Response to "Real Numbers or Estimates" (October 15, 1996 Views)

On Friday, October 18, 1996, I received a FAX from Mr. Jack Borcherding, J. B. International Market Research Consultant. Jack agrees that real numbers are always better than estimates, but thinks I oversimplified the positioning. As he stated, projectionability is a problem with all studies, even those of real numbers. In general he believes the size of the study should reflect the risk involved and we should not be looking to one method as being the answer. He believes in sequential testing where you weed out the losers early (in concept testing) then use Real Number Tests but with a lot more analysis of what the numbers are really saying (not just face value).

Generally I agree with Jack, especially in respect that no idea should ever live or die as the result of a single test. Also the size of a study is a subjective call and depends greatly on the impact of the brand in the market. Tests can be too small to read and they can miss an important population segment. On the other hand they can be so large that they attract a lot of competitive attention. My experiences with test markets are that you learn if you can make the product (effectively) and sell it to the customers. Where I do not like test markets, is in the projectionability of the sales results. It is my experience that they attract a lot of attention and competition will do everything possible to upset the study. They can and do things in the test market that they could not do on a national basis. For example, one of my last test markets involved a new product category for P&G. The major competitor, a week before the introduction of our test market, introduced a sales special of their brand (the #1 brand in the category). The sales program was a "two for one" sale. Two bottles of the competitor's brand essentially took purchasers out of the market for over six months.

The potential effectiveness of a two-store study is best seen in the last new market entry I worked on at P&G. This involved a reformulation and repositioning of an existing brand. We had conducted two simulated test markets. Both showed very little change in potential sales. The team, Mr. Steve Donovan (VP P&G), Mr. Keith Grimes (Section Mgr. Now Director P&G), Mr. Gary Walker (Brand Mgr. Now VP Dow), and myself, thought the move to be more beneficial than what was being reflected in the two studies. As a result we conducted a six month in-store test of the new positioning. The results, without any direct competitive interference, indicated a doubling of sales. We went to the market. The result was an increase of 190% in sales. But most importantly, it resulted in a paradigm shift in the product category. This was the introduction of the "ULTRA" concept in the laundry market. For the good or bad, had we not conducted the two-store study and accepted the simulated market results, we would probably not have all the "Ultra" product categories today.

Back Issues of "Views"

For the newer members of my FAX list and any of the older ones interested in back issues of the Views papers, Mr. Herb Sorensen of Sorensen Associates Inc has put together a book of the Views from January 17, 1993 to the present. A copy is available free of charge by calling Herb at 800-542-4321 or FAX a request to Herb at 503-666-5113.

Webmaster's note: Back issues are now also available online via the Views From the Hills of Kentucky home page.