• by July 29, 2005

By John Nardone

With so much of the ad world's attention currently focused on marketing accountability, you might expect a widespread use of marketing econometrics to measure and manage return on investment. Unfortunately, depending on the industry being served, this is not always the case.

Marketing econometric modeling is well established in the packaged goods industry. After the technique was pioneered in the early '90s by Marketing Management Analytics (working with General Foods, Nabisco, and Kraft), packaged goods marketers throughout the industry adopted the approach for marketing mix and price/promotion analysis. In the last few years, as marketers familiar with this technique (particularly Kraft alumni) have moved to other companies outside the packaged goods world, these marketers have begun to develop its use in other industries. But adoption is far from widespread, as is evident in the chart, right, where I quantified my perceptions of various industries' use of modeling  from the potential dollar impact of using the tool compared with how much interest in modeling companies in those industries had expressed, to the industries' actual use of the tool.

For example, retailers have unmatched potential to benefit from modeling, because they own their own detailed transactional data down to the store level. Store-level modeling can provide insights to optimize store locations, product assortment, and merchandising, in addition to marketing mix.

Pharmaceutical products have a limited life span in which to maximize profits before patents expire. Yet detailed physicians' data allows for sophisticated analysis. As such, modeling for pharmaceutical companies requires special skill and handling, but the benefits are substantial.

Telecom companies have dabbled with marketing models in the past, but now tend to be focused on the operational issues associated with recent mergers and acquisitions. Once they finish merging brands and historical data, they can leverage rich transaction data to build models that maximize the profit potential of key customer segments. Consumer electronics marketers, on the other hand, appear to be data-challenged; there is little sales data that cover key distribution channels.

In the entertainment sphere, movie studios and tv networks have become quite interested in modeling to help manage promotion. Unfortunately, since every entertainment "product" is different, modeling can only identify high-level trends and general conclusions.

Like telecom, retail bankers are assimilating many mergers/acquisitions. Nonetheless, they seem determined to go forward with modeling, to the extent their data allows. Risk modeling is an integral part of the banking world, so while marketing modeling is new, it is easily understood and is accepted quickly.

Major credit card companies have been using modeling for years to understand the media impact on charge volume. Now more card issuers are adopting the technique to help understand customer acquisition. It seems as though broker-mediated service industries have not yet caught onto modeling. Travel and hospitality marketers are just beginning to explore how marketing mix models can be added to create more complete and integrated insights for running their businesses. The auto industry has great potential for the use of this technique, but is doing very little with it.

John Nardone is executive vice president, product development and marketing, for Marketing Management Analytics. (john.nardone@mma.com)