Magazine article CRM Magazine

Forecasting a Flap: Econometrics Has Long Been the Province of Ph.D.S, but Applications for Marketers Are Emerging Via Automated User Interfaces and B1 Reporting Tools

Magazine article CRM Magazine

Forecasting a Flap: Econometrics Has Long Been the Province of Ph.D.S, but Applications for Marketers Are Emerging Via Automated User Interfaces and B1 Reporting Tools

Article excerpt

There's a chaos-theory concept that refers to how a small, seemingly unimportant action in one part of the universe creates massive change in another part of the universe. The idea is this: Does the flap of a butterfly's wings in Brazil set off a tornado in Texas? The thinking is that the infinitesimal movement of air from the flap starts a reaction that ends with something in another part of the world being affected by the wing's movement. Coined to help illustrate the relationship between two seemingly unrelated variables, the idea has lately become a way to describe the often complex relationships that exists between consumers and the marketplace.

For years, companies have sought to understand the relationship between marketplace factors and their customer bases' propensity to do business with them. Fortunately for businesses, the development of metrics applications over the past three decades has provided forecasters with the tools to take much of the mess out of figuring out the marketplace. These tools are forecasting and econometrics apps.

The fiercely competitive nature of today's business landscape is forcing organizations to operate more efficiently. Years ago the data to segment customers was often inaccurate and the ability to pull results from econometric systems on a daily basis was virtually nonexistent. Companies are now looking to the past to predict the future by using econometric and forecasting solutions to measure the impact of economic and marketplace factors. The emphasis is on automation and simplification--it's becoming more about the pull and less about the push. Results are being drawn from forecasting and econometric applications on a monthly, weekly, or even daily basis, as opposed to being pushed to a department by management once a quarter.

Mark Lush, a principal within Deloitte Consulting and coleader of the national marketing solutions practice, states his own, marketing-influenced version of the aforementioned scientific rubrik: "The butterfly flaps his wings in Beijing and suddenly lipstick sales go up in San Francisco. Econometrics is a classic example of the chaos theory. A company's customer base can be divided into literally thousands of segments. Forecasting and econometric tools identify the relationships and linkages between those segments at the highest level--they connect the dots."


Just like their CRM analytical brethren (and sistren), the big push with econometrics today is on making these solutions more innate by automating the selection and customization processes that come with operating these apps. To do their job econometrics solutions use regression and time series analysis to model the relationships between groups of variables, determine the magnitudes of these relationships, and make predictions based on those models. The models by which these solutions slice and dice the data are by no means elementary. They involve heavy doses of calculus and a firm understanding on the part of the operator of both mathematical and economic theory. While too complex to go into at length, the models themselves are data cubes, virtual-3D megacubes of customer data. "We're talking about incredibly complex analytics," Lush says. "They're pushing and pulling, slicing and dicing massive quantities of data based on preset conditions or rules that are determined by the operator."

While the variables change, the models remain a constant. Therefore, the key to these solutions lies in the ability of the operator to select and/or customize the model that will most accurately predict the type of information he is looking for. On the flipside these requirements have meant that these solutions were, for the longest time, relegated to those who had a firm understanding of the theories and mathematics involved in selecting the right model. Additionally, certain models were reserved for certain statisticians, all of which had their specialty within econometrics. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.