Tobacco Documents Online

Page 30 C) INNOVATION The mjor thrust of innovation and development that will take place in F'85 will be a complete revamping of our survey in- formation into an integrated whole which will see us through the 80's. The C.M.A. and Monthly Monitor have served us well, but they no longer suit the current dynamics of the market nor take advan- tage of today's level of informtion processing technology (in£ormation processing covers the areas of data collection through to computer reporting and analysis). Project Brand ID and various other studies leading up to that project, plus numerous opinions in and out of Research have identi£ied the draw backs of our current survey tools. An overall approach to overcome these flaws, that is not cost prohibitive, will be tackled and resolved in F'85. This undertaking is a major one and involves the formulation of a critical path which will aid us in meetinE our commitment. The main focus of this restructuring is titled Project Survey, which basically calls for a monthly telephone survey O r~D Cl- OD "--4 CD BatCo document for PFSFC 1 March 1999

Page 51 supplemented by an annual tandem in-home interview. The monthly survey will replace the C.M.A. with a shortened' computer assisted questionnaire concentrating on switching. Image data will be collected in the annual in-home survey which will also probe on issues which require, or are better delt with, using a visual stimulus. Tracking cur- rently being done in the C.M.A. over the telephone is un- responsive in providing proper measures o£ new brand or ad performance, and will be removed and replaced with a sore appropriate methodology. In summary then, the focus of the Research Group for F'83 is on three major projects: Survey, Image, and Track. There are numerous secondary projects to these three but all have implications on one another. They are currently scheduled and managed independently. In April, a master plan will be developed to indicate the critical points and implications each has on the other. The following is a description o£ each of these projects. 0 r~O Or, oo BatCo document for PFSFC 1 March 1999

Page 32 I) Project Survey The C.M.A. is currently conducted using a quota sample of urban smokers collected monthly using a telephone method- ology and typed questionnaire with numerous skip patterns and VQB' s. Various inadequacies exist because of such a methodology: - a brand identification problem due to the inability through the restricted use of words to identify the proper brand; - length of questionnaire as it is weighted down by tracking, image statements and various other questions is driving down the completion rate and quality of response and at the same time driving up cost; - because of time restrictions and the type of scaling used, image data is currently inadequate; and - lack of random sampling methodology imposes restrictions on the fore of analysis that can take place. Project Survey will initially be a feasibility study on the use of a computer assisted questionnaire administered to a urban sample which probes primarily on demographics, usual brand, cZ~ r,,J O'- CO o r,,J BatCo document for PFSFC 1 March 1999

Page 33 and previcus brand. A computer assisted questionnaire can make better use of deep probes for proper brand identification by allowing detailed question blocks for every brand on the market. Built in editing can occur to help reduce coding and secondary editing. Flow of questionnaires will be flawless and less time- consuming thus reducing length of questionnaire, thus reducing cost and increasing quality of response. As a means to secure the right supplier for this proposed design, we will be considering a proposal to be submitted by Adaom before the end of F'S/ as well as explore what other suppliers have to offer. O ~o co O ? BatCo document for PFSFC 1 March 1999

Page 34 2) Pro)ect Im~e Project Image is an umbrella project. Its ultimate goal is to develop and implement a comprehensive, quantitative instrument to measure and interpret consumers' attitudes and beliefs towards cigarettes. Phase I of this project was the development of a cons~ner smoking vocabulary. Constructs, dimensions, and their meanings, as employed by smokers, was eval- uated. Phase II of this project was a refinement of our cigarette typology, the relative positioning of brands on image dimensions, as well as further assess- ment of methodological considerations. Phase III of this project will be a large ad hoc research project designed to give relative positionings of relevant brands on image dimensions; in order to provide a comprehensive description of brand images in the Canadian cigarette market. A pilot stud)' on 40 brands and 10 dimensions will be undertaken. Areas such as: switching, smoking & health, price sensitivity and "ideal" needs and wants will be collected at the same time. The use of ~itivariate analysis will help uncover the complex relationships that exist and help in our understanding of the marketplace. 0 0 oo BatCo document for PFSFC 4 March 4999

Page 35 3) Pro)ect Track There is a need to set up and implement a system incorporating a set of benchmarks (criteria, standards) that can be utilized to track new cigarette brands both own and opposition. We presently track new brands with various tools: - retail audits; - C.M.A. - awareness, purchase, trial; and - focus groups. The system that will be set up may or may not differ fr¢m our present system. Present tools will be evaluated and possible additional/alternate tools will be looked into. Benchmarks will be set up utilizing a combination of our historical data [i.e. a~reness at time intervals of first generation, second generation, etc.) and possibly some external consultation. 'No money has been set aside for such consultation, however. 4) C.M.A. Reporting Currently, we are reporting C.M.A. consumer data based on crossed tabulation of key variables for market analysis. It is difficult for users to identify when trends occur. Also, our information is sometimes not very meaningful to most, in certain areas such O C~D CD F-O BatCo document for PFSFC 1 March 1999

Page 36 as imagery, tracking, etc. Hence, they usually rely on market research to hand out comprehensive summary. and analysis. In order to respond more adequately to our users, we must take full advantage of system flexibility. Thus, in F'83 we will undertake the necessary, steps to do so by validating, testing, and implementing modifications to our survey reporting system. The presentation of the information will be in response to various identified needs such as target group switching, imagery for non-users of a product, media target groups, etc. The project will be broken into phases that will see separate examinations of the different components of the information. Implementation of new approaches will most probably be staged depending on the above. 5) Cordcall Project Cordcall is to establish a means of measure of consumer data that does not generate a bias resulting from consumer memory. Currently under investigation is the implementation of a diary panel study of smokers to monitorat different times of the year, consumption levels of cigarettes, product defects, and purchasing behavior. Phase I currently under progress, which is a test-study 0 r,o C~ o BatCo document for PFSFC 1 March 1999

Page 37 of this methodology, should have results available by mid-March 1982. Upon positive assessment o£ the test, we would proceed with two waves of diaries. 6) Plus/Minus Project Plus/Minus delves into the areas o£ our market which we cannot predict accurately. The specific area = o£ interest is young smokers between the ages o£ IS and 19. Among the things that we do not have, besides quantitative incidence and branded information, are the causal factors that create these responses. In 1978, exploratory research was done on this issue. Subsequent to this study, switching data shows that the young are following their elders with regards to starting on lower tar products. As the youth stream from the baby boom diminishes and market potential matures, we should better understand what their smoking and quitting behav- iors are today. Have these attitudes and behaviors changed in the past three to four years? Are we respon- ding to their needs? How do they perceive current advertising trends? The purpose of project Plus/Minus is to update our portraits o£ starters and quitters, ~=~ cZD r-o C~D CZ3 BatCo document for PFSFC 1 March 1999

Page 38 explore starters' smoking history, attitudes and behaviors, explore causal factors leading to quitting and ultimately provide better predictors. A two-pronged approach will be taken on this proj- ect, one involving quantitative desk research on starters and quitters among the young as well as qualitative work on the youth, both smokers and non-smokers. 7) Smoking & Health We are currently gathering in£ormation twice a year on smokers' perceptions of smoking and health. The method o£ obtaining this in£o~tion has been con- stant since the early 7O's. Among the latest results, one can find little change over time on the issues. Horeso, our actual means of collecting the data nmy not reflect the concerns that exist today, such as social environment pressure, prices, quality of prod- uct, CO issue, etc. Project Image will explore this issue initially with the use of the consonance/disso- nance approach employed by B.A.T. The structure of our questions appear weak and could benefit from a redesign. The purpose of doing a smoking and health review is to better define and bring up-to-date 0 C~ co --4 BatCo document for PFSFC 1 March 1999

Page 39 our methodology (i.e. more in tune with reality and pertinent to the concerns of mokers). Possible ap- proaches to resolving these issues are: 1] qualitative work on smokers around a point time when certain issues are more relevant; 2] evaluate relative strengr~hs of different itess of concern :in. smokers' mind; and 3) redesign the VQB adopted for the 80's for smokers' concern. 8] Regional Information A project to be undertaken byMarketing Research in response to informtion needs involves providing better regional information; a response to the establishment of regional marketing analysis. The current information forwarded to regions is not meeting their market analysis needs. The strategy for establishing what regional infor- mation needs are, involves having the four RIA's participate on an orientation program in Research the first two weeks of March. They will be putting together a positionpaper on their in£ormationneeds. This paper will be presented r-o cr~ co BatCo document for PFSFC 1 March 1999

Page 40 on the last day of the orientation after having been exposed to all the data that exists in Research. Marketing Research will then use this as a brief from which work can proceed to better satisfying regional information needs. 9) Enhancement of Current Ansl)rtical Techniques Using Existing Data It is our concern in Research that the existing data we have has not been milked to the extent that it can be in providing an interpretation of the market. With the computing capacity that we currently have and the advance- ment in analytical techniques currently being practiced by outside suppliers, there is certainly room for improvement. In the types of analysis performed by Marketing Research, we have set an objective for ourselves that over the course of the year and prior to the annual planning update, we will incorporate new forms of analytical techniques currently in use today to better explain the market. One of these techniques will be ~Itivariate analysis which Peter Illich has assm~d responsibility for. 0 cr~ Oo o BatCo document for PFSFC 1 March 1999

Page 41 I0) Integration of Information As a result of today's technology, the ability to integrate information is at our door step. As it stands now, we have various forms of survey in- formation on line being used and analyzed in iso- lation. Data such as that collected in the C.M.A., the Monthly Monitor, massive switching studies, PMB, sales, advertising expenditures, etc., should be readily available for cross analysis. A com- mittee project headed by Carole Brohmenwill work towards the integration of this information for purposes of analysis. The initial stage of this project requires a position paper on how information can be integrated either internally or externally. Part of this project will be the elimination of redundant information. I: i+ CZ~ r~Y fy~ (30 C~D <24 BatCo document for PFSFC 1 March 1999

Page 4Z 11) Price Sensitivity Historically, the Canactian cigarette market has not segmented on price. Recently, Loblaws in Ontario introduced a no-~meproduct in its stores with ap- parent success. Is the market ripe for such a concept? With the governments raising taxes on to- bacco products continuously an4 a worsening economic climate, is there a measurable level of price sensi- tivity in the Canadian market? What has caused the sudden plateauing in the decline of the fine cut market? Why is incidence in RYO and tailor-made combined up? Our current measurements do not permit us to record accurately shifts in the market as a result of price if there are any takinz place. The day may soon come, if it hasn't already, when cigarette demand will be significantly sensitive to price an4 part of a consumers' purchasing decision will be price. Project Barker will explore the concept for low-priced cigarettes for ITL, while Project Price Sensitivity, the one discussed here, will hopefully give us insight into shifts in attitude and behavior as a result of price and allow us to monitor its magnitude on an ongoing basis. ....a. 0 P,m C~D -../ 0 tw4 C7~ BatCo document for PFSFC 1 March 1999

Page 43 12) Low-Level Incidence Currently, we are faced with increasingly high research costs. These costs increase exponentially when the product's incidence level drops. We must find research techniques to give us answers without prohibitive costs with regards to product incidence. We, therefore, will undertake to find the most appropriate research tech- niques for low incidence tobacco products. This project comes as a follow-up to Project Brand ID, in which it was found to be cost prohibitive or too impractical to implement the proposed methodologies. This project is not in response to consumer confusion towards new brand data, but rather on reliability in representativity of our survey method for quantitative approaches. IS) Pro~ ect Product This project has two overall objectives: I) assessment of our current product-testing methodology with the ultimate goal being an improvement in the instr~nent in terms of time, cost, and information supplied; and 0 r,o 0", Oo o 04 -.--,l BatCo document for PFSFC 1 March 1999

Page 44 z) integration of smoker taste and sensory perceptions to better develop our understanding of attitudes, images, and market segmentation. A meeting has been scheduled in late F'82 with Product & Packaging and R & D along with Martin Oldman to discuss issues in this area. Hopefully, a result of this session will be to set new directions in the area of product mea- surement and testing. 14) ,Expanded Industry Exchange The expanded industry exchange is an exchange of sales data by the manufacturers which measure company/brand performance in a previously defined submarket. The submarkets have been isolated with the submission of "clusters" (groups of whole- sales). As expanded sales data becomes available, regions will be approaching Research for direction on how to use and analyze this information. It is our role to understand the co~position of the information and how it can be used in analysis. 0 C~ 0 BatCo document for PFSFC 1 March 1999

Page 4S 15) Price Audits As the federal and provincial governments continue to increase taxes on cigarettes, it has become essential that we gain a better insight into price variance in Canada. Presently, the instruments ~ use to measure retail prices of cigarettes do not optimally reflect price differences and their impact on sales. The initial step in developing a test to measure price more effectively will be an analysis of all existing vehicles. At the completion of this stage, proposals wCll be made to improve our means of tracking and measuring retail cigarette prices. 16) Retail Audits There is a need to improve our ability in measuring brand performance at retail. Retail audits will be studied this year. At the present time, our sophistication is to a level of conducting mini audits in selected markets for specific brand tracking. Co -.o BatCo document for PFSFC 1 March 1999

Page 46 The reliability of these samples and their representativity is questionable. Aposition paper on the subject will be put forewardby Research recommending that such information should be purchased by a supplier. However, before designing any approach, a clearer definition of information needs is required. Are we concerned with only a measure of gross inventories to adjust sales or should we be looking into the future at store segmentation, retail n~nagement and retail inventory control? 17) Trade Inventory Armlysis Trade inventory analysis will be an in-house investigation of tobacco and associated products inventories. The objec- tive is to measure the influences of economic conditions on inventory levels at various points in time, while improving our power to predict trade reactions to changing economic conditions. This project will be done in conjunction with retail audits. 18) Sales Force Activities Sales force activities interact with virtually every other phase of the marketing system. Each year, millions of dollars are spend in support of field activities, either O rk5 c;-, co CD r-.. (ZZ~ BatCo document for PFSFC 1 March 1999

Page 47 directly, or indirectly. Yet, our only measurement of the execution of the field activity (or strategy) depends largely upon intuition. The objective of this project is to develop a statistical measurement which will assess the execution of field activ- ities. 19) Distribution Network In 1950, ITL serviced 3,700 direct accounts. By 1981, this figure had dropped to 685. The total n~nber of retail outlets selling tobacco in 1981 was 43,746, although if present trends continue, we estimate that only 21,000 retailers will sell to- bacco by the end of the 80's. This project will explore the changes which have been taking place at the wholesale and retail level, i.e. why are distributors closing? where? are there new "kinds" of distributors opening? what are the ramifications of the changing retail environment? will we need more salesmen? fewer? etc. While there are few distributors, these fewer are controlling greater amounts of merchandise, and are in a position to ulti- mtely exert a tremendous influence on what mam~acturers might 0 crx o--9 c-D J BatCo document for PFSFC 1 March 1999

Page 48 do in the marketplace. As a result, we must be aware of any. "power shift" from the manufacturer to the distributor in order to develop s~trategfes to preserve the present balance between manufacturer and distributor. 20) Project Pre Test At the present time, much of the advertising research which is conducted utilizes a qualitative methodology. Focus groups are employed to assess different creatives and to evaluate new concepts. This does provide general direction, and will remain useful in discriminating amongst ntmerous concepts and creatives. However, this methodology is inadequate in terms of deter- mining the impact of an advertisement, particularly when compared to industry norms. Further, qualitative research will never be able to fuzx~ish answers to various questions which exist. Among these are: - the use of black borders on duMaurier advertising; - red on red for dt$4aurier Light; - is the low awareness o£ the Peter Jackson Extra Light campaign a result of the creative?; - can black and white deliver the same message as color?; 0 c._-/) BatCo document for PFSFC 1 March 1999

Page 49 -what is more effective in terms of cost: half a page, full page, or dou51e page adverti'sement~?; - average view~ t~aue ~or an advertisement; - average viewing time for each specific ad element; - the percentage of respondancs who read headline and body copy', both partially and fully; - etc. There is snAmerican firm, TelcomResearch Inc., a leader in the field of eye-movement track.g, which is currently expanding in the Canadian market. They suEzest that the), can provide answers to all these areas of cnncern. As an initial step, Research intends to com,fssion a pilot study cn existing Peter Jackson Extra Light and du~urier Light creatives. The results will then be compared to various norms which, although scant>-, have already been collected in the Canadian Industry, 21) .Pr, o~ect Incidence There is growing need for a continuous, accurate measure of incidence of smoking. Historically, we have relied on two annual readings frum the spring and fall Monitors. ~bwever, various factors in the market, particularly price and the 0 r,o o-, co -..j BatCo document for PFSFC 1 March 1999

Page 50 possible volatility of ~ market to price, necessitate a more frequent measure. An agreement has been es~blished with Benson & Hedges and Canadian Facts which will allow us access to an in- cidence measure for ii months of the year. The £inal methodological design of Project Survey and Project Image mat cause a change in this arrangement however, and the issue o£ how best to measure incidence in the cigarette market in terms of accuracy and cost will be closely exa~ed. Z2) Pro~ect Librar~ The development of a research library was something that was to be completed in F'82. H~wever, due to the un- planned for complexity and size of the project, we do not expect this to be completed until sometime in F'83. As a review, the library is to provide a continuity of information that was previously lacking in Research. Almost all research projects will involve, in their early stages, a review of past research into the area in question. A research library will greatly £acili~ate this type of research. 0 CO 0 BatCo document for PFSFC 1 March 1999