Tobacco Documents Online

. , AJA1 Form 6828 - 10/8D CONFIDENTIAL State of the Brand ort ~ ..~.:_ ~.~.~_ M: ~.~e FROM: Hs. D. S. Burrows -Copies to: - ------- s.- E. N. Monahaa M. _ D. Harris . D. S. Jo nson -- - ~~' Burrows T r. W. L. Mullenix -Ms _-LBfirei.ninger_ _ Mr. C. A. Martin IrIDIC-w===~ _ - - _ ~ _ __ - -.----- - -~~--•--- --------- ---------- --- ---- -, _._ _._.. ... _ ~_ -- _ _ ~ " -- - -- - -..,, _ ,__-F_ - --- - - -- ~m,~:-~•m~--~~ ~.._~..~ .. - - - - _~- --- w . ~

SALEM STATE OF THE BRAND Management Summary StJIIIKARY OF FINDINGS INDUSTRY/RJR - - --- - MDD's ability to analyze Industry/RJR performance between September, 1982 and February, 1983 was hindered by gross distortions in volume and share-of-market data. These distortions were caused by a series of price increases, trade and consumer promotions, and trade inventory adjustments. Since these distortions appear to have been either.~ substantially reduced or eliminated du,-ing the early weeks of March, MDD believes that it is now possible to provide a reasonably accurate analysis of current performance. Analysis of Industry and RJR performance for the First Quarter, 1983, shows the following: • Industry consumptio-uhas dropr,ed 3-6X between September, 1982 and March, 1983, due to the sharp (+22x) retail price increases surrounding the 84 FET increase. Most of the impact has been seen in 85mm male-oriented brands, -- -- ----.- . _. ~_, , ~._ while the female-skewed~lOC?mm/Stqlish/Concerned styles were less affected. Generic cigarette sales have soared. • RJR has experienced a larger volume loss,--in the neighborhood of 8-12X, even after adjustment for inventory drawdowns following year-end promotional activities. °" -- • RJR's share trend has deciinect :8 to ger'haps over 1.5 share points since September, 1982. - ~ = -= _ ---- =-- _ - - • CAMEL, WINSTON and SALEM, RJR4sfol~, larger b:.ands, have accounted for most of RJR's share and volume'losses. • As across the Industry,-RJR's losses"were coiicentrated in its 85mm styles, while its 100mm styles generally performed almost on trend, with the possible exception of WINSTON and SALEM full flavor 100's. • RJR experienced more quitting than the balance of the industry which accounted for'much of our losses. In addition, we experienced above average switching losses to generics. Disproportionate switching and quitting are attributed to specific smoker groups having different price elasticities. Very little of RJR's loss appears to be going to branded competitien.

SALEM \JOLUME/SHARE • SALEM's First Quarter shipments totalled 146 billion, 21% heIow year-ago and 8% below Plan. About 707% of this variance from Plan was due to inventory adjustments not anticipated in the Plan, leaving about a 2% shortfall attributable to 1983 in-market performance. • The impact of the FET on SALEM volume is estimated to be between 6% (based on total First Quarter versus projected pre-FET trend) and 14% (based on March shipments versus projected trend). Consumer measures, possible distortions in SALEM's March shipments, and SALEM's estimated price elasticity (based on its smoker profile) suggest that the actual impact on consumption may be in the lower end of this 6-14% range. • SALEM's First Quarter SOM (7.42%) was depressed by inventory adjustments. liowcvcr, SALEM's share trend is estimated to have dropped by about .4 share points below its early 1982 trend devel. . • Most of SALEM's losses are attributable to its,85mra styles, especially King and Lights 85, which is consistent with those styles' estimated price elasticities, the highest within the Brand. These products also were inferior to competition during this vulnerable period. A Forecasting Group analysis of March shipments for SALEM 100 suggested that its in-market volume might be as much as 11% below its pre-increase trend, a falloff twice as severe as the drop they found for total established RJR 100's. SALEM 100's profile does suggest that it msy be more price sensitive than most RJR 100's and there has been some softness in the style's regular smoker share. However, this -11% estimate may be inflated by distortions surrounding the March 100mm promotion. In fact, an analysis of SALEM 100's percent contribution to RJR 100's suggests there has been no significant difference in the impact on SALEM 100 and other established RJR 100's. • Price sensitivity among SALEM styles shifted significantly during 1982, c$ue to profile changes following the Brand's repositioning, but the overall Brand sensitivity increased only slightly. • Preliminary source of trial/purchase data suggest that SALEM may be contributing some 20X of BRIGHT purchasers (the level anticipated in the Plan) but that these smokers may account for up to 40% of BRIGHT volume, which would impact SALEM more heavily than hnticipated. • 1982 switching data indicate that SALEM may be exceptionally vulnerable to generics, contributing up to 132% of its fair share. Given generics' recent growth, this could account for some .3 points of share loss for SALEM, most of its estimated .4 total share point loss. - - --- _ There is no evidence that SALEM's losses are going to branded competition other than BRIGHT. SALEM's share of Coolness increased to 43. 1% in April, while Kool and Newport both lost segment share.

• SALEM's performance trend by Sales Area has not shown significant changes. • Although there have been no significant changes in regular/occasional SALEM usage, regular smoker shares for SALEM 100, Lights 85, Lights 100, and Slim Lights have eroded during the First Quarter of 1983. This may reflect reduced Brand loyalty during this period of heavy price promotions and trial/purchase of BRIGHT. • SALEM's performance versus Plan continued to show modest, but consistent improvement through April, as has been seen through most of 1983. ---a -z CONCLUSIONS/IMPLICATIONS • Price reactions by consumers appear to be the most important factor in SALEM performance, as-across most of the Industry. This effect includes quitting and switching to generics. Couponing may be an appropriate defensive measure for styles whose price sensitivity derives primarily from older smokers (Lights 100, Ultra 85/100, FF 100, Slim Lights), but it is unlikely to be effective among the younger smokers who create price sensitivity for King and Lights 85. While price promo[ions may contribute short-term defense against generics, Marlboro's strong performance suggests that clear, consistent reinforcement of imagery may be the best defense amang SALEM's younger adult target. • BRIGHT may have more impact on SALEM than was originally anticipated. However, longer observation is needed to determine whether this is a short-term effect of trial/promotion or not. _.~ - • Continued emphasis on achieving maiintaining product super~oiity should reduce SALEM's vulnerability. • SALEM's reposiicioning has been e€fect ve in increasing the Brand's appeal to younger adult smokers. However, since these smokers currently have relatively low rate per day, Brand loyalty will be required to achieve the full long-term volume payoff from this strategy - major short-term volume gains are not likely. Success in repositioning may, in fact, be signalled by modest volume-losses as the Brand's profile changes from older to younger. - I 0 =c ~ ~ m -ca w ~

DETAYLED FINDINGS INDUSTRY OVERVIEW Actual industry volume during the First Quarter of 1983 was 7.9% lower than in First Quarter, 1982 (after adjusting for trading day differences). This decline resulted from: Wholesaler inventory ad3ustments which trace to heavy-buy-ins for ~ priceiFET _increases plus ouarterlyJyear-end nompany promationR. ThPqP distortions-obscured true market performance from the summer of 1982 through February, 1983. Decreased~ consumption as-a-result of steeply r s ng retaii cigarette - prices; precipitated by the 81 per pack FET increase. Retail prices inc=ease~proximately 22% over the last 12 months (from about 694 to 84.54 per pack) and $re estimated to have reduced total consumption by 3-6%. If industry volume remains steady during the balance of 1983, the total year will be down y-42 versus the pre-increase period in 1982. 11,923 M!~ -4.UZ 1i,iS0 V

Studies by the National Bureau of Economic Research predicted that certain patterns would occur in response to rising cigarette patterns: I a Decreased consumption would result from smokers quitting (or failing to start) rather than from -reductions in rate per day. Recent data tend to confirm this pattern, with smoking incidence reaching all-time lows in mid-1982 and First Quarter, 1983 while rate per day remained essentially stable. (Appendix A). Males, especially younger adult (18-24) and older (35+) males, would react more strongly to price increases than other smokers, while femalas (especia y 25-l=females) would be much less likely to react. This -implfes that those catEgories/companies/brands/styles whose business depends mnst heavily on male consumption are most vulnerable to price and,'thus, woald show the largest volume/share losses during late 1982 and early 2983. Market results for First Quarter, 1983 are generally _ consistent with this, as discussed in various sections below. : SEGMENTS/CA1'EGORIES - - = _ ----~--~-• . - = -= ---~. ~~. - - Category performance across the Industry was consstent with the pat~era predicted by the NBER. Male-dominated style groups (85mm and Full Flavor) showed the most severe volume and share losses, while 100+mm and Ultra Low Tar styles, which skew female, Were least affected. Menthol styles showed more severe actua7:"~volume-loss for the First Quarter than non-menthol, but this may reflect the major inventory adjustments suffered by both SALEM and Kool, r3ther than a difference in price sensitivity. (Appendix B) - - - _ -• - - -- -- _=_° -__ = - -- Segment performmncewas-mofe~ject to inventory-distortions (which tend to occur for brands/companies rather than styles) but clearly underscore the st,7ength of the highly female Stylish segment. The "M.iscellaneous" segment doubled its volume and soared 1.85 share points due to tremendous gains by generics, which are benefitting from a 15% price advantage over branded cigarettes. (Generic switching as a price reaction was not considered in the hBER analyses.) (Data in Appendix B) When compared to its trend prior to 8eptember, 1982 (when retail prices began to climb sharply), RJR volume has declined some 8-12% - about twice the industry's rate of loss. RJR volume for the First Quarter was down (-3.3 billion) versusJPlan and down (-9.3 billion) versus the trend projected by the MDD Forecasting Group based on the pre-increase portion of 1982. . . Consequently, RJR has shown rapid s are losses since Novemberr 1982, f~rom about 33.1% to 32.3% SOM. After adjusting for introductory pipelining on = = --_= BRIG#iT, R3R's current s Zl:ing rate is estimated to be 32.1% - down 1.0 share points from the pre-November level. RJR VOLtJME/SHARE ----'- - 7~i-- - cs -- co: - - cr+

t 1 The 1983 Operating Plan assumed an RJR inventory adjustment of about 5.0 billion units in 1983. The actual inventory build-up in late 1982 was about 7.0 billion. Thus, about 2 billion of the First Ouarter 3.3 billion variance versus Plan can be attributed to inventory adjustments. Also, this 2 billion units which was loaded into 1982 accounted for about .7 share points of the Company's year 1982 SOM, which explains the difference between RJR's estimated 32.9% - 33.1% selling rate in 1982 and its reported MSA share of 33.55%. RJR's disproportionate losses may reflect the Company's franchise, which is older, more blue collar, and more male than the Industry average and, therefore, possibly more price sensitive. It is also possible that RJR's emphasis on price promotions has tended to attract smokers with above-average price sensitivity, increasing the Company's overall price vulnerability. _ _ _ - - =-- _ -- --~ __ _ In general, RJR's losses have been patterne as pre cted by the NBER, with the male-oriented 85mm styles accounting for 86% of the total loss. Most of the Company's volume drop was attributable to CAMEL, WINSTON, and SALEM (which appeared down 14% on average-during March) while NOW, MORE, and VANTAGE (which skew older/female/upscale) have recently performed more in line with the 1oss. - ~- Industry's average 4% l979 DATii 7}i21XJmi f*3tIL 193 --TOTA. R. 3. REYtCLD6 tONX,1M rREan 1980 1981 1982 1983 61t1t73T CCAMfiL FILTtit tl9' 31.31 BAisTAC.E 5pf»-Offr

SALEM PERFORMANCE Vo I ,ime SALEM shipments duri~g""the First Quarter of 1983 totalled 9.8 billion, 21X less than the same period in 1982 (af ter adj ustment for trading day differences) and 8% below the level projected in the 1983 Operating Plan. Most of this variance versus Plan was clearly attributable to inventory adjustments following the Corporate end-of-year promotional efforts, to which SALEM contributed roughly .6 billion units (Forecasting Group estimate) be ond the amount anticipated in the 1983 Plan. This accounts for more than 70% of the Brand's total variance versus Plan. SALEM VOLUME (MM) •1ST_QUARTER, 1983 1983 Operating Plan Shipnents Actual Variance hm 10; 639 - - 9; 822 - 837 Actual Year End Load Load Anticipated in Plan Excess Load/Payback Net Variance Not Attributable to Load/Payback 2,000 Est. 1,400 Est. 600 Est. - 237 MM - 2.2% _s-.--w- ; -L 3y projecting SALEFi's v31ume trend during March-September, 1982 through thes First Quarter, of 1983, the Forecasting Group estimated that, if all market factors had remained the same, SALEM's First Quarter conseumption would have totalled 12.6 billion. Thus, SALEM's actual shipments (9 'liillio ), adjusted for payback (+2.0 billion), were still .8 billion or ab t 667%- low than the projected cc+nsumption trend established prior to the p riod o `retail price increases. If this 6% difference represents the pact of price on SALEM consumption, then SALEM's price sensitivity would appear roughly similar to the Industry average. This would be in line with the price elasticity estimated for SALEli based on the Brand's age/sex profile and the NBER studies (see Appendix D). However, by comparing this trend level to shipments during the first three weeks in March (the first span of time in 1983 when overall market distortions seem to have subsided), the apparent impact on SALEM consumption could be estimated as high as 14%. v -- c~ ~ ~

ESTIMATES OF REDUCTION IN SALEM CONSUMPTION - ~ESTIMATED----- -- - - ---- O P IO - - - - --- - _TREND_-__~....~_.. "._..._ ° (BSED ON A(;TUAL NSUM N C REDUCTION PRE - 9/82)--SiiIPMENTS -(VARIANCE) (MM) (MM) (tL*1) TX) Total 1st. Qtr. 12,600 11,800* - 600 - 6.3% Avg. Wk., 3/7-3/25/83 1,007 866 - 141 -14.0% f , *Adjusted for payback. These estimates can be reconciled in two ways: (1) SALEM's.performance may have deteriorated during thot three months of 1983, i.e.,wMarch consumption may have been significantly lower than January or February: Iiowev'er; this seems unlikely - because: _ ~ - - - - -- ~_ > -- = • Consumer Alert data do not show any significant loss of regular/occasional SALEM smokers. (Appendix E) • No market share gains were found among likely SALEM competitors, except generics. (Appendix E) • Quitting in response to price is likely to be most severe immediately after the increase, and have less effect as some smokers gradually return. (2) SALEM shipments may have been somewhat depressed in early March by temporary factors, such as: f t • Early shipments for the Company's 100mm Promotion which, at that time, excluded SALEM 100. "Payback" following SALEM's $1.50/carton FSI, which may have led retailers/consumers to stock up. (-Appendix F) • Carton coupons on BRIGAT, dropped in February. (Appendix F) • Continuing inventory adjustments as wholesalers/retailers fine tuned stocks in line with reduced consumption. (This is more likely to affect large brands, like SALEM or WINSTON, than small brands.) i a_.. __ seasonality). (See Appendices G&iI) _ -= -=- - - :- : -- Although April volume reflected payback from the March/100m:a promotions, SALEM continued to show modest, but consistent, improvements versus previous month (which could be'seasonal), versus year-ago, and versus Plan (which considers

Relative volume performance for SALEM styles/flavor groups has been consistent with the relative price sensitivities predicted from their consumer profiles. The SALEM repositioning appears to have created shifts in style profiles (and therefore price sensitivity) during 1982, but did not greatly affect the overall sensitivity of the Brand Family (Appendix D). , Elasticity (2nd Half 1982) April YTD Var. Vs. Plan E Rank % Rank Lt. 85 :412 1 -14.8 1 King -.395 2 - 8.5 2 100 -.364 3 - 3.7 3 Ult. 85 -.364 4 - 3.3 5. TLt. 160----: _,.:...__->.r_.289-= _--__: -= 5 - 2.9 6 3 ' " 4 4 SSL _ -. 2 8 , 6 - .2 - -Ult. 100 -.221 7 - .9 7' -_ ---- - TOTAL SALEM -. 360 - 7.6 Most of the~Brand's volume shortfall has'_been att=ibutable to the 85mm styles, which the Forecasting Group estimated to have suffered as much as a 21% drop in consumption (by comparing March to the pre-increase 1982 trend). - By the same-method, SALEM 100 was found to be sh pping 13% below lts proj ected March trend, which represents nearly twice the impact suffered by,other established RJR 100mm styles (average 5.2% estimated consumption drop). It appears likely that SALEM 100 would be experiencing some in-market softness, since it seems to be one of the most price sensitive of RJR's 100mm styles (elasticity of -.364 compared to a -.333 average among RJR 100mm) and consumer measures suggest--some loss of regular SALEM 100 smokers (Appendix M) during the First Quarter. However;-the 11% estimate based on March may be inflated by shipment distortions related to the March•100mm promotion (from which SALEM 100 was excluded until the last minute). Also, an analysis of SALEM 100's percent contribution -to RJR-~06mEn volume indicates that SALEM 100's performance versus other established RJR 100's is essentially on the same trend which has been seen for two years, i.e., has not changed/worsened significantly during the price-impact period. -