HANDLING UNCERTAINTIES
Unfortunately, most of the numbers that we use in our analyses are not precisely known. In some cases, we can get pretty good estimates, but not always. This is particularly true in projecting revenues. So what do we do? There are a variety of approaches to select from, depending on the purpose of your analysis. All of these approaches derive from our main goal of proving that our project is profitable and better than competing projects.
One approach is to use a worst-case estimate. By worst case, I mean a value so poor that the people reviewing your analysis will concede that reality will probably be better. This takes a bit of psychology to implement. You might firmly believe that your new production line will bring in revenues of $2 million per year, but your comptroller has doubts. He might be willing to concede that it will bring in $1 million per year or better, so use his number; in that way, he won't disagree with your conclusions. Of course, if your analysis shows that you aren't profitable at only $1 million per year, you'll need to find some other way to convince him.
Another approach is the break-even analysis. Run your analysis for various values of the unknown number until it just barely looks good enough. Then you can argue that, as long as the true value is better than the assumed value, your project will be a good one. If the break even value is low enough. this might be an easy case to make. For example, you might have no idea what your revenues will be, but you know that projects with a rate of return over 25% often get funded. Find the revenue stream that gives you exactly 25% return. Let's say it turns out to be $1.5 million per year. Now your pitch is, all we need is revenues of $1.5 million per year to make at least 25%. If your audience finds it easy to believe in the $1.5 million, they will accept your pitch.
The what-if analysis is another way you might choose to go. For the unknown value, make four or five guesses, and run your analysis each way. Put the results into a table, showing what your rate of return (or present value, or whatever you choose) would be for each possible value of revenue, labor costs, or whatever the unknown number was. In this way, the reviewers of the analysis can make their own decisions about what is likely to happen. Hopefully, if your project is a good one, they will receive a favorable impression of the probable outcomes.
What about inflation? This is certainly an unknown, especially over a long term project, say 20 years or more. Fortunately, in many cases, we can simply ignore it. Costs will go up over time in the face of inflation, but so will revenues. In many situations, they will approximately cancel out. This is especially true when our costs are heavily weighted toward the early years of the project and the incomes are weighted toward the later years; in those scenarios, ignoring inflation gives us a conservative estimate of profitability. Annual cost streams, such as mortgage payments, remain constant in terms of dollar values, but decrease over time with respect to value, so again we are in a conservative estimate situation. The only time ignoring inflation will get you into trouble is when you make your revenues up front and have payments spread out into the future. Fortunately, those situations are rare in the engineering world, and mostly plague financiers. Handling this situation is not difficult, but is beyond the scope of our discussion here.
The bottom line on handling uncertainty is to remember that your goal is to argue convincingly that your project is a winner; if you use conservative estimates that nobody could disagree with and your analysis still shows a profit, you've made your point.
THE SOFTER STUFF
Those numeric quantities discussed above weren't too hard to come by; at the very least we can get some arguable estimates. However, other issues can be much fuzzier. First and foremost, is there business out there for you to go after in the first place? You'll need to assemble an argument showing that the revenues you describe will actually come through; that you're offering something the current suppliers are not offering (better price, quality, delivery schedule, convenience, etc.). If there are no current suppliers, how do you know there are any customers? This work has probably already been done for you by Marketing; see what they know.
Consider symbiotic effects of this venture, both good and bad. Will you be cannibalizing your own business in other areas? If your division makes a killing, but another division in your company is the one that gets killed, your board of directors will not be impressed. On the other hand, symbiotic effects can be good; two products that go together, from the same supplier, can boost market share for both products (just ask Microsoft).
Sometimes you need to invest in a project just to stay in business at all. For instance, setting up a recall center or a warranty support office does not make you any money on its own, but it lets you remain competitive in your main line of work. In that case, you don't need to show a profit, just that you are getting the job done for the least cost possible. You are competing not with other revenue-generating ideas, but with other solutions to the same problem.
Finally, always consider the reputation and image of your organization. If your short-term profits soar but your good name suffers, you might be in trouble in the long run through loss of other business. Will you be teaming with someone your customers cannot abide? Are you selling to your best customer's prime competition? These do not necessarily make a project untenable, but they do indicate that you should give it some serious thought.
