Tobacco Documents Online

n°~~ SECRET No. 3 31 By._-Z,, Declassifica#ion s New Brands and Strategic Research Report TO: Mr. A. T. Sterling FROM: Ms. M. A. Albert Mr. R. F. McCarthy ' PROJECT XG FFNM 85 SP PROTOTYPE TEST FINAL -REPORT COPIES TO: Mr. L. W. Hall, Jr. Mr. E. J. Fackelman Mr. J. D. Weber Mr. H. T. Parks Ms. R. S. Turlington XG Fact Book MDIC PUBLISHED BY THE MARKETING DEVELOPMENT DEPARTMENT R.J. REYNOLDS TOBACCO COMPANY, WINSTON-SALEM, N.C. 27102 P NJAT Form 7368 -*tOJet

PROJECT XG FFNM 85 PROTOTYPE TEST EXECUTIVE SUMMARY (MDD #85-42606) PURPOSE: To report final results on the performance of two XG FFNM 85 SP prototypes (CT 552-A and B) versus Marlboro King among target smokers. BACKGROUND: In recognition of the importance of younger adult smokers to RJR's growth, Project XG will be positioned to compete aggressively with Marlboro. To effectively meet this goal, Project XG must incorporate a superior product which is superior in smoothness to Marlboro, while providing comparable strength and tobacco taste. Plans currently call for XG to be test marketed in January, 1986. Brand is preparing for test market by developing/testing prototypes in three configurations: FFNM 85 SP, FFLTNM 85 SP, FFNM 80 Box. RESEARCH OBJECTIVES: The objectives of this test were to: • Provide a quantitative assessment of the two XG FFNM 85 SP prototypes (CT 552-A and B) versus Marlboro King (and WINSTON King) among male Marlboro 85 smokers, aged 18-24. • Provide input into further development and refinement of the XG products. KEY CONCLUSIONS/FINDINGS: • Neither prototype (CT 552-A and B) achieved action standard of superiority versus Marlboro King among male Marlboro 85 smokers, aged 18-24. Both prototypes were rated inferior to Marlboro King on overall acceptance. While Prototype 1 (552-A) was also inferior to WINSTON King, Prototype 2 (552-B) was rated at parity with WINSTON. The prototypes performed at parity with each other. Marlboro King WINSTON King Prototype 2 (CT 552-B) Prototype 1 (CT 552-A) 7+ Among Sample Target Size Smokers 300 60.5 < 300 > 58.5 ~ 297 C 54.3 282 49.6 ] < Significant difference (80% confidence level) with arrow pointing to superior product.

: I • Versus Marlboro King The prototypes performed similarly versus Marlboro King on most product attributes. Although Prototype 1 (CT 552-A) was inferior to Marlboro King on overall performance, it was, surprisingly, at aQ rity on all but one product attribute--where the prototype had the advantage of more natural taste than Marlboro. Relative to Marlboro King, Prototype 2 (CT 552-B), which was also inferior on overall performance: -- Was, surprisingly, at ap rity on all but four attributes. -- Had the advantages of less harshness and a less strong aftertaste. -- Was milder. -- Had only one disadvantage -- harder to get smoke through the filter. - Both prototypes were aQ rity with Marlboro King on key attributes of: -- Smoothness Not attaining superior goal -- Strength ~ Achieving parity goal -- Tobacco Taste • Versus WINSTON King - Although Prototype 1 (CT 552-A) was inferior to WINSTON King on overall performance, attribute performance did not support the inferiority rating. The prototype was parity with WINSTON on all but two attributes -- where it had advantages of fresher aftertaste and easier to get smoke through filter. Prototype 2 (CT 552-B) was parity with WINSTON King on overall performance and was at parity with WINSTON on all but four attributes. Versus WINSTON King, the prototype: -- Had the advantages of more smoothness and a fresher aftertaste. -- Was milder. -- Had only one disadvantage (less strength). . A thorough follow-up analysis was conducted to examine why XG FFNM 85 SP prototypes (CT 552-A and B) were inferior to Marlboro King, but attribute performance did not fully support inferiority. Overall, the follow-up analysis provided no major insights to the discrepancy between %70+ and attribute ratings. MDII RECOMMENDATIONS/NEXT STEPS: • To further follow-up on the discrepancy between %70+ and attribute ratings for Prototypes 552 A and B: - Sensory Panel Testing of CT 552-A, CT 552-B, and Marlboro King has been requested. - Product testing is continuing to address this issue across brands. We will report information as it becomes available. ~kax~ a, Qul__ Melanie A. Albert Marketing Development Department

: XG FFNM 85 SP PROTOTYPE TEST MDD #85-42606 BACKGROUND: In recognition of the importance of younger adult smokers to RJR's growth, Project XG will be positioned to compete aggressively with Marlboro. To effectively meet this goal, Project XG must incorporate a superior product which is superior in smoothness to Marlboro, while providing comparable strength and tobacco taste. Plans currently call for XG to be test marketed in January, 1986. Brand is preparing for test market by developing/testing prototypes in three configurations: FFNM 85 SP, FFLTNM 85 SP, FFNM 80 Box. • XG FFLTNM 85 SP In recent testing, none of four prototypes (CT 443-A and B; MDD #85-42602; CT 558-B and C; MDD #85-42606) achieved action standard of superiority versus Marlboro Lights 85 among XG target smokers, as all four prototypes were parity with Marlboro on overall acceptance. • XG FFNM 80 Box Recent testing on WINSTON Box resulted in a superior prototype (CT 446-D), without B-9 in filter, to Marlboro Box (and in-market WINSTON Box) among WINSTON's target smoker group. As a result, this product was put into Phase II manufacturing. The recent post-production test on WINSTON Box (CT 606 A) and the supplement (on WINSTON Box and Marlboro Box) among XG target smokers, however, showed that the Post-Production WINSTON Box product (CT CT 581-A/606-A) was inferior to Marlboro Box among male Marlboro smokers, aged 18-24, not achieving action standard of superiority. • XG FFNM 85 SP In the first test of conventional XG FFNM 85 prototypes (MDD #84-42602/ prototypes CT 444-A and C), neither product achieved action standard of superiority versus Marlboro. As a result, further development/testing is being pursued. R&D developed two prototypes (CT 552-A and 552-B) for testing. Prototype CT 552-A was designed to primarily address strength and smoothness, while Prototype CT 552-B was designed to address harshness and strength as follows: Prototype 1 (CT 552 A) Prototype 2 (CT 552-B) Designed To CAMEL Filter blend without stems High T/N Ratio blend Achieve ideal harshness CENTURY color tipping CENTURY color tipping Non-diluted Non-diluted Increase tar and thereby strength No filter flavor No filter flavor Increase smoothness Part II casing adjust- Increase smoothness ments (Increase cocoa ~ ; and B2, eliminated o 4V licorice and B4) B9 applied at final B9 applied at final ~` top dressing stage top dressing stage o a+ N

4 ACTION STANDARD: Full Flavor XG Soft Pack 85 prototypes should achieve overall superiority versus Marlboro King among male Marlboro smokers, aged 18-24. METHODOLOGY: The standard NFO two-pack monadic design with telephone callback was used in this study. The two XG FFNM 85 prototypes (CT 552-A and B) were each tested among approximately 300 male Marlboro 85 (SP, Box, and 25's) smokers, aged 18-24. Products were mailed out the week of April 29, 1985. Results are compared to Marlboro King and WINSTON King (Waves 1-3/1985 of the Competitive Product Series plus a supplement). This current study used a new product testing questionnaire containing the following new attributes: Hard to get smoke through the filter (replacing "draw"), Fresh aftertaste, Strong aftertaste, Drying aftertaste, Lasting aftertaste. Sample sizes for these attributes are small (about 60-70) for Marlboro King and WINSTON King. DETAILED FINDINGS: A. Overall Performance 1. Among mong male Marlboro 85 smokers, aged 18-24, neither prototype (CT 552-A and B) attained action standard of superiority versus Marlboro Kin . • Both prototypes were rated inferior to Marlboro King on overall acceptance. • While Prototype 1 (552-A) was also inferior to WINSTON King, Prototype 2 (552-B) was rated at parity with WINSTON. • The prototypes performed at parity with each other. Overall Performance 7+ Among Sample Target Size Smokers Marlboro King 300 60.5 < WINSTON King 300 > 58.5 Prototype 2 (CT 552-B) 297 54.3 Prototype 1(CT 552-A) 282 49.6 < Significant difference (80% confidence level) with arrow pointing to superior product. -2-

z B. Attribute Performance 1. Relative to Marlboro King The XG FFNM 85 SP prototypes performed similarly versus Marlboro King on most product attributes (see Attachment 1 for more detail). • Prototype 1 (CT 552-A) was rated inferior to Marlboro King on over- all performance. Surprisingly, however, it was at ap rity on all but one product attribute, where the prototype had the advantage of a more natural taste than Marlboro. • Prototype 2 (CT 552-B), which was also rated inferior to Marlboro King on overall performance, was also surprisingly rated at parity on all but four attributes as shown below: The prototype had the advantages of less harshness and a less strong aftertaste versus Marlboro King. In addition, the prototype was milder than Marlboro King. As shown, the prototype had only one disadvantage versus Marlboro -- harder to get smoke through filter. Parity Superior Inferior Different Strength Less harsh Harder to get (Both at ideal) Smoothness Less strong smoke through Milder Tobacco taste aftertaste filter Bitter taste Rich taste Fresh aftertaste Long lasting aftertaste Drying aftertaste • On key attributes, both prototypes were ap rity with Marlboro King on: - Smoothness -- not attaining superior goal - Strength ~ I achieving parity goal - Tobacco Taste • Both prototypes had ideal levels of mildness. Additionally, Prototype 2 (CT 552-B) had an ideal level of harshness, as desired. • A thorough follow-up analysis (described in Section D) was con- ducted to help explain why the XG FFLTNM 85 prototypes (CT 552-A and B) were inferior on overall performance, but attribute performance did not fully support inferiority. o 49 -3- .v ~ 0 o+ w

z Relative to WINSTON King • Although Prototype 1(CT 552-A) was inferior to WINSTON King, attribute performance did not support the inferiority rating. The prototype was parity with WINSTON on all but two product attributes (where it was superior), as shown: Parity Su rior Harshness Fresher aftertaste Smoothness Easier to get smoke through filer Mildness Strength Tobacco taste Bitter taste Rich taste Artif icial taste Strong aftertaste Drying aftertaste Long lasting aftertaste • Prototype 2 (CT 552-B) was parity with WINSTON King on overall performance and was at parity with WINSTON on all but four product attributes, as shown below. The prototype had the advantages of more smoothness and a fresher aftertaste than WINSTON. The prototype was milder than WINSTON. And, relative to WINSTON, Prototype 2 only had one disadvantage (less strength). Parity Superior Inferior Different Harshness Smoother Less strength (Both at ideal) Tobacco taste Fresher Milder Bitter taste aftertaste Rich taste Artificial taste Strong aftertaste Drying aftertaste Long lasting aftertaste Getting smoke through filter 3. Comparison of Prototypes The prototypes were at parity with each other on overall performance, and were at parity on all but three product attributes -- where Prototype 2 (CT 552-B) had advantages of a fresher aftertaste and less strong aftertaste, but disadvantage of being harder to get smoke through filter than Prototype 1(CT 552-A). -4-

z C. Product Refinement Opportunities , If development efforts were to continue on prototypes CT 552-A and B, key refinement opportunities to attain the desired goal of superiority versus Marlboro King would be as follows: Prototype 1(CT 552-A) Prototype 2 (CT 552-B) Increase smoothness/reduce harshness Increase strength/tobacco taste Increase strength Provide fresher, less long last- ing, less strong aftertaste Increase smoothness Other less important attributes and ETA ratings, which show correlation between %70+ and product attributes are shown in Attachment 4. D. Follow-up Analysis A follow-up analysis was conducted to examine why XG FFNM 85 SP prototypes (CT 552-A and B) were inferior to Marlboro King, but attribute performance did not fully support inferiority. The following areas have been examined to help determine why prototypes were inferior to Marlboro King. • Attribute differences between prototypes and Marlboro at a lower than normal confidence level (65%) and higher than normal level (90%). • Open-ended dislikes • Acceptors (70%) versus Rejectors (60-) • Marlboro Soft-Pack versus Marlboro Box smokers • 21-24 versus 18-24 year old male Marlboro smokers • Comparison of Marlboro King from the two XG FF prototype studies • Distribution of scores - attributes and 70+ • Analytical data • Direction for improvement (ETA's) -5-

z SUMMARY OF FOLLOW-UP ANALYSIS: ^ • Results could not be explained by: - Open-ended dislikes--no major dislikes - Marlboro King's performance did not change between the two XG FF prototype tests - 18-24 and 21-24 year old smokers rated the products similarly • Analysis of distribution of attribute scores showed that Marlboro King distributions appeared to be tighter, while distributions for prototypes were flatter using the scale more broadly. No significant skews in attribute ratings existed. • Both prototypes were inferior to Marlboro King on overall performance among total smokers due to inferiority among Marlboro Box smokers. - Among Marlboro Box smokers, however, both prototypes were parity with Marlboro King on all attributes. • Attribute performance of prototypes versus Marlboro King (.65 CL) indicated that: - 55 2-A Disadvantage Advantages Drier aftertaste (.73) More natural taste (.84) More tobacco taste (.71) Less strong aftertaste (.68) Was milder (.71) (Prototype and Marlboro at ideal) - 55 2-B Disadvantage Advantages Less strength (.67) Less harshness (.84) Less strong aftertaste (.95) Was milder (.91) (Prototype and Marlboro at ideal) -6-

z • Analysis of Acceptors versus Rejectors suggested somewhat more variability in prototypes' taste perceptions than in Marlboro. Across all three products, rejectors found the product to be harsher, less smooth, more artificial, more bitter and less rich than acceptors. Rejectors also found: Marlboro King 552-A 552-B Less Tobacco Taste Less Mild Less Tobacco Taste Harder Draw Less Strength Less Fresh Aftertaste Stronger Aftertaste • Rejectors of Prototype 552-B had more (attribute) reasons to reject the prototype (not strong enough/too mild) than did rejectors of Marlboro King. • Product analyticals reflected only: - 552-A - higher tar than MK or 552-B - Both prototypes - higher moisture and draft resistance than Marlboro King • ETA's provided product development direction: 552-A 552-B Increase smoothness Fresher aftertaste Increase strength Increase strength Decrease harshness Less long lasting aftertaste Less strong aftertaste Increase tobacco taste suggesting an aftertaste problem with 552-B. • Overall, follow-up analysis provided no major insights to discrepancy between 70+ and attribute ratings. NEXT STEPS: • Sensory panel testing of 552-A, 552-B, Marlboro King has been requested. • Product testing is addressing this issue across brands. ATTACHMENTS : • Attachment I: Attribute Performance Relative to Marlboro King • Attachment II: Attribute Performance Relative to WINSTON King • Attachment III: Mean Attribute Ratings • Attachment IV: XG FFNM 85.SP Development Direction (ETA Ratings) • Attachment V: Analytical Data for XG FFNM 85 SP Prototypes (CT 552-A and B) • Attachment VI: Product Testing Schedule -7-