[Update 2007/02/12: Welcome TamsPalm readers from *Carnival of the Capitalists*. We hope you like what you find and stick around. Let us know what you think!]

Return on investment calculation is critical to using ROI for prioritizing requirements. We’ve discussed how to forecast return on investment by estimating costs and predicting benefits. Here are five tips to help you when calculating return on investment.

## Summary of ROI Tips

The following ROI calculation tips are detailed below:

- Recognize the Risks
- Discount Future Cash Flows
- Separate Sales From Expenses
- Overcome Ozymandias Syndrome
- Ignore Infinite Elvises

Here are the details…

## 1. Recognize the Risks

You can’t predict the future with certainty. You have to account for the range of possible returns on our investment, or the risks to those returns. One way to account for risk is to assign a set of probabilities to the possible outcomes, and determine the expected value of the return. If you believe there is a 50% chance that your investment will return $1,000 and a 50% chance that it will return $2,000, then the expected value of your investment is $1,500.

Another approach would be to use PERT estimation to calculate your expected return. To use PERT, you identify the worst case, most likely case, and best case returns. You may not have the data you need to accurately calculate a return with a PERT estimate. If you do not have the data, you should use a simple expected value calculation.

The PERT estimate is a representation of a beta distribution. The math is pretty complex, and applying it to a single estimate can give us a false sense of precision. Remember that the numbers we put together in the first place are our best guesses, so doing more precise math on rough estimates doesnâ€™t give us a more precise PERT estimate, just more math.

The beta distribution is very similar to a normal distribution (the familiar bell curve), when it is balanced. This similarity is used to simplify the application of PERT. The standard simplification in using PERT is to treat estimates as if they represented the following probabilities:

- Optimistic: 5% of the time we will complete our task in less than the optimistic time estimate.
- Likely: 50% of the time we will complete our task in less time than the likely time estimate.
- Pessimistic: 95% of the time we will complete our task in less than the pessimistic time estimate.

The discussion above applies to calculating the returns. We also need to calculate the costs in order to determine ROI.

Costs are usually represented by a combination of known expenses (investments, purchases, etc) and predicted expenses (labor to create something, relative labor increases, etc). There may also be probabilities (risks) associated with your cost forecasts. You should use the same approach for predicting costs as you do for predicting returns. It is more likely that you will be able to use PERT estimation for labor cost estimates.

Our costs can also be a combination of fixed costs and variable costs. We have to take that into account as well, and again, adjust for risk.

## 2. Discount Future Cash Flows

A dollar a year from now is worth less than a dollar today. Think about it – if I offered you $100 today in exchange for $100 a year from now, you would take it. But if we reverse the exchange, you’d be foolish to take it. Economists call this *discounting* future cash flow. To determine *today’s* value of *tomorrow’s* dollar, you compute the net present value (NPV) of that dollar. A future dollar is discounted by the interest rate that you could earn if you chose to invest your money (risk free) today. Check out our net present value tutorial for more details.

## 3. Separate Sales From Expenses

Sales dollars and expense dollars are apples and oranges.

At first glance, the following two investments might look to be equivalent:

- Spend $10,000 to increase sales revenue by $50,000.
- Spend $10,000 to cut costs by $50,000.

Sales dollars represent *top-line* revenue (money coming in the door), and expense dollars represent *bottom-line* earnings (money going out the door). What many people overlook is that sales dollars do not equal profit dollars. Profit dollars are usually called *margin* dollars and are equivalent to expense dollars when calculating return on investment.

## Determining Margin Dollars

When you sell a product you have costs associated with that sale. Those costs can be separated into fixed costs and variable costs. The fixed costs are fixed – an increase in sales will not change them. But variable costs will increase with increasing sales. For every new dollar in sales revenue you get, you also incur additional variable costs. The difference between sales dollars and variable costs is called *gross margin* dollars.

An increase in sales of $1,000 with a corresponding increase in variable costs of $800 would yield a gross margin increase of $200. The $200 is your increase in profits – and this is the amount you should use in your ROI calculations.

## Determining Expense Dollars

Your investment might be a cost reduction. If so, every dollar you save flows through to the bottom line as a dollar of profit. These are known as *net margin *dollars.

When an investment will both increase sales and reduce costs, you can’t add the numbers together – you can only add the profit dollars together. The same holds true if your costs are going up (say your investment is to hire an additional salesman). You have to calculate your return on investment based on the increase in gross margin dollars relative to the increase in expenses.

## 4. Overcome Ozymandias Syndrome

[Photo taken by Hajor, December 2001. Released under cc-by-sa and/or GFDL]

Nothing lasts forever. You may have a proposal that will reduce costs by $100,000 per year. What is the total savings? $100,000, $500,000, infinity? For how many years are you predicting that the savings will be realized?

I always think of this quote from the poem Ozymandias:

[...]

Look on my works. Ye Mighty, and despair!’

Nothing beside remains. Round the decay

Of that colossal wreck, boundless and bare

The lone and level sands stretch far away.

Nothing lasts forever. There are three time limits that you can apply to constrain your ROI calculations.

- Limit by Uncertainty
- Limit by Financial Impact
- Limit by Compelling Event

## Limit by Uncertainty

You can’t predict the future. Even with good estimates for potential savings, growth and costs, you can’t predict *outside factors* that could make your investment irrelevant. Netscape couldn’t predict that Microsoft would offer Internet Explorer for free. Our younger readers should know that browsers used to cost money. Microsoft’s action all but obliterated the business model for Netscape. The company certainly imploded.

In software, three years is about as long as you can conservatively plan on *relevance*. The IRS guidelines allow for depreciation of a software purchase for either three or five years. This makes for a good sanity check – in theory, your capital purchases should be completely depreciated at the time that they have no value.

## Limit By Financial Impact

When you calculate net present value, you applied a discount rate to determine the value *today* of a cash flow stream from the future. When calculating NPV, you see that a dollar next year is worth less than a dollar this year. Similarly, a dollar two years from now is worth less than a dollar next year. On some future date, that future dollar will be immaterial in today dollars.

There isn’t a firm definition of “material financial impact”, but 10% feels like a reasonable number to us. In other words, the first year in which a future dollar is worth $0.10 or less is the first year to ignore. If your discount rate is 10%, that would be year 25. If your discount rate is 20%, you would look no more than 12 years into the future.

In software development, the *relevance* factor will always constrain us first. In industries with longer time horizons (like transcontinental shipping, or toll-road building), the financial limit is more likely to be the more constraining item.

## Limit By Compelling Event

A compelling event would be a change in strategy, the elimination of the need for the product you’re evaluating, or *another* future improvement that will be implemented. In short, anything that drives obsolescence.

## 5. Ignore Infinite Elvises

Here’s a quote from 1994:

When Elvis Presley died in 1977 there were 37 Elvis impersonators in the world. Today [Ed: 1994] there are 48,000. If the current trend continues, by the year 2010, one out of every three people in the world will be an Elvis impersonator.

We’re running out of time.

There is a good point in here – there is huge danger in extrapolation of trends. Even when you think you understand the underlying causes of growth, it is unreasonable to expect the trend to continue. Venture capitalists regularly rant and/or joke about people who project *infinite Elvis* growth, almost as much as they complain about people who can’t support why they will get “just 1% of a zillion dollar market.”

## Summary

You can’t over-estimate the importance of return on investment calculations when making decisions. Many of our articles touch on ROI, even when we’re writing about seemingly unrelated topics. In addition to those articles, remember these 5 tips for improving your ability to calculate return on investment:

- Recognize the Risks
- Discount Future Cash Flows
- Separate Sales From Expenses
- Overcome Ozymandias Syndrome
- Ignore Infinite Elvises