sample="quota" bates="675109665" isource="bw" decade="1980" class="ni" date="19830603" BROWN & WILLIAMSON TOBACCO CORPORATION INTERNAL CORRESPONDENCE TO: G. T. Reid DEPT. C. C. to D. N. Finley B. J. Morrison B. R. Pellett G. Herzog I. B. Brand R. Vail (C&W) C. L. Lomicka B. A. Sproule C. Dillon (C&W) B. Ryan A. A. Simon K. Conorqui (C&W) FROM R. P. Medicus DATE: June 3, 1983 SUBJECT SUMM Strategies for KOOL I. BACKGROUND A. WHAT IS SUMM SUMM (Single Unit Marketing Model) is a model that determines what characteristics of a brand should be advertised to a consumer in order to increase the probability that they will buy that brand. In early 1982, approximately 1000 smokers were identified through a national random sample. B&W smokers were oversampled to increase reliability and then weighted back to their true share of market. All smokers were weighted by the amount of cigarettes they smoke. The model is based on 32 pre-determined attributes - 26 product and 6 imagery - related. These attributes are used to establish 2 types of profiles - a desire profile and a belief profile - for each smoker. The desire profile indicates what a smoker wants and how much he wants it. (Each respondent rates the desirability to him of each variation of the 32 attributes and the importance of the attributes overall.) The belief profile indicates the extent to which a smoker believes various brands have the attributes he wants. (Each respondent rates his own brand plus 12 "banner" brands , indicating which attributes he perceives each brand to have.) The model is used by conducting "what-if" simulations. Several simulations have been played for KOOL and are reported in this memo. B. LIMITATIONS OF SUMM FOR KOOL Two important factors must be taken into account when applying SUMM to KOOL. KOOL Filter King, BARCLAY, BARCLAY Menthol, BELAIR, RALEIGH, RALEIGH Lights, VICEROY, VICEROY Rich Lights, Marlboro, Virginia Slims, Carlton, and More. 1) Only KOOL Filter King is rated by all smokers. Therefore all simulations are based on this style only. 2) The demographic profile of KOOL Filter Kings and SUMM differs markedly from weal world: a) Blacks may be uderrepresented in the model as a whole. b) Young smokers are underreepresented in the SUMM KOOL Filter King franchise. While these subgroups are underrepresented these data do provide useful insight into possible future actions. Simulations should be used as guidelines or for hypothesis generation from which further testing can be done. II. OVERALL FINDINGS Of the product-related simulations, the following yielded the most positive results for KOOL: - increased satisfaction - package change - improved aftertaste - improved tobacco quality - more pleasant aroma Of the simulations that required creation of a KOOL line extension, the following yielded the most positive results: - increased low tar awareness (KOOL Filter King was the only KOOL style rated. In order to determine the leverage for increased Lights awareness, it is necessary to treat "low tar" as a line extension.) -10's pack Imagery related to attributes, by themselves, do not appear to have much impact in terms of brand share change. III. HOW TO INTERPRET THE TABLES Four columns are listed on the table of simulations: topic, strategy, simulation, and net share change. TOPIC - the attribute(s) that is being changed. STRATEGY - the reason for the change. SIMULATION - the "input" into the model based on assumptions of the KOOL research group. NET SHARE CHANGE - the output from the model; i.e., the result of how our changes impact consumer beliefs and desires Net Share Change should not be interpreted as an absolute number. Rather, it indicates the magnitude of one change relative to others. NOTE: Given that KOOL is a large established brand, it is probably not realistic to expect that all smokers beliefs can be changed. Therefore, all simulations have been played with the assumptions that only 30% of smokers who have a belief about KOOL will be affected. How to read the table of simulations: (An Example) (1) Tar/Flavor is one of the 32 attributes smokers evaluated KOOL on. Given the performance of KOOL low-tar styles to date... (2) a strategy might be "what happens if we make smokers more aware of a KOOL low-tar line extension". (3) the simulation is to take existing beliefs about KOOL Filter King and create a line extension with the same characteristics...except for tar/flavor. Here, smokers now perceive that this line extension offers low-tar and a lot of flavor. (4) The result is a net gain of +.20 for the KOOL family. IV. SIMULATION DEMOGRAPHICS In simulations that resulted in large share gains (>10) or losses (-.08), we have examined where these gains or losses are coming from. Four demographic breaks are addressed: gender, age, tar level, and type (menthol vs. non-menthol). Simulations that result in gains primarily from a given subgroup are summarized below. V. CURRENT PERCEPTIONS OF KOOL VS. OTHER BRANDS KOOL is better known than most of the remaining 11 brands (except Marlboro) rated by all consumers. Relative to Marlboro, KOOL is perceived to have - a less attractive package - a less easy draw - lower quality tobacco - less smoothness - a less pleasant aftertaste - less satisfaction R. P. Medicus