Detailed Observations 



Total sales of fresh and frozen shrimp products of the 12 stores 

 participating in the experiment rose from 151.^ pounds during the first, 

 to 2^5 pounds during the second, and to 279.1 pounds during the third, 

 biweekly period. Figure VI - 9 shows both the combined, as well as the 

 separate, increases for fresh and packaged frozen products. The big boost 

 in sales of fresh shrimp effected during the second 2-week interval has 

 to be credited principally to the addition of fresh shrimp to the line of 

 a Chicago supermarket which prior to that time had limited itself to the 

 marketing of frozen breaded uncooked shrimp. 



Separate data for the four metropolitan areas covered are shown 

 in figure VT - 10. Fresh shrimp was handled only by 3 of the 12 stores, 

 2 of them located in New York, 1 in Chicago, Sales of frozen products 

 reacted favorably throughout the duration of the experiment in New York 

 and Chicago J in Harrisburg they dipped during the third 2-week period after 

 a rise during the second 2-week period j in Waterloo they declined originally 

 but recovered substantially in the last 2 weeks to a level well above the 

 first biweekly interval. 



More light is shed on the experiment when the operations of 

 the individual participating stores are einalyzed. Altogether 9 stores 

 out of the 12 audited appeared to benefit from the additional sales 

 stimuli provided. Sales in three stores showed progressive declines 

 over the three survey intervals. The Nielsen Company comments on these 

 differences in results as follows: "It has been our experience that in 

 any test procedure it is unusued. that the consumer sales trend would be 

 identical in all participating stores. Causes are frequently external, 

 resulting sometimes from competitive product, and even competitive store, 

 activities. Over-all gains made during promotions are often ccanposed of 

 raulti -directional trends in all stores included in the sample". 



A special effort was made to identify the reasons for the 

 individual departiires from trend. Tv;o stores in the Harrisburg sample 

 showed significsint decreases in sales daring the third survey interval 

 after a favorable response during the second 2-v;eek period. It was 

 surmised that competitive product promotion around the Thanksgiving 

 season had some influence on this reversal. One of the Waterloo stores 

 failed to add new items on the grounds that the time of the year was 

 unfavorable for the sale of shrimp. The substantial increase in sales 

 in this city during the third period was the result of added newspaper 

 advertising by another local store. Two of the stores in the Chicago 

 metropolitan area reported that their wholesalers were out of stock, 

 one of the stores showing the effects by failiire of its sales to pick 

 up during the course of the experiment. There was no obvious explana- 

 tion for the relatively indifferent showing of one Brooklyn, New York, 

 store in the third biweekly interval. 



36 



