based on the same criteria. The process of developing the model is perhaps more important than the actual model which is generally a combination of various standard formulae employed to analyze anticipated costs and savings over time.

The process has the following steps:

  • Review the overall corporate strategy
  • Review specific functional strategy and objectives in the area where the investment is to be made.
  • Review objectives and directions of functions related to the area where the investment is to be made.
  • Identify project goals based on strategy and objective reviews
  • Develop one or more quantifiable measure for each project goal.
  • Incorporate quantifiable measures into an ROI model

It is the up front work related to strategy, objectives, and relationships to other business functions that will enable you to make the investment decision based on more than just the internal rate of return, net present value, or payback period. The process of developing the ROI model will help to ensure that the functional benefits you attain through the investment and project implementation are consistent with the direction of your firm.

Appendix A

An example of an ROI model (This model is not yet available on the World Wide Web. If you would like a copy please call me at (703) 486-8497

The following are the descriptions of each of the Summary of Annual Cost Reductions section line items. They are examples of objectives associated with the implementation of a distribution control system. The heart of any ROI model is its objectives.

When a functional improvement is being accomplished through technology, it is equally important to ensure that the approach taken is consistent with other supporting functions. In the example of a business application, the functionality is how the investment will achieve the paybacks. However, a platform or language that is not familiar to the information systems staff or is not consistent with their strategic direction may add additional support costs and internal resistance.

A) Profit Gains

In a distribution environment sales are adversely impacted by problems such as shipping the wrong product, shipping the product late, and invoice errors. Often an objective for implementing a distribution control system is the reduction and elimination of problems that result in lost sales and therefore profits due to poor order fulfillment. The system provides a competitive advantage that can reduce lost sales and in many cases increase sales through customer loyalty and a larger share of a customer's business. Customers want to do business with reliable suppliers.

reduce lost sales by .5% to 5% (typical range) of total sales.

B) Floor Line-Offs/Backorders

Floor line-offs occur when stock cannot be found or is not shippable because its inventory status is not known. Additional order administrative costs are incurred when an order cannot be shipped complete.

reduce number of orders with line-offs due to inventory accuracy problems thereby reducing unnecessary additional administrative ordering processing costs..

C) Normal Transportation Costs

When product must be shipped separately due to a line-off additional transportation costs are incurred.

reduce the number of line-offs due to inventory inaccuracy and thereby reduce unnecessary transportation costs.

D) Air Freight Costs

Often line-offs and short shipments due to inventory inaccuracies at the location level result in emergency shipments via air freight.

reduce the number of emergency shipments and their related costs due to inventory inaccuracies.

E) Accounts Receivable

Days outstanding on accounts receivable can be reduced by reducing the number of disputed invoices related to order fulfillment inaccuracies.

improve customer service in the area of order fulfillment accuracy thereby reducing the number of disputed invoices and the amount of capital tied up in accounts receivables.

F) Trailer Loading

Various functions of distribution control systems such as cross docking, system controlled dock floor areas, and reverse sequenced order staging reduce the amount of time it takes to load a trailer

reduce the labor costs associated with loading trailers.

G) Overtime Hours

Overtime in a distribution environment is often caused by manual problems related to receiving, putaway, replenishment, picking, packing, and shipping. Directing these functions through the discipline associated with a distribution control system reduces overtime requirements

reduce overtime hours by streamlining operations through system directed disciplines.

H) Searching

Distribution environments that do not employ system controlled directed putaway and picking inherently have some degree of searching for product at location.

reduce the level of searching and optimize location travel sequence through system directed putaway and picking.

I) Annual Physical Inventory

Distribution control systems enable cycle counting for verifying accuracy and identifying procedural problems. As a result the annual full physical and related costs can be eliminated.

Eliminate the full physical inventory count and related costs.

J) Shipping Errors

Shipping errors such as wrong quantity or product shipped effect customer service levels and also have associated with them additional transaction processing costs.

improve shipping accuracy thereby reducing administrative processing costs associated with order corrections.

K) Inventory

The implementation of a distribution control system and the resulting improvements in inventory accuracy enables the reduction of inventory levels with improved customer service levels. Reduces inventory levels translate into lower levels of capital invested in inventory.

improve inventory accuracy in order to reduce inventory levels and therefore the amount of capital invested.

L) Safety Stock

Safety stock is used to hide inventory inaccuracies and support unexpected surges in demand. Distribution control systems enable firms to achieve very high inventory accuracy levels and, through cross docking respond quickly to receipts, passing them through the facility from receiving dock to shipping dock.

Reduce or eliminate safety stock levels and the investment in them through improvements in inventory accuracy and the ability to quickly receive material and direct it immediately to the shipping dock.


Other objectives not included in this model might include:

  • Reduce costs of maintaining a legacy system or outdated equipment
  • Eliminate penalty costs of not being in compliance with a legal, industry or key customer requirement. Common examples here include hazardous material reporting requirements and Advance Shipment
  • Notification (ASN) labeling and EDI requirements.
  • Delay a capital expense for expanding a facility by improving cube utilization and throughput of existing facility.
  • Reduce inventory writeoffs. Lower inventory levels translate into lower writeoffs due to obsolescence. This is especially important in an environment where the product life-cycle is very short.

Appendix B

Strategic Modeling

John Martin of IBM recently published an article in the December 1993 issue of APICS - The Performance Advantage titled "Use Strategic Modeling to Evaluate Warehouse Automation." This article describes financial modeling at a strategic level. That is, the evaluations of various options (in this case the examples are warehouse automation solutions) take into account the firms's overall financial position. Outcomes of various options are analyzed based on their impact on the balance sheet and income statement and related ratios.

The premise of "The Process of Developing a Model for ROI Analysis" is that strategic objectives must be incorporated into ROI models as they relate to desired functional outcomes of investment options. The model will then be valid and be a useful tool for comparing investment options. John Martins article provides additional insight into how investment options can be analyzed for their overall impact on an organization's financial performance.

The distinction between the two is that the first incorporates strategic objectives into a model used for tactical analysis. The latter provides tools for analyzing tactical outcomes on strategic performance.

Suggested Readings

Blomquist, J.A. and T.W. Speh, "The Financial Evaluation of Warehousing Options: An Examination and Appraisal of Contemporary Practices," Miami University, Oxford Ohio: The Warehousing Research Center and Affiliate of the Warehousing Education Research Council, May 1988.

Burr-Brown Corporation, Economic Justification Workbook, (Tucson: 1991).

Chase, R.B. and Aquilano, N.J., Production and Operations Management, A Life Cycle Approach, (Homewood, Illinois: Richard D. Irwin, Inc., 1981), pp. 117-124.

Dadzie, K.Q., and Johnston, W. J., "Innovative Automation Technology in Corporate Warehousing Logistics," Journal of Business Logistics, Volume 12, Number 1, 1991, pp. 63-82.

Martin, John, E. "Use Strategic Modeling to Evaluate Warehouse Automation," APICS - The Performance Advantage, December, 1993, pp. 19-22.

Mentzer, J.T. and B.P. Konrad, "An Efficiency/Effectiveness Approach to Logistics Performance Analysis," Journal of Business Logistics, Volume 12, Number 1, 1991.

Copyright 2003, Latitude Associates