Home » Business

Develop Proprietary Measurements to Anticipate Shifts in Powerful Trends

Mar 28, 2008
Barometers measuring irresistible forces that aren't available to others will be more useful to you than ones that everyone can access. If everyone learns how to forecast an event, behavior will change, and the event will no longer turn out the same way. To make the process easier for you to understand, the emphasis so far has been placed on publicly available measures. Now we turn our attention in different and less visible directions.

What proprietary irresistible force barometers can you find that work better for your business than publicly available ones?

The best barometers are those that can be devised from what occurs within a business itself, and will be invisible to others. In the oil rig business, you might find that the number of requests you have to bid on rig rentals from a certain set of companies where you are the dominant supplier could be a uniquely effective advance indicator of oil company perceptions of future oil prices.

Notice that here we are focusing on oil company perceptions rather than actual oil price trends. This kind of shift in focus would have been preceded by an analysis that showed that oil company perceptions of future oil prices were a more powerful irresistible force than the actual trend in oil prices. If that is the case we should be happy, because oil companies historically have been bullish about oil price trends a lot more often than those trends actually occurred.

The best way to get started is to assemble all the internal data that you have been keeping track of for a long time and compare those data to your company's later performance, employing a time lag between events and results. In the oil rig example, the performance you want to predict could be defined as placing more rigs at higher prices for longer periods of time.

You could develop an index that captured this pattern. Then you would use single variable statistical regressions to compare that success index to earlier patterns within the business, such as the previously mentioned number of requests to bid on rig rentals. If you don't know how to do these statistical regressions, chances are that some of the recent business, economics, or math majors in your organization do. They can help you.

Start working with them by showing them this article. The statistical relationships that emerge should be tested for the frequency with which they give accurate signals, as well as the logic of why they should be related.

Once you begin to identify some of these relationships as barometers to the success index, you may find it helpful to check if combinations of the barometers have worked better in the past than individual barometers. For example, if all of them give you the same signal, then the forecast's accuracy will usually be higher than if only a few gave that signal.

You can also employ multivariate regressions (using more than one variable at a time to match your success index) to improve the weight you give to different factors. If you are working with people who do not know the subject area well for these regressions, be sure to check that the relationships make sense. Multivariate regressions will often produce some statistical relationships that are probably wrong (such as if the regression says that a faster growth rate of the economy leads to lower petroleum prices).
About the Author
Donald Mitchell is an author of seven books including Adventures of an Optimist, The 2,000 Percent Squared Solution, The 2,000 Percent Solution, The 2,000 Percent Solution Workbook, The Irresistible Growth Enterprise, and The Ultimate Competitive Advantage. Read about creating breakthroughs through 2,000 percent solutions and receive tips by e-mail by registering for free at

http://www.2000percentsolution.com .
Please Rate:
(Average: Not rated)
Views: 122
Print Email Share
Article Categories