This article aims to discuss what an inexact science forecasting is but still how can business leaders exploit the value in forecasting. While many business leaders tend to fall in love with forecasts (it has something to do with human tendency to believe something more when it has a number attached to it), they must also remember that a forecast is at best a wild guess. No matter the amount of data used or the sophistication of models employed, a forecast is literally meaningless in most cases except to give you a very high-level understanding of a trend. Therefore, if you are arguing about the growth rate being 5.3% or 3.9%, or the market size being 189.33 vs. 165.48, you are essentially wasting your time and focusing on the wrong stuff.
It is in this context that the comments of Stephen Chazen, CFO of Occidental Petroleum, are very insightful. In a recent interview by CFO magazine, take a look at his answers. We are so glad to hear that he does not live in a dream world of forecasts.
Q. What (oil) prices do you use for your own budgeting and forecasting?
Chazen: ...we use three cases: a low-case price, a mid-20s (US$) case, and a high-20s case. But our ability to predict is nil.
Q. The Energy Information Administration said they expected $37 a barrel through 2005. Where do they get their numbers?
Chazen: They're making them up.
This is exactly what happens when consultants and market research firms come up with forecasts that are so often cited by magazines and then end up in all kinds of business plans. The forecasts are essentially made-up numbers based on some manipulation of data and a study of who is saying what in press releases. And we all know that it is not difficult to come up with evidence to support that X is going to happen.
So how to deal with the future?
It is in this context that the comments of Stephen Chazen, CFO of Occidental Petroleum, are very insightful. In a recent interview by CFO magazine, take a look at his answers. We are so glad to hear that he does not live in a dream world of forecasts.
Q. What (oil) prices do you use for your own budgeting and forecasting?
Chazen: ...we use three cases: a low-case price, a mid-20s (US$) case, and a high-20s case. But our ability to predict is nil.
Q. The Energy Information Administration said they expected $37 a barrel through 2005. Where do they get their numbers?
Chazen: They're making them up.
This is exactly what happens when consultants and market research firms come up with forecasts that are so often cited by magazines and then end up in all kinds of business plans. The forecasts are essentially made-up numbers based on some manipulation of data and a study of who is saying what in press releases. And we all know that it is not difficult to come up with evidence to support that X is going to happen.
So how to deal with the future?
- Don't fall in love with your or anyone else's forecast. We at iProceed always tell our clients to use any forecasts (including ours), particularly those that come attached with numbers supposedly accurate to the second point of decimal, for entertainment purposes only.
- Use a scenario planning approach instead and while you can base the scenarios on some numbers, it is best to stick to ranges than hard numbers that are inflexible. (For example, it is best to talk of growth in a range of 1-2X of GDP as opposed to 4.72%.
- There is no evidence that we have found that one should rely more on forecasts generated from interviews with industry participants. Often times, folks in the industry basically regurgitate what they hear from someone else. We would like to point out, though, that by talking to folks in the industry you can sometimes a fair and more real understanding of what the trends are (who is buying what and in what amount, etc.) as opposed to a consultant in a market research firm who has never stepped out of a cubicle, has never seen the product that s/he is forecasting, and is relying heavily on what is out there on the Net.
- While supply is far more easy to forecast than demand (though there are uncertainties involved in this case as well as we are seeing with oil markets), it does not mean that supply data should form the foundation of your strategy.
- Finally, always use more than one source. Never rely on one research firm's forecast and build your business on it. Once you do get a piece of data, keep testing it with industry participants.
Recommended article: How to pick a market research and forecasting firm?