UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Sales, advertising and distance Lockhart, David Culton 1969

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1969_A4_5 L62.pdf [ 5.9MB ]
Metadata
JSON: 831-1.0102341.json
JSON-LD: 831-1.0102341-ld.json
RDF/XML (Pretty): 831-1.0102341-rdf.xml
RDF/JSON: 831-1.0102341-rdf.json
Turtle: 831-1.0102341-turtle.txt
N-Triples: 831-1.0102341-rdf-ntriples.txt
Original Record: 831-1.0102341-source.json
Full Text
831-1.0102341-fulltext.txt
Citation
831-1.0102341.ris

Full Text

SALES, ADVERTISING AND DISTANCE — A REGRESSION ANALYSIS by DAVID CULTON LOCKHART B.Comm., University of B r i t i s h Columbia, 1968 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF BUSINESS ADMINISTRATION in the Department of COMMERCE AND BUSINESS ADMINISTRATION We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA APRIL, 1969. In presenting th is thes is in p a r t i a l fu l f i lment of the requ i re -ments for an advanced degree at the Un ivers i ty of B r i t i s h Columbia, I agree that the L ibrary s h a l l make it f ree ly ava i lab le for reference and Study. I further agree that permission for extensive copying of th is thesis for scho la r ly purposes may be granted by the Head of my Department or by his representat ive . I t i s understood that copying or pub l i ca t ion of th is thes is for f i n a n c i a l gain s h a l l not bo allowed without my wr i t ten permission. Department of Commerce and Business Admin is t ra t ion . The University of B r i t i s h Columbia Vancouver 8, Canada. Date: A p r i l 21st., 1969. ABSTRACT Most theories of promotion deal almost exclusively with behavioral parameters, and only s u p e r f i c i a l l y with the a l l -important action component. There have been few p r i o r studies examining the sales effects of advertising. The major purpose of t h i s thesis i s to analyze the association between a number of advertising variables and the weekly sales volume of an automobile dealership. A c o r o l l a r y objective i s to ascertain the role of average price and distance as related to sales. Relationships are tested by a multiple regression analysis on empirical data. Among the more important findings are: 1) Weekly Sales Dollars = -42.78 01 Dealer ' s Newspaper Lineage t-l"f-.47 Average Distance + 39.47 Average Price. N = 51 , R 2 = .40 , F«10.97 2) Weekly Sales Dollars—-34.31 •+-.01 Dealer's Newspaper Lineage t - l - h 39.51 Average Pri c e . N — 51 , R2=-.39 , F =--15.81 Both equations were s i g n i f i c a n t at the .001 l e v e l . Regression estimates indicate that the dealer's newspaper advertising, average price of the automobile and distance t r a v e l l e d by consumers are related to weekly sales. The study i s unable to conclude whether the r e t a i l e r ' s broadcasting expenditures or the manufacturer's l o c a l advertising outlays are s i g n i f i c a n t sales determinants. i i TABLE OF CONTENTS CHAPTER PAGE I. INTRODUCTION 1 Purpose of the Study 5 Units of Analysis 6 Advertising and the Automobile Dealer 8 Limitations of the Study 12 Advertising and Distance 16 Organization of the Report 17 I I . SURVEY OF THE LITERATURE 19 Introduction 19 Theoretical R e l a t i o n s h i p — A d v e r t i s i n g to Sales 19 Theoretical Relationship--Advertising to Distance 23 Sales and Advertising: A Review of the Literature 26 General 27 Department Store Sales 29 Novels 31 Lettuce 32 Cookware 33 Apples 34 Oranges 35 Dairy Products 37 i i i CHAPTER PAGE Coffee and Cleansers 38 Lamb 39 Cigarettes 40 Drugs 42 Automobiles 46 Summary and Conclusions 48 I I I . PRELIMINARIES TO REGRESSION ANALYSIS 50 Introduction 50 The Primary Data 51 Newspaper Advertising 51 Radio Advertising 52 Tele v i s i o n Advertising 53 Broadcast Dollars 53 Manufacturer Advertising 53 Total Advertising 54 Weather 54 Sales 55 Distance 55 IV. THE INTERRELATIONSHIP BETWEEN PROMOTION AND OTHER VARIABLES 57 Introduction 57 Advertising Performance and Other Factors .... 57 Consumer Product Preferences 57 Competition and Location 59 Seasonal Demand and Price Levels 60 i v CHAPTER PAGE Credit, Disposable Income and Intentions to Buy 61 The E f f i c i e n c y of Personal S e l l i n g .... 62 V. THE RESULTS 65 Introduction 65 The Correlation Matrix 65 General Comments 65 Pr e c i p i t a t i o n Variable 67 Sales and Advertising 68 General Comments 68 Sales and Total Advertising 74 Sales and Dealer Newspaper Advertising. 77 Sales and Distance 77 Sales and Price 82 Other Predictive Equations 83 Several Variables As Predictors of Sales Volume 86 Summary and Conclusions 86 Sales and Dealer Newspaper Lineage .... 87 Distance 87 Price 88 Other Variables 88 Further Research 88 BIBLIOGRAPHY 90 V CHAPTER PAGE APPENDIX A. Data on Manufacturer's Newspaper Advertising 99 APPENDIX B. Computer Program Used to Compute Median and Average Distances 102 v i LIST OF TABLES TABLE PAGE I . Cumulative Sales Volume, Gross and Net P r o f i t Resulting From $1000 Advertising Appropriation Increments 20 I I . Trend of Sales Expense When Advertising Increases as a Percentage of Sales 29 I I I . Cookware Units Per 1000 Female Heads of Households 34 IV. Sales Increases Attributed to Advertising .. 38 V. Relative Movement of Sales Associated With Di f f e r e n t Advertising P o l i c i e s 46 VI. Association Between Saturday Evening Post and Chevrolet Registrations Per 1000 People in 500 Areas 47 VII. Correlations—Automobile Sales and Advertising 48 VIII. The Correlation Matrix 66 IX. Percentage D i s t r i b u t i o n of Annual Automobile Sales 68 X. Automobile Dealer's Media Choice 70 XI. Results: Units Sold Correlated With 7 5 XII. Results: Sales Dollars Correlated With 76 XIII. Results: Distance and Advertising 80 XIV. Results: Sales and Distance 81 XV. Results: Sales Volume Regressed On 84 v i TABLE PAGE XVI. Results: Units Sold Regressed On 85 XVII. Regression cf Units Sold On Dealer and Manufacturer Advertising Variables 100 XVIII. Comparison of Manufacturer's Monthly Sales and Advertising Cycle 101 v i i i L I S T OF FIGURES FIGURE PAGE 1. The Advertising-Sales Relationship 21 2. The A d v e r t i s i n g - P r o f i t Relationship 22 3. Relationship of Travel Cost to Distance Travelled 23 4. Relationship of Travel Cost to Advertising Effectiveness 25 5. Net Regressions of Value on Advertising Expenditures 36 6. A Drug Products Advertising-Sales Relationship.. 43 7. Dealer's Media Mix 69 8. Dealer's Monthly Sales and Advertising Cycle ... 71 9. Relative Change In Sales Versus Relative Change i n Advertising 73 10. Dealer's Weekly Sales and Advertising Cycles ... 74 11. Advertising Expenditure and Out-of-town Sales 78 12. Unit Sales, Median and Average Distance 79 ix ACKNOWLEDGEMENT I wish to express ray sincere thanks to D r . J . D. Forbes for h i s computational assistance and guidance i n preparing t h i s paper. I wish also to extend thanks to the management of the dealership who w i l l i n g l y supplied the data on which t h i s study i s based. D . C . LOCKHART CHAPTER I INTRODUCTION The Importance of the Problem Recently an advertising agency executive published "a perfect measure of advertising's contribution to the firm: F i r s t , make a l i s t of a l l the working functions of the business (research and development, maintenance, accounting, sales e t c . ) . BUT DO NOT INCLUDE ADVERTISING. To each one of the l i s t e d functions a l l o c a t e the exact amount of sales or p r o f i t which can properly be credited to that a c t i v i t y . Add up the a l l o c a t i o n s . Deduct the sum of these all o c a t i o n s from the known t o t a l of sales or p r o f i t for the business. ^ What remains i s the contribution of advertising". Unfortunately, no one has yet devised an exhaustive technique for measuring the productivity of advertising. Wallace's "perfect" formula embodies a defeatest a t t i t u d e , prevalent i n many advertising c i r c l e s today: the feeling that the sales effects of advertising cannot be measured simply because the quest thus far has proved f r u i t l e s s . Irrespective of the many d i f f i c u l t i e s involved, there i s great pressure i n modern business for the development of inexpensive, p r a c t i c a l techniques to evaluate t o t a l advertising effectiveness. In the past, advertising researchers have estimated •"•James M. Wallace, "A Perfect Measure of Advertising's Contribution to Marketing," Journal of Marketing, 30:16, July, 1966. 2 what was easier to measure. As a r e s u l t , we now have reasonably sophisticated techniques to determine r e c a l l , impact and a host of other "intervening variables" which have no necessary 2 rel a t i o n s h i p to ultimate consumer purchasing behavior. "Such substitute a c t i v i t i e s may be completely j u s t i f i a b l e , but t h e i r vindication can come only from a successful assault on the 3 basic objective". Advertising research has simply side-stepped the main concern of p r a c t i c i n g advertisers—what are the sales e f f e c t s of advertising? In making the decision to advertise, the marketing manager i m p l i c i t y assumes that advertising w i l l stimulate sales. He i s not interested i n such effects as "noted", "seen-assoc-iated" or "read most", for "the ultimate measurement of 4 advertising performance i s buying action." The part-whole f a l l a c y i s no more evident than in advertising research. Just because three advertisements are judged e f f e c t i v e , does not imply that the sales results of the combination of these advertisements are e f f e c t i v e . I t would appear much easier to analyze the whole advertising campaign rather than attempting to ascertain the contribution of i n d i v i d u a l parts. See,for instance, K r i s t i a n S. Palda, "The Hypothesis of a Hierarchy of E f f e c t s : A P a r t i a l Evaluation," Journal of  Marketing Research, 3:13-24, February, 1966. 3 Harry V. Roberts, "The Measurement of Advertising Results," Journal of Business, 20:131, July, 1947. Daniel Starch, Measuring Advertising Readership and  Results (New York: McGraw H i l l Book Company, 1966), p. 178. Probably at no other time i n the h i s t o r y of advertising has the need for measurements of advertising's productivity been more acute than i t i s today. A recent study d e t a i l i n g increases i n the "average cost" of advertising i n Canada between 1961 and 1966 shows With t h i s trend toward increased costs for time and space, management i s demanding more objective evidence that adver-t i s i n g i s an economical and proven stimulator of sales. Management's concern i s c e r t a i n l y not unfounded, as the following table t e s t i f i e s : FROM 1961 — 1966 PERCENTAGE INCREASE Personal Savings 123 Personal Disposable Income 47 Consumer Expenditures 43 Advertising Expenditures 37 From a macro viewpoint, the rate of growth i n advertising expenditures has been slower than that of disposable income or consumer expenditures. Since the marketed value of goods and services has grown more dramatically than advertising 0. J . Firestone, The Economic Implications of  Advertising (Toronto: Methuen Company" 1967), p. 125. MEDIUM PERCENTAGE INCREASE IN AVERAGE COST Tele v i s i o n Radio Daily Newspaper Weekend Publications Consumer Magazines Outdoor: F u l l Showing 28 16 12 7 16 19 6 I b i d , p. 144. 4 expenditures, "the question arises whether in prosperous periods a r e l a t i v e l y lesser amount of s e l l i n g e f f o r t i s 7 required...". C l e a r l y , the prudent businessman w i l l want to examine his advertising-sales r e l a t i o n s h i p for whatever causal influence may be exhibited. I f the p o t e n t i a l consumer i s exposed to more than 1500 8 commercial messages d a i l y , each advertiser faces intense competition i n attempting to complete his communication process. In the l i g h t of growing advertising budgets, i t i s somewhat i r o n i c to f i n d that "few companies r e a l l y know whether they are spending too much or too l i t t l e for advertising, or what the e f f e c t s of an increase or decrease in t h e i r advertising appropriations might be". The problem becomes even more complex when one considers the gap between advertising theory and p r a c t i c e . "Today there i s l i t t l e evidence i n the l i t e r -ature to indicate whether business firms are i n fact doing what the theorists say they ought to be doing or conversely, whether the problems the theorists have attempted to solve by mathematical or other means are r e a l l y the ones the companies 10 are faced with". 7 I b i d , p. 8. Q °Harry D.Wolfe, James K.Brown and G.Clark Thompson, Measuring Advertising Results (Studies i n Business Policy No. 102, New York:National I n d u s t r i a l Conference Board,1962),p.2. 9 I b i d , p. 2. *°Donald C. Marchner, "Theory Versus Practice i n A l l o c a t i n g Advertising Money," Journal of Business, 40:286, July, 1967. 5 Purpose of the Study Every businessman should be anxious to determine, even crudely, the r e l a t i o n s h i p between the costs of his advertising and i t s sales benefits. Often there are clues within a company's past experience that are laying dormant, awaiting proper analysis. The major purpose of t h i s study i s to examine a series of sales and advertising volumes for whatever causal influence may be indicated. Measurement of past advertising e f f i c i e n c y i s desirable so as to provide the basis for greater productivity i n future advertising operations. I n a b i l i t y to determine the sales and p r o f i t contributions of parts of the marketing program does not hinder a good management from trying to secure some measure-ment of these parts. Without measurement, e f f e c t i v e advertising may be discarded because i t s contribution has not been recog-nized, or conversely, i n e f f e c t i v e advertising may be continued because someone wrongly assumes i t has been successful. In planning the "optimal" advertising program, i t i s necessary to assemble a l l information that completely explains variations i n sales over time. Unfortunately, such perfect information i s rarely available. I f one cannot ascertain the net effectiveness of advertising, at l e a s t he can determine the basic advertising-sales r e l a t i o n s h i p . Perfection i n ad-v e r t i s i n g research i s an unknown phenomenon, and a small step toward that state should prove rewarding.H With t h i s modest l^For a succinct discussion of the goals of advertising 6 goal i n mind, the writer has analyzed the sales and advertising volumes of a large Vancouver automobile dealership. Units of Analysis In d r i v i n g an automobile, one must continually readjust h i s speed to accomodate new information from the changing environment. In operating an automobile dealership, one must constantly reappraise his advertising expenditure i n the l i g h t of changing sales trends. Whereas the automobile i s a purely mechanistic system f a c i l i t a t i n g simple research, advertising productivity i s a Gestalt, requiring complex inves t i g a t i o n . T h e " e f f i c i e n c y of advertising" implies two d i s t i n c t aspects. "One i s the productivity of the advertising process, and t h i s refers to the effectiveness of advertising i n inducing pot e n t i a l sales. The other i s the productivity of the operations of t h e advertiser as a r e s u l t of, or at least i n part induced 12 b y t h e advertising process." Obviously, the f i r s t dimension implies a cost, the second, a benefit. In the marketplace, t h e cost of advertising for a l l automobile dealerships i s the same; what each receives from his advertising expenditure i s highly v a r i a b l e . This d i f f e r e n t i a l i n advertising effectiveness conceivably could b e so pronounced as to be transformed into a Research See, Edwin B. Parker, Stewart A. Smith and John Scott Davenport, "Advertising Theory and Measures of Perception," Journal of Advertising Research, 3:40, December, 1963. 12 O. J . Firestone, op_. c i t . , p. 64. $ 7 powerful competitive advantage. Using h i s t o r i c a l data, the costs of advertising can be e a s i l y calculated. C r i t e r i a for di v i d i n g the t o t a l promotional budget into i t s advertising, sales promotion and personal s e l l i n g categories were not required i n t h i s study, since the c o n t r o l l e r of the dealership had already performed t h i s function. However, i t i s another matter to diagnose the benefits of advertising, simply because there i s l i t t l e agreement as to what the precise advantages are. In a f i n a l analysis, i t would appear that the benefits of advertising are b a s i c a l l y the buying power of the advertising d o l l a r . Purchase of an auto-mobile i s a d e f i n i t e , ultimate measure of advertising perform-ance, and i n the present context, w i l l serve as the measure of advertising e f f e c t i v e n e s s . ^ This use of sales data as a c r i t e r i o n of effectiveness must be c l a r i f i e d . The theory of the firm assumes that business st r i v e s toward maximization of p r o f i t . Advertising i s only one marketing input designed to achieve t h i s l o f t y objective. In theory, then, the universal c r i t e r i o n on which to measure advertising e f f i c i e n c y i s contribution to p r o f i t . Using a marginal approach, the firm w i l l dispense advertising dollars u n t i l the l a s t d o l l a r spent equals the p r o f i t r e s u l t i n g from That i s , the firm's advertising objective i s assumed to be synonoraous with i t s marketing objective. For a contrary viewpoint see, Russel H. Colley (ed.), Defining Advertising  Goals For Measured Advertising Results (New York: Association of National Advertisers, 1961), 8 that expenditure. Rarely, however, does the firm have s u f f i c i e n t information to equate the marginal revenue product of i t s advertising d o l l a r s to i t s marginal c o s t . ^ By industry standards, the automobile dealership under consideration i s extremely p r o f i t a b l e ; i t s high sales volume has contributed to this success. While contribution to p r o f i t i s the i d e a l determinant of advertising productivity, i f we assume a d i r e c t c o r r e l a t i o n between p r o f i t and sales volume, then the l a t t e r may be used as the c r i t e r i o n of effectiveness. This assumption i s necessary because p r o f i t figures were not available for the study. Since the time period to be analyzed spanned only one year, weekly sales and advertising volumes were a r b i t r a r i l y chosen as the most suitable unit of study. When both series are subjected to multiple regression analysis, i t i s anticipated that underlying relationships w i l l be c l a r i f i e d . The s p e c i f i c units employed i n the regression equations are discussed further i n Chapter I I I . Advertising and the Automobile Dealer As a prerequisite to obtaining his franchise, the auto-mobile dealer contractually undertakes to implement an ambitious For a b r i e f summary of the "advertising production function" see, K r i s t i a n S. Palda, "Sales Effects of Advertising" Journal of Advertising Research, 4:14, September, 1964. 9 promotional p o l i c y . The importance of such promotion i s well demonstrated by the following clause extracted from the fran-chise agreement of a leading automobile manufacturer: The Dealer s h a l l promote vigorously and aggressively the sale of Products, using as f u l l y as i s p r a c t i c a l the Company's advertising and sales promotions and merchand-i s i n g material, and s h a l l develop energetically and s a t i s f a c t o r i l y the p o t e n t i a l i t y for such sales and obtain a reasonable share of the market i n the Dealer's T e r r i t o r y . Whether of not the Dealer s h a l l have vigorously and aggressively promoted such sales and s h a l l have obtained a reasonable share of market s h a l l be determined by r e f -erence to such c r i t e r i a as the Company may develop from time to time. In the case of Vehicles these c r i t e r i a may include, but s h a l l not be limited to, the relat i o n s h i p . of the Dealer's r e t a i l sales of Vehicles to users located i n the Dealer's T e r r i t o r y to (a) the t o t a l r e g i s t r a t i o n s of Vehicles i n the Dealer's T e r r i t o r y , (b) the f a i r and reasonable r e t a i l sales objectives of Vehicles established for the Dealer for the Dealer's T e r r i t o r y , and (c) the reg i s t r a t i o n s of automobiles (or trucks) of other manu-facturers selected by the Company and generally competitive with Vehicles i n price and product c h a r a c t e r i s t i c s i n the Dealer * s T e r r i t o r y . Perhaps the ro l e of advertising from the automobile dealer's perspective i s further i l l u s t r a t e d by considering the following description of automobile shopping behavior: ^ The f i r s t stage i s one of "preliminary exploration". I t may be described as a gradual evolution of a state of readiness to buy. I t culminates i n a d e f i n i t e decision, expressed i n verbal commitments, to enter the market. This exploration i s set i n motion almost imperceptibly, i n response both to external events and to i n t e r n a l psych-o l o g i c a l changes. A man arrives at this state, for example when his car has passed a certain aye (which he defines as "old" in comparison with the model year of cars owned by people he considers s i g n i f i c a n t ) . Or he may be impelled from the stage of unconsciousness s e n s i t i v i t y to one of active readiness by a change i n -°Leo Bogart, Stategy In Advertising (New York:Harcourt, Brace and World, Inc., 1967), pp. 77-78. 10 l i f e circumstances—a marriaye, a change of job or res-idence, the b i r t h of a new c h i l d , a death i n the f a m i l y — anything which a l t e r s his functional or status requirements. A buildup of tension takes place gradually over months for most people. During t h i s period, anything that goes wrong with the prospect's current car assumes an importance i t did not have before. I f the car burns o i l or gets less mileage to the gallon of gas, or i f minor parts have to be replaced, this now becomes a matter of family conver-sation. (Family discussion, i n fact, accompanies a l l of the subsequent stages in the process of decision-making). As the car owner finds himself talking more about cars, he also becomes more sensitive to advertising and e d i t o r i a l matter dealing with automobiles. He pays more attention to new models he sees on the road. Automobile shows and new model introductions i n t e n s i f y the process by creating fresh occasions for conversation and r e f l e c t i o n . As t h i s preliminary exploration goes on, the car buyer comes more and more to think of himself as being a c t i v e l y engaged i n the market. He may define his inte r e s t now i n terms of a target date for h i s purchase. This may be related to the introduction of the new models, to the end of winter, or to the anticipated lowering of prices at the end of the old model year. Often some decisive incident or event (a major repair job, or an occasion to celebrate) forces the prospect to begin his active shopping. Even before he reaches the point of decision to buy, he has become highly conscious of the d i f f e r e n t makes and models, and of the dealers i n hi s area of residence. Once he has made up his mind to buy a car, the customer starts to look into showroom windows and even to browse r through f l o o r '"displays. He starts studying the ads to compare features, and p r i c e s . He consults the "experts" of h i s acquaintance—like f i l l i n g s t ation attendants. He chats with friends about his own past experience. At the f i n a l stages he talks to dealers,about features, prices and trade-ins and, on the basis of comparison, closes h i s deal. In a majority of cases he makes up his mind within a matter of weeks after he had made the dec-i s i o n to buy. Throughout the c r u c i a l weeks, the prospect's feeling about the reputation of a make i s buttressed now by what he sees i n the advertising to which he has suddenly become s e n s i t i v e . But the ads he sees with new inte r e s t produce t h e i r e f f e c t by reactivating a l l the other adver-t i s i n g — n o t to mention a l l the news and rumor—that has reached him from that company over the years. This general description depicts the automobile purchase decision as e s s e n t i a l l y a two stage process: 11 1) The prospect f i r s t enters the market, and from the universe of accessible automobile dealerships, selects a sample of dealers from which to seek further information. 2) On the basis of offers received, the i n d i v i d u a l decides to accept or reject a p a r t i c u l a r proposition. Now, dealership advertising functions only as a means of stimulating potential buyers to contact a p a r t i c u l a r s e l l e r . Advertising entices prospects to walk into the automobile show-room, but personal s e l l i n g ensures that potential customers become buyers. Most dealers have a vague idea regarding the e f f i c i e n c y of t h e i r advertising. A popular measure i s a comparison of the number of automobiles sold with those models featured i n the dealer's recent newspaper advertisement. I t i s one thing to glance at sales figures and r e c a l l the advertising recently placed. I t i s quite a d i f f e r e n t matter to i s o l a t e the e f f e c t of one variable, advertising. An advertising manager must perform an herculean task in comparison to the sales manager. When the l a t t e r approaches the dealership president regarding probable benefits of h i r i n g an additional salesman, he says, "This man w i l l cost $ X i n salary, $ Y i n commission and $ Z i n fringe benefits. I am •••"According to Richard L. Nelson, The Selection of R e t a i l Locations (New York: F. W. Dodge Corporation, 1"95~8) , p. 5 3 , the automobile dealer generates 95% of his business through advertising. 11 1) The prospect f i r s t enters the market, and from the universe of accessible automobile dealerships, selects a sample of dealers from which to seek further information. 2 ) On the basis of offers received, the i n d i v i d u a l decides to accept or reject a p a r t i c u l a r proposition. Now, dealership advertising functions only as a means of stimulating potential buyers to contact a p a r t i c u l a r s e l l e r . Advertising entices prospects to walk into the automobile show-room, but personal s e l l i n g ensures that potential customers become buyers. Most dealers have a vague idea regarding the e f f i c i e n c y of t h e i r advertising. A popular measure i s a comparison of the number of automobiles sold with those models featured i n the dealer's recent newspaper advertisement. I t i s one thing to glance at sales figures and r e c a l l the advertising recently placed. I t i s quite a d i f f e r e n t matter to i s o l a t e the e f f e c t of one variable, advertising. An advertising manager must perform an herculean task i n comparison to the sales manager, when the l a t t e r approaches the dealership president regarding probable benefits of h i r i n g an additional salesman, he says, "This man w i l l cost $ X i n salary, $ Y i n commission and $ Z i n fringe benefits. I am •"•"According to Richard L. Nelson, The Selection of R e t a i l Locations (New York: F. W. Dodge Corporation, l"95"8) , p. 53, the automobile dealer generates 95% of his business through advertising. 12 projecting he w i l l return $2X+$3Y+ $ 4 Z i n increased sales". A l t e r n a t i v e l y , the advertising manager i s unable to place a s i m i l a r accurate monetary value on advertising's contribution to the sales objective. But, i f he can roughly demonstrate the sales response to advertising, the advertising manager i s i n a better position to obtain scarce promotional funds. From a p r a c t i c a l standpoint, the advertising manager would not expect today's advertisement to r e s u l t i n a sale two years hence. Yet sales effects may occur a day or a month a f t e r the advertising has been run. The cumulative ef f e c t s of advertising, whether they be of a residual, sleeper or boomerang nature, complicate delineation of the advertising-sales r e l a t i o n s h i p . In accordance with the above model of automobile shopping behavior, we s h a l l assume that the dealer's patronage advertising i s of the " d i r e c t action t y p e — a s e l l i n g document, complete enough in i t s e l f to induce immediate action by the consumer either i n the form of placing an order or that of seeking additional information before placing an o r d e r " . ' Given the automobile dealer's annual sales target, we are r e a l i s t i c a l l y assuming that the sales effects of advertising during any given year are more or less immediate. Limitations of the Study "Advertising effectiveness" i s a curious phrase. I t N e i l H. Borden, The Economic Effects of Advertising (Homewood, I l l i n o i s : Richard D. Irwin, Inc., 1952), p. 105. 13 i s at once both ex c i t i n g and somewhat embarrassing to advertiser, reasearcher and management. I t i s exciting because of i t s promise, but embarrassing because of i t s elusiveness. Why does a solution to the problem of advertising productivity remain so obscure? The answer probably stems from the fact that advertising researchers "cannot for the l i f e of them t e l l how to i s o l a t e the contribution of one cause to an e f f e c t of many causes (experimental design), how to learn what commun-i c a t i o n caused which sales (measure both), how to project from few to many (probability sampling), or how to do a l l t h i s within the organizational constraints of the advertising bus-iness."^- 9 More s p e c i f i c a l l y , the p r a c t i c a l d i f f i c u l t i e s to be encountered i n studying the sales effects of advertising can be reduced to four basic factors: u 1) The Marketing Mix Situation - Because the automobile dealer exercises many tools of marketing action, i t i s rea d i l y apparent that advertising, per se i s not the sole stimulator of sales. I t i s much easier to measure results of a t o t a l ^ • ^ i c h a e l H. Halbert, "A P r a c t i c a l and Proven Measure of Advertising Effectiveness," (Proceedings Sixth Annual Conference, New York: Advertising Research Foundation, 1960), p. 77. 19 Charles K . Raymond, "Are We Downhearted," Journal of  Advertising Research, 6:64, September, 1966. 20p o r a m o r e detailed discussion, see Paul E.Green, "Bayesian Decision Theory i n Advertising," Journal of Advertising  Research, 2:33, December,1962. Also, A l f r e d A.Kuehn, "How Advertising Performance Depends Upon Other Factors," Journal of Advertising Research, 2:2, March 1962. 14 marketing e f f o r t , a l l factors i n the marketing mix, than to i s o l a t e the effects of only one variable. "Even i f the results produced i n the past by each com-ponent of the mix could be discovered, information as to what a d i f f e r e n t mix would have produced would be lacking. And information about past results would not point with certainty to the results l i k e l y to be produced by a sim i l a r or i d e n t i c a l 21 mix i n the future". In other words, the f i r s t complication i s the i n t e r n a l dimension—the effectiveness of advertising decisions i s dependent upon other courses of action available to the firm. 2) The Environmental Situation - While the advertiser can control the mix si t u a t i o n , his a b i l i t y to manipulate important ecol o g i c a l factors a f f e c t i n g demand, i s far less c e r t a i n . Next to a house, the purchase of an automobile i s the largest expenditure most people make i n a l i f e t i m e . Such economic considerations as competitor's prices or the consumer's a b i l i t y to buy are la r g e l y beyond the control of any one dealership. In the same vein, the individual's psychological constitution, h i s buying habits and desires, are only minimally influenced by dealership advertising. 3) The Temporal Situation - I t i s generally recognized that a lag exists between the moment the advertising expend-i t u r e i s made and the time the sales results are obtained. 2lAlbert W. Frey, How Many Dollars For Advertising'" (New York: Ronald Press Company, 1955), p. 36. 15 Palda has summarized why such a lag may e x i s t : 1) Continued brand preference, though probably maintained by s a t i s f a c t i o n with the qual i t y of the product, may have i t s o r i g i n i n the action of a single long forgotten advertisement. 2) I t may take a series of advertisements to break through a threshold of buying resistance. 3) The poten t i a l customer, persuaded though he may be by the advertisement i s not immediately i n the market for the product. 4) A p a r t i c u l a r l y lengthy lag may re s u l t when ^2 a product can only be used from a certain age on. Advertising research cannot take place in a vacuum— to i s o l a t e the sales effects of advertising one must know the duration and amplitude of the lag. However, i n examining advertising's productivity for a given year, we are, i n e f f e c t , taking a s t a t i c view of the dynamic marketplace. This ab-st r a c t i o n from an obscure r e a l i t y i s necessary, but, of course, i s an inherent l i m i t a t i o n of the study. 4) The Correlative Situation - Most techniques currently employed to measure aspects of advertising effectiveness display methodological weaknesses—their r e l i a b i l i t y and v a l i d i t y should be subjected to close scrutiny. In analyzing the sales responses to advertising, these methodological d i f f i c u l t i e s are compounded. The advertiser should not interpret a high (low) corr e l a t i o n between sales and adver-t i s i n g as i n d i c a t i v e of high (low) advertising e f f i c i e n c y . For instance, i f the advertising appropriation i s budgeted on K r i s t i a n S. Palda, The Measurement of Cumulative  Advertising E f f e c t s (Englewood C l i f f s , New Jersey: Prentice H a l l , Inc., 1964), p. 9. 16 a percentage of sales basis, one would expect to f i n d a r e l a t i v e l y high, yet spurious c o r r e l a t i o n , between advertising and sales volume. Thus, one must supplement s t a t i s t i c a l analysis with informed judgment i n analyzing the findings of th i s report. Advertising and Distance While the study i s mainly concerned with the sales e f f e c t s of advertising, i t also focuses on the re l a t i o n s h i p between advertising and consumer s p a t i a l behavior. I f pro-motion r e a l l y exerts " p u l l i n g power", i s i t s potency s i g n i f -icant i n time and space? More p a r t i c u l a r l y , do customers t r a v e l greater distances to purchase an automobile when advertising e f f o r t i s auymented? C e n t r a l i t y i s a basic p r i n c i p l e of consumer b e h a v i o r — other things equal, people prefer to purchase products near t h e i r homes, thereby~minimizing t r a v e l costs. I t i s r e l a t i v e l y easy to document h i s t o r i c a l patterns of consumer s p a t i a l behavior; i t i s very d i f f i c u l t to analyze and inte r p r e t these s p a t i a l r e l a t i o n s h i p s . The attraction-repulsion continuum, depicted i n the following diagram, seems a f i r s t step toward better understanding the many interactions between sales, advertising and distance. Size and/or increasing Penetration increasing _ of Trading ^ Area A t t r a c t i o n Repulsion Advertising Distance Advertising i s viewed as a p o s i t i v e f o r c e — i t attempts to a t t r a c t customers over an increasing geographical area. As a counterforce, distance l i m i t s the extent to which pote n t i a l consumers are w i l l i n g to t r a v e l for the purpose of purchasing goods. At some point, there i s a trade-off between the con-sumer's p o s i t i v e expectations of purchase, i n s t i l l e d by advertising, and the negative expectations of nonpurchase, associated with distance. The size of the r e t a i l e r ' s trading area i s determined at t h i s trade-off point. Research i n marketing geography i s a r e l a t i v e l y new area of endeavour, and^ to the writers knowledge, there has been no attempt to r e l a t e sales, advertising and s p a t i a l behavior. This would seem to j u s t i f y the present study, a l b e i t of an exploratory nature. Organization of the Report With the preceding limitations i n mind, the paper now turns to a b r i e f description of the t h e o r e t i c a l relationships between advertising, sales and distance. The l a t t e r part of Chapter II i s devoted to a review of previous research findings on the advertising-sales r e l a t i o n s h i p — t h e paucity of relevant 18 studies w i l l soon become apparent. Pr i o r to regression analysis, i t i s necessary to select and obtain accurate data on a number of factors. The units in which these chosen variables are to be measured are discussed i n Chapter I I I . Much work remains to be done and future research must e x p l i c i t l y consider a variety of endogeneous and exogeneous variables. What d i r e c t i o n these variables may take is outlined, and where possible, implications for marketing strategy are suggested. The f i n a l chapter presents the more s i g n i f i c a n t regression r e s u l t s . The a b i l i t y of the independent parameters to explain sales fluctuations i s deduced u t i l i z i n g t h e o r e t i c a l and s t a t i s t i c a l c r i t e r i a . CHAPTER II INTRODUCTION The aim of this chapter i s t o review previous findings on the nature of the advertising-sales r e l a t i o n s h i p . However, a conceptual framework i s f i r s t developed so the reader may better understand the academic relevancy o f advertising to sales. Then, a theore t i c a l interpretation of consumer s p a t i a l behavior i s developed, with p a r t i c u l a r reference to the advertising-distance dichotomy. THEORETICAL RELATIONSHIP—ADVERTISING TO SALES At the outset, note that the models presented below are not profound, nor are they new. The representations are a n aid to analyzing the advertising expenditure problem but owing to t h e i r s i m p l i s t i c assumptions, the models are c e r t a i n l y not to be construed as the method of analysis. Underlying any study of the advertising-sales r e l a t i o n i s the concept of advertising e l a s t i c i t y : i f advertising expenditures are increased "X" percent what percentage increase i n sales w i l l r e s u l t . To calculate the advertising e l a s t i c i t y c o e f f i c i e n t , one would require such f i n i t e information as that presented i n Table I . In placing the advertising expenditure problem into Adapted from, Albert V 7 . Frey, How Many Dollars For  Advertising , (New York: Ronald Press Company, 1955), p. 33. 20 the i d e a l framework of contribution to p r o f i t , i t i s apparent that the optimal advertising appropriation i s $5000. Beyond that amount, additional sales are achieved at excessive cost and net p r o f i t declines r a p i d l y . TABLE I CUMULATIVE SALES VOLUME, GROSS AND NET PROFIT RESULTING FROM $1000 ADVERTISING APPROPRIATION INCREMENTS TOTAL ADVERTISING APPROPRIATION TOTAL $ SALES TOTAL SALES COSTS TOTAL GROSS PROFIT ($) TOTAL NET PROFIT ($) 0 1000 600 400 400 1000 5000 3000 2000 1000 2000 10000 6000 4000 2000 3000 14000 8400 5600 2600 4000 17333 10399 6934 2934 5000 20139 12113 8576 3566 6000 22689 13613 9076 3076 7000 24689 14813 9876 287 6 The data i n Table I i l l u s t r a t e an important p r i n c i p l e , namely that advertising i s subject to the law of diminishing returns: ...we can see that the f i r s t units of advertising have l i t t l e e f f e c t i n overcoming the i n e r t i a of the buying public; that successive units receive more and more res-ponse at a rapidly r i s i n g rate to a point where the r e l -ationship for the moment i s t h e o r e t i c a l l y l i n e a r , and that from there on additional amounts of advertising r e s u l t i n progressively smaller additions to sales. At some point to be determined through analysis of the ind i v i d u a l enterprise's break-even points, the additional advertising expenditure w i l l not produce i t s own worth. Below that point, reducing the advertising expenditure w i l l only r e s u l t i n s t i l l more diminished returns. 2 Sidney Hollander, "A Rationale for Advertising Expenditures," Harvard Business Review, 27:83, January, 1949. 21 Although he offers l i t t l e proof, Hollander suggests that "the return per d o l l a r of advertising i s more l i k e l y to be expressed as an S-shaped curve—up to a certain point, 3 increasing; past that point decreasing". This hypothesis i s shown graphically i n Figure 1. Advertising Dollars Sales i n Units FIGURE 1 THE ADVERTISING-SALES RELATIONSHIP Adding the cost element, the academic rela t i o n s h i p of advertising expenditures to sales and p r o f i t i s depicted i n Figure 2. With the exception of advertising, a l l factors affecting demand are held constant—thus i n Figure 2 varying sales levels are attributed to fluctuating advertising expenditures. This relationship between costs and benefits i s further explained as follows: The r i s i n g diagonal l i n e represents advertising cost and the curved l i n e labeled gross p r o f i t represents the difference between sales and a l l costs except advertising. 3 Ibid, p. 8 0 . T h e r e f o r e , the only c o s t not yet s u b t r a c t e d to o b t a i n gross p r o f i t , i s a d v e r t i s i n g c o s t . Hence, net p r o f i t i s the shaded area between gross p r o f i t and a d v e r t i s i n g c o s t . The net p r o f i t curve, when p l o t t e d s e p a r a t e l y on the h o r i z o n t a l a x i s , takes the i n v e r t e d bathtub shape as shown i n the diagram. -Saturation L e v e l FIGURE 2 THE ADVERTISING-PROFIT RELATIONSHIP 4 Gordon E f M i r a c l e , A Method of Measuring the  P r o d u c t i v i t y o f A d v e r t i s i n g (Ann Arbor, Michigan: U n i v e r s i t y M i c r o f i l m s , I n c . , 1962) , P. 1 3 0 . 23 Consequently, i f p r o f i t i s the c r i t e r i o n of e f f e c t -iveness, the optimal advertising expenditure i s the amount "OA", which i n turn, induces a sales l e v e l of "OS" and net p r o f i t "OB". THEORETICAL RELATIONSHIP—ADVERTISING TO DISTANCE A substantial body of l i t e r a t u r e has demonstrated the importance of distance in analysing trade movements. Most studies conclude that t r a v e l costs, v;hcther i n temporal or monetary units, exert a potent influence on s p a t i a l patterns of consumer behavior. The oft-cbserved inverse relationship between distance and the density of consumers highlights the importance of consumer t r a v e l costs. The t h e o r e t i c a l r e l a t i o n s h i p of t r a v e l cost to distance i s depicted i n Figure 3 . Travel Cost Distance FIGURE 3 RELATIONSHIP OF TRAVEL COST TO DISTANCE TRAVELLED 24 The curve "AA" begins somev.-hat above the o r i g i n — t h i s demonstrates that any shopping t r i p involves certain fixed costs, either i n expended time or monetary units. Beyond t h i s fixed amount, the t r i p expenditure r i s e s slowly with incremental increases in distance, u n t i l , at some point, i t r i s e s disproportionately. This point of i n f l e c t i o n varies with the consumer's perception of his needs, the distance he i s w i l l i n g to t r a v e l , and his opportunity cost. At least hypothetically, the impact of t r a v e l costs i s to c o n s t r i c t the r e t a i l e r ' s trading area. While the e f f e c t of distance in shaping the firm's demand curve i s doubtless an important concept, i t must not be overemphasized. The consumer's l i f e space, the areas i n which he works, shops and s o c i a l i z e s , has increased immensely i n recent years. Rising incomes, greater spending on shopping goods and the dynamics of merchandising tend to lessen the importance of distance. Moreover, the automobile i s the epitomy of the term "shopping good", that i s , one "for which the probable gain from making price and qua l i t y comparisons i s thoughtto be large r e l a t i v e to the time and e f f o r t needed to shop properly for the good". 5 Since t r a v e l costs are somewhat i n s i g n i f i c a n t , r e l a t i v e to the price of the automobile, consumers may v i s i t several showrooms. While the automobile dealer can never be certain that even the consumer next door E.Jerome McCarthy, Basic Marketing;A Managerial  Approach, (Homewood,Illinois: Richard D. Irwin, Inc.,1964), p. 398. 25 w i l l patronize his showroom, he might r e a l i s t i c a l l y assume that the nearer the household, the more l i k e l y the consumer's automobile shopping w i l l include a v i s i t to his location. To the consumer, there are l i m i t s of time and cost, but such boundaries should t h e o r e t i c a l l y become more ambiguous the greater the e f f e c t i v e l e v e l of advertising e f f o r t . Figure 4 depicts t h i s phenomenon: the curve increases slowly, but eventually, irregardless of advertising effectiveness, r i s e s r a p i d l y . Time Distance Expenditure Advertising Effectiveness FIGURE 4 A RELATIONSHIP OF TRAVEL COST TO ADVERTISING EFFECTIVENESS Advertising enables the consumer to gain some a p r i o r i knowledge of the p r o b a b i l i t y that an advertising dealership might s a t i s f y his automobile requirements. The automobile showroom i s t r u l y a generative location, that i s , "one to which the consumer i s d i r e c t l y attracted from his place of 26 residence; to shop there i s the p r i n c i p a l purpose of the consumer i n leaving his residence". I f such advertising i s e f f e c t i v e , the consumer exhibits a willingness to incur greater t r a v e l expenditures i n a n t i c i p a t i o n of greater need s a t i s f a c t i o n . Consequently, advertising may tend to increase the r a t i o of t r i p distances the consumer i s w i l l i n g to t r a v e l , and simultaneously, to expand the dealer's trading area. SALES AND ADVERTISING : A REVIEW OF THE LITERATURE In surveying the l i t e r a t u r e of the advertising-sales r e l a t i o n s h i p , one i s immediately impressed by the paucity of relevant studies. There are at least three reasons for t h i s s c a r c i t y of data:(1) research has t r a d i t i o n a l l y applied measures of "attention, i n t e r e s t and desire" as estimates of advertising's effectiveness, (2) the expense of measuring the advertising-sales relationship often cannot be j u s t i f i e d r e l a t i v e t o the cost of the advertising being evaluated, and (3) companies which have documented the r e l a t i o n s h i p are reluctant to disclose t h e i r findings. That research which has found i t s way t o the publisher i s p a r t i c u l a r l y i n t e r e s t i n g and relevant to the present study. To examine th i s research systematically, i t i s necessary to structure the following discussion with reference to the product investigated. Since 6Richard L. Nelson, The Selection of R e t a i l Locations , (Chicago: F. W. Dodge Corp., 1958), p. 45. 2 7 delineation of the advertising-sales relationship i s beset with many methodological d i f f i c u l t i e s , i t i s indeed revealing to note the many techniques employed by previous researchers. General A survey by Vidale and Wolfe affords a departure point i n reviewing previous empirical research and, at the same time broadens our understanding of advertising theory. Having analyzed the sales performance of many consumer products, these 7 t h e o r i s t s propounded a model based on three parameters. The general model i s of the form, ds — rA(t) M-S dt M-X. S and i s interpreted as follows: the increase i n the rate of sales, ds/dt, i s proportional to the i n t e n s i t y of the advertising e f f o r t , A, reaching the f r a c t i o n of potential consumers, M-S/M, minus the number of customers being lost/»S. The three para-meters merit further attention. 1} Sales Decay Constant ) - In the absence of advertising Vidale and Wolfe found that sales of many products decline as a constant percentage l o s t annually. For established products, the sales decay rate i s r e l a t i v e l y small, but grows larger for mature products facing highly competitive market conditions. For a more detailed discussion, see M. L. Vidale and H . B . Wolfe, "An Operations-Research Study of Sales Response to Advertising," Operations Research, 5:370-81, June, 1957. 28 2) Saturation Level (m) - Depending upon the product type and the effectiveness of media, there i s a p r a c t i c a l l i m i t to the sales volume that can be generated by a s p e c i f i c advertising campaign. The researchers found "a declining rate of increase i n sales per d o l l a r of advertising, suggesting existence of a l i m i t , beyond which additional advertising expenditures i n the same media w i l l be superfluous". 3) Response Constant (r) - The t h i r d dimension, sales generated per advertising d o l l a r , i s almost impossible to measure. Were the marketer able to estimate the response constant, his problems would be over'. The Vidale-Wolfe model i s an oversimplification, yet i t i s apparently based on empir-g i c a l data. I t i s u n r e a l i s t i c i n that i t completely omits a consideration of repeat purchasing phenomena. On a more p r a c t i c a l l e v e l , K o l l i n e r examined the marketing costs of 893 companies to determine i f greater advertising expenditures implied lower sales expense. He found that "the consistent downward trend i n sales expense continues when advertising increases as a percentage of sales 10 expense". K o l l i n e r s data, summarized in T a b l e l l shows that Ibid, p. 379. 9 •^ An application of the model i s presented i n William R. King, Quantitative Analysis For Marketing Management, (New York: McGraw H i l l Book Company, 1967),pp. 37 0-37 3. l°Sim A. K o l l i n e r , "New Evidence of Ad Values," I n d u s t r i a l  Marketing, 48:82, August, 1963. 29 for various size firms, a higher advertising to sales expense r a t i o decreases the t o t a l cost of sale s . TABLE II TREND OF SALES EXPENSE WHEN ADVERTISING INCREASES AS A PERCENTAGE OF SALES (SOURCE; SIM A. KOLLINER, "NEW EVIDENCE OF AD VALUES ," INDUSTRIAL MARKETING, 48:83, AUGUST, 1963.) ADVERTISING AS PERCENTAGE SALES EXPENSE—ADVERTISING OF SALES PLUS DIRECT SELLING COST Under 9.0 11.3 9 — 14.2 11.8 14.3 — 18 10.4 18.2 — 24.8 9.3 24.9 — 33.9 8.9 Over — 34.2 8.4 This inverse r e l a t i o n s h i p may be a completely spurious finding, for i t neglects the nature of each firm's s e l l i n g a c t i v i t y . In aggregating the data of many firms, K o l l i n e r ' s survey implies economies of scale. This finding i s i n sharp contrast to that of Weinberg, who, i n a s i m i l a r study concluded that sales of an i n d u s t r i a l product declined logarithmically as a function of a d v e r t i s i n g H o w e v e r , the two studies are not d i r e c t l y comparable, because they employ d i f f e r e n t variables. Department Store Sales - Cover et a_l examined the sales of women's clothes i n selected Chicago department stores between 1926 and 1931.* 2 The advertising series studied was units of See, Robert S. Weinberg, An A n a l y t i c a l Approach to  Advertising Expenditure Strategy, TNew York, Association of | National Advertisers, 1960). i 1 2John H. Cover, et a l . , "Department Store Sales and Advertising." journal of Business, 4:227-44, J u l y . 1931. 30 newspaper lineage; sales were viewed as percentage increases or decreases from a p a r t i c u l a r month to the corresponding month i n the subsequent year. Trend and seasonal factors were then measured and eliminated. Using graphic analysis, the research-ers found that newspaper advertising generally conforms to c y c l i c a l fluctuations i n department store sales. The fact that as sales decrease, advertising f a l l s o f f , i s a l o g i c a l f i n d i n g . To what extent department stores create demand, as opposed to following consumer desires, was an unsolved problem. In a l a t e r inquiry, Brown and Mancina hypothesized that the r e l a t i o n s h i p between the sales and advertising expenditures of 108 department stores could be expressed by a l i n e a r function of the following general type: Q—AV*' V*L VK* V * " O / \ A / A i A 3 •••A/) where: S refers to sales volume A i s the advertising constant X-^t X , X^, X n are s e l l i n g inputs K l ' K 2 ' K 3 ' K n a r e c o n s t a n t s ' Hence, advertising i s proportional to t o t a l s e l l i n g e f f o r t . While th i s l i n e a r function appears to imply that advertising should be a fixed percentage of sales, the authors emphasize that "a perfect relationship (between sales and advertising) should not be expected, for the theory implies that the e f f e c t of advertising i s continuous over an i n f i n i t e range, wheras i n r e a l i t y there i s a certain minimum amount that 13 can successfully be spent on advertising". Brown and l^George H.Brown and Frank A.Mancina, "A Note on the Relationship Between Sales and Advertising of Department Stores," Journal of Business, 13:8, January, 1940. 31 Mancina found that the r a t i o of advertising expense to t o t a l s e l l i n g expense varied between nine and f i f t e e n percent. Although t h e i r analysis of variance technique yielded only rough r e s u l t s , they concluded "that the relationship between sales and advertising i s describable by a function homogeneous i n the f i r s t degree".^ Although he offers no empirical data, Maranz states that one measure of the effectiveness of department store advertising "can be obtained by noting i t s impact on the weekly moving average sales curve". The curve i s calculated as follows: M1+ Tu2+W3 + Th4 + F5+36 = Weekly Moving Total One Tu2 + W3 +Th4+-F5+ S6-+-M8 = Weekly Moving Total Two where each l e t t e r represents d a i l y sales and each number represents the date. I f advertising does influence sales, the smoothed sales curve w i l l deviate from the normal curve, (that i s , sales without advertising). Novels Berreman analyzed the sales and advertising figures of 234 novels selected from publishers' l i s t s between 1933 and 1938. I n i t i a l l y , he noted an apparently high c o r r e l a t i o n between the t o t a l amount spent on advertising and the t o t a l -^Ibid, p. 1. l^Marcel Marantz, "Evaluating Department Store Advertising, Journal of Advertising Research, 7:16, March, 1967. sales of novels. Looking closer at the temporal dimension, Berreman found c o n f l i c t i n g evidence—he discovered that books which receive "the most early advertising can generally be counted on to'have superior s a l e s " . " 0 Yet Berreman concluded that "the bulk of advertising of best s e l l e r s accompanies or follows sales rather than precedes them, and i t i s impossible to predict from advertising even a month after publication-date, the r e l a t i v e success which best s e l l i n g novels w i l l 17 achieve". Then, too, one wonders how much sales volume act u a l l y depends on p o s i t i v e reaction by reviewers, public acclaim and word-of-mouth advertising. Lettuce Assuming that variations i n the per capita consumption of lettuce are a function of the area's per capita income, i t s temperature, r e t a i l p r i c e and the i n t e n s i t y of advertising e f f o r t , Meissner performed a multiple regression analysis of lettuce sales i n 22 c i t i e s . In the period 1950—1955, "only four variables, p r i c e , temperature, fieldman and nev/spaper advertising had a s i g n i f i c a n t influence on lettuce consumption The eleven independent variables analyzed by Meissner "explained" only h a l f of the t o t a l v a r i a t i o n i n the per capita J o e l V. Berreman,"Advertising and the Sale of Novels, Journal of Marketing, 7:237, January, 1943. 1 7 I b i d . l 8Frank Meissner,"Sales and Advertising of Lettuce," Journal of Advertising Research, 1:6, March, 1961. 33 consumption of lettuce—38 percent "was explained" by price, income and weather, and only twelve percent by the eight advertising variables. As a result, the author did not provide any concrete conclusions regarding the advertising-sales r e l -ationship. Moreover, his use of p a r t i a l c o r r e l a t i o n c o e f f i c -ients as i n d i c a t i v e of causal influence on sales has been challenged. "^ Cookware Following declining sales of Teflon-coated cookware, Dupont began an advertising experiment employing a f a c t o r i a l design. Thirteen American c i t i e s received three levels of advertising e f f o r t i n late 1362 and 1963. "The major research measurement of sales during each test period was a wave of telephone interviews conducted with samples of 1000 female 20 heads of households m each of the test markets" . Beclcnell and Mclsaac's findings are summarized below. 19 See, John C. Maloney, "Letters," Journal of  Advertising Research, 1:32, June, 1961. ^°James C. Becknell and Robert S. Mclsaac, "Test Marketing Cookware Coated with 'Teflon'," Journal of  Advertising Research, 3:3, September, 1963. 34 The researchers also investigated the carry-over e f f e c t of advertising, and concluded that the promotional campaign "worked at the high l e v e l but had no discernible e f f e c t at 21 the lower l e v e l . . . " . TABLE III • .' COOKWARE UNITS PER 1000 FEMALE HEADS OF HOUSEHOLDS (SOURCE: JAMES C. BECKNELL AND ROBERT W. McISAAC, "TEST MARKETING COOKWARE COATED WITH 'TEFLON'," JOURNAL OF ADVERTISING RESEARCH, 3:5, SSPTEM3ER 1963) . FALL 1962 WINTER 1963 HIGH ADVERTISING LOW OR NO ADVERTISING HIGH LOW OR NO ADVERTISING ADVERTISING Total Units ( A l l Types) 404 317 268 221 Units Coated with Teflon ( A l l Types) 38 16 59 27 S k i l l e t s and Griddles Coated with Teflon 2e 16 27 13 Teflon Market Share 9% 5% 22% 12% Apples Tousley compared per capita advertising expenditures to the consumption of apples i n 38 di f f e r e n t areas. "A consumption trend for each market was obtained by computing the percentage change i n the average annual c a r - l o t unloads of Washington apples between two periods, 1932 — 1935 and 1937 —1339" . 2 2 2 1 I b i d , p. 8. 22Raymond D. Tousley, "Advertising Fresh F r u i t s and Vegetables I I , " Harvard Business Review, 5:81, A p r i l , 1927. 3 5 Unfortunately, the test markets were rather atypical and Tousley was forced to conclude that the advertising-sales relationship was not e n t i r e l y consistent.'" Oranges Investigating promotion's role in the sale of oranges, Nerlove and Waugh analyzed the 1 9 1 0 — 1 9 5 9 advertising expend-itures of Sunkist Growers, Incorporated. An interesting theory of cooperative advertising without supply control i s f i r s t p r o p o u n d e d . T h e writers then studied the value of the orange crop with respect to quantity sold (Qt) , consumer income (Yt), the current advertising appropriation (At) and the mean advertising expenditure (At) over the 5 0 year period. Several regressions were run using numerous lagged values of the advertising a p p r o p r i a t i o n — a unique two parameter form of d i s -tributed lag gave better results than an exponentially d i s t r i b -uted lag. The estimating equation obtained was: Ut= -2.939 - 0.390Qt+ .924Yt+.233At+ 1.03At , ' , ^However, Tousley's a r t i c l e i s an excellent summary of many studies investigating the effectiveness of advertising for farm products. ^^A highly mathematical discussion of the monopolist's advertising problem may be found i n . Marc Nerlove and Kenneth Arrow, "Optimal Advertising Policy Under Dynamic Conditions," Economica, 39:129-142, May, 1962. ^Marc Nerlove and Frederick V.Waugh,"Advertising Without Supply Control:Some Implications of a Study of the Advertising of Oranges," Journal of Farm Economics, 43:832, November, 1961. 36 The authors concluded that advertising and consumer income were omnipotent factors in the expansion of orange sales; the equation indicates that " i f orange production remained constant, an added d o l l a r of advertising would rais e the gross returns to orange producers by over twenty d o l l a r s " - Other regressions demonstrated that marginal returns from advertising rapi d l y decline as advertising expenditures increase. As F i g u r e s shows,2^ the e f f e c t of advertising has been to increase farm value by much more than the cost of advertising. Farm Value ($ per capita) 1.50 1.00 .50 - -Current Advertising Expenditures ($ per capita) .005 .020 10 Year Average Advert-i s i n g Expenditures ($ per capita) FIGURE 5 NET REGRESSIONS OF VALUE ON ADVERTISING EXPENDITURE 2 6 I b i d , p. 835 27 Ibid, p. 832. 37 Disregarding the low advertising-sales r a t i o maintained by orange growers, Nerlove and Waugh may be a t t r i b u t i n g far more to advertising than i s r e a l i s t i c a l l y j u s t i f i e d . The model i s e s s e n t i a l l y s t a t i c , and as they point out, "when supplies are uncontrolled, i t i s impossible to judge the long run effects of advertising without taking account of such matters as the long run e l a s t i c i t y of supply and external economies or d i s -28 economies". F i n a l l y , the writers compute an optimal adver-t i s i n g budget, assuming that the above estimating equation remains v a l i d beyond the range of the data. As any economet-r i c i a n well knows, such an assumption i s highly suspect. Dairy Products In a simple p i l o t study, Dickens presented evidence that sales of dairy products can be increased by point of purchase 29 advertising i n grocery stores. Dicken's findings are summ-arized i n Table IV; however, her before-after study raises many more questions than i t answers. Clement measured the e f f e c t of three increasing levels of advertising expenditures on t o t a l milk sales. Government sales data, supplemented by r e t a i l audits and waves of telephone interviews gave an accurate measure of milk consumption. In t h i s elaborately designed experiment, "sales increased with each higher l e v e l of expenditure and expenditures i n the previous • °Ibid, p. 836 29 Dorothy Dickens, "Advertising Dairy Products In Rural Grocery Stores."Journal of Marketing, 19:268-270, January, 1955. 38 period had an e f f e c t on sales in the present period". From the experimental data, Clement succeeded i n determining the optimum advertising expenditure by substituting the optimum sales l e v e l (10.4 thousand pounds of sales per day) i n the estimated advertising cost function: Total Cost = 78.3 - 10.7+ 1.52X2 Total Cost = $131 TABLE IV SALES INCREASES ATTRIBUTED TO ADVERTISING (SOURCE: DOROTHY DICKENS,"ADVERTISING DAIRY PRODUCTS IN RURAL GROCERY STORES," JOURNAL  OF MARKETING, 19: 269, JANUARY, 19551 PRODUCT SALES DURING PERCENTAGE INCREASE IN CONTROL PERIOD SALES DURING TEST PERIOD Whole Milk 457 15 Buttermilk 341 - - 4 Chocolate Milk 50 - 16 Chocolate Drink — 150 25 Cheddar Cheese 231 18 Ice Cream 216- 22 Evaporated Milk 651 4 Condensed Milk 52 - 2 Coffee and Cleansers Using multiple regression analysis, Banks found that brand preference, price and in-store promotional e f f o r t were s i g n i f i c a n t i n determining the market shares for various brands of scouring cleanser. A l t e r n a t i v e l y , brand preference 3 0Wendeil E. Clement, "An Analysis of the Advertising Process and I t s Influence on Consumer Behavior," Paper presented at the 1968 F a l l Conference Proceedings. (Denver, Colorado: American Marketing Association), p. 3. (Mimeographed). 39 was the only obvious determinant of brand share for coffee. Incidentally concerned with the sales effects of advertising, Banks makes a p a r t i c u l a r l y i n t e r e s t i n g assertion: whereas Maxwell House was spending 40 percent of the t o t a l coffee advertising volume i n Chicago, i t was receiving only sixteen percent of t o t a l coffee sales. "In contrast, Manor House was spending only nine percent of the t o t a l advertising volume, but receiving 30 percent of t o t a l sales" Lamb Hoofnagle c i t e s two controlled experiments to guage the e f f e c t of promotion on lamb sales. The 1956 study employed a crude subdivided time series approach wherein r e t a i l sales were checked before, during and after, the advertising period. This procedure did not give d e f i n i t i v e r e s u l t s . Using a double changeover design, the T961 experiment attempted to assess the effectiveness of a cooperative and consumer advertising program. "Results of the experiment showed that the combined weekly lamb sales for (three) northeastern and midwestern c i t i e s aver-aged 26 percent higher for cooperative advertising and ten per-cent greater for the regular promotion program than for com-3 2 parable periods of no advertising and merchandising support". 31 S eymour Banks, "Some Correlates of Coffee and Cleanser Brand Shares," Journal of Advertising Research, 1:27, June, 1961. 3 2Wi l l i a m S, Hoofnagle, "The Effectiveness of Advertising For Farm Products," Journal of Advertising Research, 3:5-6, December, 1963, 40 Cigarettes Examining the r e l a t i o n between the consumption and price of cigarettes, Schoenberg b r i e f l y considered the effects of promotion: "advertising data were introduced into the analysis on the assumption that the consumption of cigarettes depends on the p r i c e , on the amount of advertising and on other fac-t o r s . . . the relationship between the foregoing variables i s almost perfect, the c o e f f i c i e n t of multiple c o r r e l a t i o n being 33 almost unity (.998)". Wagner, too, correlated cigarette sales and advertising indices. While both indices tended to more coincidentally, he found that "beginning i n 1937, the advertising index rose higher than the sales i n d e x " . 3 4 This r e f l e c t s diseconomies, since increased advertising d o l l a r s had l i t t l e a f f e c t of sales. In a widely acclaimed study, Telser used regression analysis to investigate the r e l a t i o n of sales to advertising for three large cigarette brands. Given that p r o f i t maxim-i z a t i o n requires diminishing marginal productivity of adver-t i s i n g , he examined four regressions with d i f f e r e n t implications 35 concerning the nature of the returns to advertising: 33 E. H. Schoenberg, "The Demand Curve For Cigarettes," Journal of Business, 6:27, January, i933. 34 Louis C. Wagner, "Advertising and the Business Cycle," Journal of Marketing, 6:133, October, 1941. 35;Leon Telser/'Advertising and Cigarettes," Journal of  P o l i t i c a l Economy, 70:478-80, October, 1962. 41 1) Advertising has a constant marginal product, the second derivative of sales with respect to advertising i s zero. q t = A 4- Bxfc + Cy t-f Dp t + E t 2) Advertising exhibits a decreasing marginal product; the second derivative of sales with respect to advertising i s -B1/x2< 0. q f c= A1^- B L l o g x f c+ C 1 l o g y t 4 D 1 l o g p ^ E 1 t A regression on f i r s t differences of the above variables was also run. 3) Advertising has an increasing marginal product wherein the second derivative of sales with respect to advertising i s b 1 2 q ^ 0 . log q f c = a+-bx t 4-cy t 4-dp t + efc Note i n the above equations that: q^ . Sales i n bullions of cigarettes x f c Advertising outlay i n mil l i o n s of dol l a r s y t National income divided by the consumer price index p t R e t a i l price per package of cigarettes divided by the consumer price index t Linear trend Although Telser's "research" may be c r i t i c i z e d on grounds that i t i s merely a c o l l e c t i o n of s t a t i s t i c s to prove a predet-ermined viewpoint, this censure c e r t a i n l y does not invalidate his findings. The best regression for predicting the marginal . M l ' . • sales effects of advertising was the f i r s t difference form of equation t w o — i t s residuals had the lowest s e r i a l c o r r e l a t i o n , and i t had the lowest standard error of p r e d i c t i o n . Telser also investigated the competitive effects of advertising by s c r u t i n i z i n g the r e l a t i o n between market share and advertising expenditures. He suggests there was "closer competition between the advertising on Camels and Lucky Strike than between any one of these brands and the C h e s t e r f i e l d ex-penditures". Perhaps more important, Telser determined that "advertising outlays b u i l t up a fund of goodwill that depreciated 37 at a rate varying between f i f t e e n and 20 percent per year". Of course, t h i s depreciation decreased as the r e l a t i v e adver-t i s i n g expenditures increased. Drugs i . Hollander i s o l a t e d six variables which apparently expl-ained the sales variations of "a nationally advertised drug product". Using graphic multiple c o r r e l a t i o n , he determined the quantitative r e l a t i o n s h i p between advertising and sales. To measure the cumulative effects of advertising, Hollander successfully tested a "moving average formula b u i l t on assump-tions inherent i n the advertising s i t u a t i o n regarding the rate 38 at which the advertising e f f e c t i s b u i l t up and runs out". Unfortunately, Hollander presented no empirical data. 3 6 I b i d , p. 487. 3 7 I b i d , p. 498. 3 8 H o l l a n d e r , o£. c i t . , p. 85, 43 Robinson reported a case study i n which a pharmaceutical firm, content with i t s present sales, decided to drop i t s adver-39 t i s i n g program. As shown i n Figure 6 , t h i s strategy proved disastrous and the advertising was soon restored. Advertising "Normal" "Normal" Sales No Advertising Advertising Again TIME FIGURE 6 A DRUG PRODUCTS ADVERTISING-SALES RELATIONSHIP 39T Patrick J . Robinson, "Management Science In Marketing: Capsule Cases From Seven Years 0 R Experience," i n Better  Measurements of Advertising Effectiveness Proceedings, F i f t h Annual Conference (New York: Advertising Research Foundation, 1959), pp. 79-94. 44 Applying multiple regression techniques to consumer panel data, Roberts uncovered evidence that firm A's "adver-t i s i n g was much more productive with respect to A's sales than was B's advertising with respect to B's s a l e s " . 4 0 In addition, he found that the r e l a t i o n between advertising and sales was c u r v i l i n e a r ; because of t h i s , "one can i n f e r that a r e a l l o c a t i o n of i n s e r t i o n schedules among media may r e s u l t i n increased p r o d u c t i v i t y " . 4 1 Palda applied many regressions, lagged and nonlagged, i n an attempt to explain annual variations i n the sales of a 42 vegetable compound. More s p e c i f i c a l l y , h i s d i s s e r t a t i o n questioned whether the measurement of cumulative advertising e f f e c t s could be improved by using Koyck's model of d i s t r i b u t e d lags. A p r i o r i reasoning suggested that the time-shape of the sales reaction to advertising could be expressed by a d i s t r i b u t e d l a g . Unlike previous models, the simple Koyck model employs only one lagged and one nonlagged exogenous variable. " I f this s u b s t a n t i a l l y simpler model were to give nearly as good a p i c -ture of r e a l i t y as the more complicated ones, i t would be 4^Harry V. Roberts, "The Measurement of Advertising Results," Journal of Business, 20:141, July, 1947. 4 1 I b i d , p. 145. ^ K r i s t i a n S. Palda, The Measurement of Cumulative  Advertising E f f e c t s , (Englewood C l i f f s , New Jersey: Prentice H a l l , Inc., 1964). 45 advantageous to employ i t " . 4 3 Applied to the advertising-sales s i t u a t i o n , the Koyck model hypothesizes that movements in sales assume the d i s t r i b -ution of a geometric progression from time period one, onward: Lagging t h i s equation one time period, multiplying i t by J\ and subtracting the r e s u l t from the above, one arrives at the basic Koyck equation: -AS^J -Ad-r^AA^f t °vtV\^> . . .-hut-/' Applying numerous s t a t i s t i c a l t ests, Palda concluded that the d i s t r i b u t e d lag model, with a logarithm of advertising as an independent variable, gave the best f i t to the data. Not only did the "cumulative models tend to outperform the non-cumulative ones", 4 4 but the "semilogarithmic functional forms of the regression models gave consistently better results than those i n which the advertising variable was not used i n logar-ithmic form", 4^ Because some noncumulative models performed almost as well as similar cumulative ones, Palda's findings are not wholly conclusive. The results show advertising as a major stimulator of sales. Whether the Koyck hypothesis can be successfully employed by advertisers other than Lydia Pinkham 4 3 I b i d , p. 15. 44 Ibid, p. 77. 4 5 I b i d , p. 77. 46 i s c e r t a i n l y questionable. Like mail order data, the r e l a t i o n -ship between Pinkham's advertising and sales was completely measureable—there was an absence of intervening marketing inputs. Today, few companies u t i l i z e only one facet of the marketing mix, and fewer survive unaware of t h e i r competitive environment'. Automobiles V a i l e analyzed the r e l a t i v e variations i n the sales and i advertising volumes of eighteen automobile manufacturers between AC. 1920 and 1924. ° His data i s summarized below. Apparently, the difference between the sales of those firms which increased t h e i r advertising lineage and the sales of those who decreased i t , i s r e l a t i v e l y small. Nevertheless, on the basis of numerous other studies, V a i l e concluded i t was b e n e f i c i a l to increase advertising expenditures during a "depression" so as to gain a d i f f e r e n t i a l advantage over competitors. What such a p o l i c y does to the firm's p r o f i t picture was c a r e f u l l y avoided'. TABLE V RELATIVE MOVEMENT OF SALES ASSOCIATED WITH DIFFERENT ADVERTISING POLICIES (SOURCE; ROLAND S. VAILE "THE USE OF ADVERTISING DURING DEPRESSIONS," HARVARD BUSINESS REVIEW, 5: 328, APRIL, 1927.). POLICY 1920 1921 1922 1923 1924 Increased Advertising 100 80 i 109 98 95 No Advertising 100 100 100 100 100 Decreased Advertising 100 77 98 85 88 46Roland S. V a i l e , "The Use of Advertising During Depressions," Harvard Business Review, 5:323-330, A p r i l , 1927. 47 Also relevant to the present paper i s Cowan's 1936 sales analysis study i Assuming "that i f any chosen type of advertis-ing affects sales, the e f f e c t should be greater i n some areas 47 than others", he examined the association between Chevrolet reg i s t r a t i o n s and the c i r c u l a t i o n of the Saturday Evening Post. The automobile was advertised i n th i s magazine for many years. TABLE VI ASSOCIATION BETWEEN SATURDAY EVENING POST AND CHEVROLET REGISTRATIONS PER 1000 PEOPLE IN 500 AREAS (SOURCE; DONALD R. G. COWAN "SALES ANALYSIS FROM THE MANAGEMENT STANDPOINT," JOURNAL OF  BUSINESS, 9:175, JULY, 1936.) COPIES OF SAT-URDAY EVENING POST ESTIMATED CHEV-ROLET REGISTRA-TIONS PER 1000 PEOPLE WITH EACH SUCCESSIVE ADDITION OF 10 SATURDAY EVENING POST PER 1000 PEOPLE INCREASE IN CHEV-ROLET REGISTRATIONS % INCREASE 0 6.0 10 13.5 7.5 125.0 20 -25.0 11.5 85.2 30 35.0 10.5 42.0 40 44.0 8.5 23.9 50 51.0 7.0 15.9 60 57.0 6.0 11.8 70 61.0 4.0 7.0 80 65.0 4.0 6.6 90 69.0 4.0 6.2 100 72.0 3.0 4.3 Cowan's marginal data c l e a r l y demonstrates that heigh-tened automobile advertising i s f i r s t accompanied by increased 4 7Donald R.G. Cowan, "Sales Analysis From the Management Standpoint," Journal of Business, 9:175, July, 1936. 48 sales, but ultimately economies of scale are evidenced. By comparing automobile sales and advertising, Wagner was able to demonstrate that both indices tended to move to-gether. "Correlations run to determine the timing of f l u c t u -ations i n advertising and i n sales indicated that changes i n 48 advertising volume preceded changes in sales". With advertising lagged two months, the highest c o r r e l a t i o n was evident: i TABLE VII CORRELATIONS - AUTOMOBILE SALES AND ADVERTISING (SOURCE; LOUIS C. WAGNER, "ADVERTISING AND THE BUSINESS CYCLE," JOURNAL OF MARKETING, 6:2, OCTOBER, 194.1, p. 133) CORRELATION (r.) ; ADVERTISING .73578 .81197 ,84501 .85659 .84382 Summary and Conclusions Although i t i s generally accepted that economies of scale do e x i s t i n a d v e r t i s i n g , ^ 9 this b r i e f review of the l i t -erature shows that the evidence i s inconclusive. Researchers 4 8Wagner, o£. c i t . , p. 132. 4 9 A n opposite viewpoint i s elaborated by J u l i a n L. Simon, "Are There Economies of Scale i n Advertising?" Journal  of Advertising Research, 5: 15-20, June, 1965. Ahead one month S ame month Lagged one month Lagged two months Lagged three months 49 have been struggling with the advertising-sales r a t i o for many years, and i t appears impossible to derive s p e c i f i c axioms at thi s time. Moreover, the many methodological d i f f -i c u l t i e s inherent i n an investigation of the advertising-sales relationship w i l l only be overcome i n small stages. CHAPTER III INTRODUCTION As the l i t e r a t u r e survey revealed, controlled experim-entation i s frequently used i n documenting the sales response to advertising. The small businessman has neither the expertise nor f i n a n c i a l capacity to design such elaborate marketing experiments despite an urgent need to discover the underlying quantitative r e l a t i o n s h i p between advertising and sales. To t h i s end, multiple regression analysis has become fashionable as a successful alternative to experimentation. Using only accurate company records, the regression technique should enable one to guage the degree of association between sales and advertising. Multiple regression analysis b a s i c a l l y employs quant-i t a t i v e evaluations of data—consequently, the p a r t i c u l a r units chosen to express the primary data used i n t h i s study must be outlined. F i r s t , a few introductory comments may be i n order. I t i s important to recognize that the automobile dealer's advertising appropriation was allocated on an ad hoc basis. As a r e s u l t , the c o r r e l a t i o n between advertising and sales may not be purposefully high, though i t may be low. Moreover, the influence of competitive advertising on the dealer's promotional strategy i s of course, an unquantifiable phenomenon; doubtless, t h i s variable i s important. To the academic mind, the adver-t i s i n g expenditures may appear i n s i g n i f i c a n t r e l a t i v e to the 5 1 sales volume. To the p r a c t i c i n g businessman, however, any promotional expenditure which does not increase sales becomes in t o l e r a b l e . Previous empirical research has demonstrated that many factors influence automobile sales. Such socio-economic data are t y p i c a l l y c o l l e c t e d i n yearly form, and hence cannot be included i n the regression analysis. By the same token long term secular factors are assumed to a f f e c t a l l automobile dealers i n s i m i l a r manner—there i s l i t t l e need to consider these variables further. Because the sales effects of advertising r e f l e c t both media and copy q u a l i t i e s , the exclusion of the l a t t e r variable from the analysis i s an o v e r s i m p l i f i c a t i o n . Since careful evaluation of the advertising copy suggested that most promotion emphasized immediate buying action, this oversight should not contaminate the r e s u l t s . THE PRIMARY DATA Raw data subjected to regression analysis was gleaned from various company records and occasionally secondary sources: Newspaper Advertising The newspaper was the p r i n c i p a l medium for advertising. This variable was defined to include advertising placed i n d a i l y and weekly newspapers c i r c u l a t e d throughout the greater i metropolitan area. Advertisements were mainly two types: 52 display and c l a s s i f i e d . "Display advertising usually involves i l l u s t r a t i o n s and appears throughout the paper; c l a s s i f i e d advertising i s usually concentrated on special pages of the newspaper under headings that c l a s s i f y the various items''.^" Over ninety percent of the newspaper advertising expend-i t u r e was concentrated i n Vancouver's two d a i l y editions; suburban weekly publications accounted for the remainder. Because c i r c u l a t i o n figures were f a i r l y stable throughout the year, i t was decided not to include this factor among the independent variables. Invoices provided the information necessary to determine the two newspaper advertising s e r i e s . A l l b i l l i n g s were red-uced to weekly lineage and d o l l a r expenditure t o t a l s . While t h i s procedure proved tedious, i t was e s s e n t i a l for comparing dealer and manufacturer newspaper advertising volumes. Radio Advertising Commercial announcements were aired on f i v e l o c a l stations, the i n t e n s i t y of advertising pressure being greatest during the i n i t i a l and l a t t e r weeks of the year. A physical index, such as "gross media audience" was contemplated as a measure of radio advertising. Because the rate structure i n the broadcasting industry e x p l i c i t l y assumes d i f f e r e n t i a l audience s i z e s , weekly expenditures on a l l radio stations were J^.ames M. Ferguson, The Advertising Rate Structure i n  the Daily Newspaper Industry,(Englewood C l i f f s , New Jersey: Prentice H a l l , Inc., 1962), p. 7. 53 thus subjected to regression analysis. At best, audience figures are an extremely crude indicator of advertising effectiveness. T e l e v i s i o n Advertising During January, February and March spot announcements were placed on two t e l e v i s i o n stations. Again, weekly expend-it u r e s on t h i s medium were t o t a l l e d from invoices. Although t e l e v i s i o n was used sparingly, i t was included i n the analysis because of i t s r e l a t i v e high cost. Only to the extent that sales are responsive to t e l e v i s i o n advertising i s t h i s higher t cost j u s t i f i e d . Broadcast Dollars This independent variable i s merely the summation of weekly radio and t e l e v i s i o n expenditures. Manufacturer Advertising During the year, three methods of communication were used by the manufacturer: national t e l e v i s i o n , magazines and l o c a l newspapers. Although i t was desirable to c o l l e c t this data, the national figures could not be secured. However, this deficiency i s not c r u c i a l . Had such information been available, i t would be severely understated by the amount of "splash-over advertising" from the United States. The manufacturer's l o c a l newspaper advertising lineage and d o l l a r expenditure must therefore serve as a rough i n d i c -ator of advertising i n t e n s i t y . Two further limitations on this data must be recognized. Whereas the manufacturer paid almost 54 $1.25 per newspaper l i n e , the automobile dealer paid approx-imately f o r t y cents. Since t h i s rate d i f f e r e n t i a l between national and l o c a l advertising i s not d i r e c t l y applicable to units of newspaper lineage, this variable may be a better indicator of advertising volume. Secondly, the manufacturer's advertising series was available only i n monthly form. Because t h i s data was pro-rated on a weekly basis, the figures w i l l be overstated i n some weeks, but understated i n others. Operating within the confines of a franchise system, the automobile dealer bears a s i g n i f i c a n t part of the manufacturer's promotional expenditure. Whether the franchisor's l o c a l adver-t i s i n g s t a t i s t i c a l l y influences dealer sales i s a question of concern to a l l f r a n c h i s e e s — i t was thus included i n the analysis. Total Advertising The dealer's weekly expenditure on a l l media are summar-ized i n t h i s v a r i a b l e . Note that " t o t a l advertising" does not include advertising placed by the manufacturer. Weather Just as the automobile industry's health t r a d i t i o n a l l y r e f l e c t s that of the economy, so automobile sales mirror c l i m a t i c v a r i a t i o n s . To assess the influence of weather upon sales, p r e c i p i t a t i o n was chosen as an independent variable. This factor was defined to include the measurable amount of r a i n and snow f a l l i n g on any given day. Records were secured from the weather o f f i c e at Vancouver International Airport, and summed to a weekly basis. 5 5 Sales Every automobile salesman recognizes that the s e l l i n g p r i c e of a vehicle i s an outstanding factor i n the choice of a dealer. As the dependent variable i n t h i s study, "sales volume" refers to the t o t a l r e t a i l s e l l i n g p rice of a l l automobiles sold during the week. Such figures exaggerate the dealer's return i n that the trade-in value of the old model i s deducted from the r e t a i l p r i c e . Net cost to the consumer was not employed as a measure of sales because the actual mechanics of the buyer agreement are the function of the salesman. Since advertising copy mentioned only gross s e l l i n g p r i c e , t h i s figure was u t i l -i z e d . I t was also convenient to measure sales volume i n physical u n i t s . Then for each week, another variable, average s e l l i n g p r i c e was calculated; this i s simply the t o t a l r e t a i l s e l l i n g p r i c e divided by the number of units sold. Distance , To assure p r o f i t a b i l i t y , management should know which areas of the c i t y are being successfully exploited. From buyer agreements, each customer's address was obtained and plotted on a map of the Greater Vancouver area. A g r i d of one mile by one mile squares was then placed upon the map. Using a space-distance approach, the g r i d enables a row and column number to h i t approximately every household which purchased an automobile. Although the g r i d could have been drawn so precise as to i n t e r -sect at each customer's residence, t h i s would have proved extremely cumbersome. As a r e s u l t , a l l residences within a 5 6 zone are treated equally—we assume that customers are located at the nearest intersecting coordinates. The origin-destination technique f a c i l i t a t e s a general picture of the s p a t i a l relations between s e l l e r and buyer. By merely subtracting the coordinates of the customer's address from those of the dealership and applying the Pythagorean theorem, the approximate distance between the r e t a i l e r and every customer was computed. F i n a l l y , weekly average and median distances were calculated, using the computer program displayed i n the Appendix. CHAPTER IV INTRODUCTION The previous chapter discussed a number of variables that could be used to explain fluctuations i n an automobile dealer's sales volume. This thesis i s but a b r i e f investigation of the subject, for there ex i s t numerous seemingly unquantifiable factors which influence the automobile buying process. A d i s -cussion of these parameters i t the focus of this chapter. Where applicable, guidelines for future inquiry are suggested. ADVERTISING PERFORMANCE AND OTHER FACTORS Consumer Product Preferences Within a geographical area, the franchisees of a given manufacturer can anticipate only a li m i t e d market share. To a v large extent, each dealer's c e i l i n g sales l e v e l i s determined by customer product preferences. The automobile buying process i s probably the most documented aspect of patronage behavior. Whether one believes the assertions of Vance Packard or Franklin Evans i s immaterial. I t i s c r u c i a l to recognize that many studies consistently f i n d that brand preference among automobile purchasers, c e r t a i n l y not mutable, i s somewhat consistent. The used car showroom of any automobile r e t a i l e r substantiates this axiom. An examination of trade-ins in the present study suggests .that repurchase of the currently owned make occurs almost three times as often as purchase of a d i f f e r e n t make. 58 That brand l o y a l t y toward the currently owned make i s r e l a t i v e l y high has important repercussions for marketing strategy. The franchise system of d i s t r i b u t i o n guarantees there w i l l be no "mass market" for the " t o t a l product" offerings of a given dealership. Rather, the market i s l i m i t e d to the extent of p o t e n t i a l buyers des i r i n g , manifestly or l a t e n t l y , the prod-ucts of each manufacturer. At t h i s l e v e l , each franchise system competes v i s - a - v i s every other,- s i m i l a r l y , each franchisee's advertising competes with the promotional "noise" of a l l con-tiguous dealers. I f consumers prefer a s p e c i f i c manufacturer's product, intra-dealership competition also exists within the same franchise system. Whether inter-franchise or intra-franchise competition i s more s i g n i f i c a n t must ultimately depend upon the consumer. I f brand loyalt y i s high, prospects w i l l generally be more receptive to the advertising of those franchisees market-ing the currently owned or o r i g i n a l l y preferred model. On the other hand, the choice of an automobile i s defin-i t e l y not an habitual act. Customer product preferences change through time, as the Volkswagen and Mustang successes t e s t i f y . I f consumers perceive the offerings of several manufacturers as homogeneous, they may meander from r e t a i l e r to r e t a i l e r seeking that one l a s t p r i c e . As the i n t e n s i t y of shopping increases, both inter and intra-franchise promotional competition w i l l probably heighten. However, evidence suggests that the search w i l l be of a limited nature, perhaps considering only two or three alternative makes and dealers. The burden on advertising 59 w i l l be great, e s p e c i a l l y i f consumers choose the desired model before shopping'. Competition and Location Models of the advertising process rarely mention r e t a i l a v a i l a b i l i t y as an e x p l i c i t parameter. I t i s readily apparent that t h i s variable i s linked to the factor discussed above. In automobile r e t a i l i n g , " a v a i l a b i l i t y " implies not only a consid-eration of the competitive interface, but also the influence of location upon advertising strategy. Since an automobile franchise i s d i f f i c u l t to obtain, the number of newly-appearing dealerships i s r e s t r i c t e d . Such a s i t u a t i o n appears conducive to analyzing the sales effects of franchisee advertising. Yet the universe of dealerships remains s u f f i c i e n t l y large that competitors' marketing a c t i v i t i e s con-ceivably influence purchasing behavior. A University of Nevada survey of advertising by small business concluded that automobile dealers "set t h e i r advertising expenditures i n the l i g h t of com-p e t i t i v e a c t i v i t y i n twenty eight percent of the cases versus an average of only eleven percent".^" Within bounds, this method is c e r t a i n l y j u s t i f i a b l e ; knowing where, when and how competitors are advertising assists the dealer i n planning his own promotion. From a research standpoint,the d i f f i c u l t y obviously l i e s i n quantifying competitive advertising a c t i v i t y . ^•University of Nevada,"The Extent of R e t a i l Advertising As A Management T o o l — I t s Scope and Importance i n Small Business," Washington: Small Business Administration, 1 9 6 1 . p. 67. (Mimeogr aphed) 60 Location i s another potent merchandising variable often neglected i n advertising research. However, the a c c e s s i b i l i t y and q u a l i t y of a location complement the effects of an aggress-ive advertising campaign. If the dealer possessed an "optimal location" r e l a t i v e to that of his competitors, he could prob-able allocate less dollars to advertising. At best, marketing geography i s s t i l l very much an art; research i s o l a t i n g s p a t i a l contributions to sales remains i n the f e t a l stage. To what extent human ecology may be manipulated by advertising should be the subject of much inquiry. Over the years, Vancouver automobile dealers' land requirements have been increasingly met by s i t e s located on the outer fringe of the built-up urban area. Locating on major r a d i a l s and ribbons has proved popular. Yet the spacing of dealers i s reasonably c l o s e — t h i s i s no doubt a function of geographically concentrated purchasing power, the constrained Vancouver transportation network and the l i m i t e d time-distance people are w i l l i n g to t r a v e l . As a r e s u l t , advertising w i l l continue to play a major role i n demand stimulation. Seasonal Demand and Price Levels Even i f advertising were of no consequence i n automobile r e t a i l i n g , one would expect weekly movements i n sales volume. External influences, such as holidays, weather and other events cause sales to fluctuate. As the year progresses, automobile sales usually increase, peak and decline. Planned obsolesence. 61 i n the form of new model introductions, i s designed to expand demand when i t i s otherwise at a low ebb. Throughout the annual sales cycle, the price l e v e l of the automobile may vary. I f the "economic man" foresees price re-ductions during the "clean-up" model period, he can e a s i l y forego purchase. Conversely, i f his need is urgent, he may purchase immediately. This dichotomy i l l u s t r a t e s a perplexing problem to advertising researchers: does a high c o r r e l a t i o n between current sales and advertising indicate that the dealer i s just s e l l i n g i n March what he would otherwise have sold i n September? At present,the answer i s indeterministic. C r e d i t , Disposable Income and Intentions To Buy Measurements performed on one year's advertising prod-u c t i v i t y probably include a hi s t o r y of previous promotion and market conditions-*,, Because each year contains d i s s i m i l a r events that a l t e r the f i r m ^ ^ demand curve, i t i s suggested that business * i research be of a longitudinal nature. In t h i s way, the e f f e c t of r i s i n g disposable income and consumer willingness to entertain debt may be more accurately assessed. A p o t e n t i a l l y r i c h source of information, often overlooked i n advertising research, i s consumer intention to buy data. In Canada, t h i s service i s offered by the MacLean Hunter Research Bureau. One might use t h i s data to determine whether a r i s e i n expectations has h i s t o r i c a l l y been associated with an increase i n company sales. Is analysis of the automobile "consideration 62 class" useful in planning the advertising campaign? Do they become members of the"purchasing class"? By t h e i r very nature, expectational data summarize a host of factors influencing the demand for durables. Because the MacLean Hunter survey i s published quarterly, the reader may f e e l t h i s information i s outdated for purposes of operational-i z i n g current advertising strategy. However, a recent study indicates that "both lagged purchasing plans and plans at time 2 t had about the same predictive a b i l i t y " in estimating Canadian automobile sales. Highlighting the importance of disposable income, Murray found that the regression equation for B r i t i s h Columbia automobile sales (1960-1967) was: SALES=.214 It - h 2.143 Yt-1 - 12150.33 where purchasing plans (It) and disposable income from the previous period (Yt-1) are the explanatory variables. The s i g n i f i c a n t c o e f f i c i e n t of multiple determination, .842, indicates that buying intentions may prove a useful variable in future studies of advertising productivity. The E f f i c i e n c y of Personal S e l l i n g An understanding of the interdependency between advertising and personal s e l l i n g i s prerequisite to sound advertising research. Within the automobile r e t a i l i n g context, a convincing argument can probably be voiced e x t o l l i n g the function of the sales force Alex. J . Murray, "Canadian Consumer Expectational Data: An Evaluation," Journal of Marketing Research, 5:56, February, 1969. 63 and minimizing that of advertising. This proposition i s not true. I f the "economics of advertising" i s an inconclusive concept, then the "economics of personal s e l l i n g " i s far less c e r t a i n . In theory, the complexities involved in measuring the productivity of each function are s i m i l a r . In practice, more tangible c r i t e r i a may be applied to documenting each salesman's performance than to each advertisement's e f f e c t . I s o l a t i n g personal s e l l i n g ' s contribution from that of advertising seems v i r t u a l l y impossible. Although the persuasive and communication aspects of each d i f f e r , both inputs are complementary. Advertising probably functions at an early point i n the consumer's purchase cycle, salesmanship during the l a t t e r stages. Such a "push-pull" concept i s only partially-applicable to automobile r e t a i l i n g . Nevertheless, i t i l l u s t -rates t h i s e ssential degree of complementarity; simultaneously, the d i f f i c u l t y of i s o l a t i n g i n d ividual contributions is apparent. Advertising studies notoriously relegate personal s e l l -ing to the "other things equal" category. This is an esp e c i a l l y dangerous assumption in automobile r e t a i l i n g . The mobility of automobile salesmen often reaches velocity proportions and con-sequently, the c a l i b r e of the sales force i s i n a constant state of f l u x . Future research on advertising e f f i c i e n c y might employ "dummy" variables to represent sales force q u a l i t y . Such an approach has i n t u i t i v e appeal and i s based on a simple premise— the salesman who i s immobile between dea l e r s h i p s i s less dependent upon a d v e r t i s i n g as a generator of showroom t r a f f i c than i s h i s wandering counterpart. By examining personnel and sales records, sales force c a l i b r e may be indexed over time and i t s e f f e c t s t a t i s t i c a l l y i s o l a t e d from that of a d v e r t i s i n g . Note that t h i s design would s t i l l recognize, both v a r i a b l e s as "investments", f o r i n the w r i t e r ' s view, there i s no other type of business a c t i v i t y i n which the term " f r a n c h i s e b u i l d i n g " i s more a p p l i c a b l e than to the modern automobile dealer'. CHAPTER V INTRODUCTION This section presents a s t a t i s t i c a l analysis of the automobile sales and advertising data outlines i n Chapter I I I . i Only the more interesting results are discussed, and where possible, implications for marketing strategy are suggested. THE CORRELATION MATRIX General Comments An i n i t i a l step i n the analysis i s to present a corr-e l a t i o n matrix for some of the variables, this appears i n Table VTII. The lagged forms of important predictors were not included i n the matrix due to computer programming d i f f i c u l t i e s . 2 However, most correlations can be derived from the R values presented i n the simple regression estimates of succeeding Tables. The variables included i n the matrix are: 1) UNITS - the number of automobiles sold eachuveek. 2) SALES - weekly d o l l a r r e t a i l sales volume. 3) TOTAD - dealer's t o t a l weekly advertising expenditure. 4) BROAD - dealer's weekly expenditure on radio and t e l e v i s i o n . 5) RADIO - dealer's weekly expenditure on radio. 6) NEWS $ - dealer's weekly newspaper advertising expenditure 7) T V $ - dealer's weekly expenditure on t e l e v i s i o n . 8) DLNWL - dealer's weekly newspaper advertising lineage. 9) MNNWL - manufacturer's weekly newspaper advertising lineage. 10) MAN $ - manufacturer's weekly newspaper advertising expenditure. 11) AVEDI - average distance t r a v e l l e d by buyers each week. 12) MEDIS - median distance t r a v e l l e d by purchasers each week The relationships i n the cor r e l a t i o n matrix support the TABLE VIII CORRELATION MATRIX - DEALER'S SALES ADVERTISING AND DISTANCE UNITS SALES TO TAD NEWS $ BROAD RADIO T V $ DLNWL MNNWL MAN $ AVEDI MEDIS UNITS 1 . 0 0 SALES . 9 5 1 . 0 0 TO TAD . 3 2 . 2 6 1 . 0 0 NEWS $ . 4 9 . 4 0 . 7 7 1 . 0 0 BROAD - . 2 0 - . 2 0 . 5 6 . 0 1 1 . 0 0 -* RADIO - . 1 4 - . 1 3 . 5 5 . 0 7 . 8 9 1 . 0 0 T V $ - . 2 0 - . 2 1 . 2 7 - . 0 8 . 6 5 . 2 5 1 . 0 0 DLNWL . 4 9 . 4 2 . 7 6 . 8 2 . 0 1 . 0 4 - . 0 5 1 . 0 0 MNNWL - . 1 1 - . 1 5 - . 3 0 - . 2 2 - . 2 6 - . 1 2 - . 3 7 - . 1 5 1 . 0 0 MAN $ - . 1 0 - . 1 4 - . 2 9 - . 2 2 - . 2 7 - . 1 3 - . 3 7 - . 1 3 . 9 9 1 . 0 0 AVEDI . 2 5 . 2 5 . 2 1 . 2 5 - . 0 2 . 1 1 - . 2 4 . 2 1 . 2 6 . 2 7 1 . 0 0 MEDIS . 3 1 . 3 0 . 1 2 . 2 3 - . 1 4 - . 0 6 - . 1 9 . 1 6 . 1 4 . 1 5 . 6 6 1 . 0 0 67 regression equations presented in the following section. The only unanticipated finding i s the negative relationship between the sales variables and all. the advertising factors except the dealer's t o t a l advertising, newspaper lineage and newspaper expenditure variables. This negative relationship i s probably due to the infreguency with which these media were employed. The P r e c i p i t a t i o n Variable After the correlation matrix was computed, i t became obvious that many other factors influence r e t a i l sales. In periods of high p r e c i p i t a t i o n consumers may confine their shopping to immediate needs and may forego extensive shopping u n t i l weather conditions improve. Because data were readily available, p r e c i p i t a t i o n was introduced as an explanatory-variable. P r e c i p i t a t i o n must be viewed as a random shock. At no' time did the parameter explain more than one percent of the weekly sales v a r i a t i o n . This finding contrasts with the known sales pattern of the dealership. In aggregating d a i l y r a i n f a l l figures to a weekly basis, the r e a l influence of p r e c i p i t a t i o n i s d i s t o r t e d . Consequently, the factor was not considered further. 68 SALES AND ADVERTISING General Comments As a point of departure, i t would be useful to determine how the case dealer's sales compared to those of a l l Vancouver dealers. Weekly data summarizing automobile sales within the metropolitan area were not available. However, one measure of the dealer's r e l a t i v e performance i s a percentage comparison of his monthly sales to a l l r e t a i l automobile sales i n the Vancouver market. TABLE IX 1 PERCENTAGE DISTRIBUTION OF ANNUAL AUTOMOBILE SALES ALL GREATER CASE ' M O N T H - VANCOUVER DEVIATION DEALERSHIPS DEALER - % • % % January 7.3 6.4 .9 February 8.4 6.9 - 1.5 March 8.7 8.7 0 A p r i l 8.6 7.7 .9 May 8.2 9.1 .9 June 9.4 8.2 - 1.4 July 9.0 7.8 * - 1.2 August 9.3 12.4 3.1 September 6.9 6.9 0 October 7.5 8.1 .6 November 8.2 7.3 .9 December 8.5 10.5 2.0 The deviations i n Table IX seem important. As an example the dealer' s sales were r e l a t i v e l y higher than those of i t s competitors during August and December. Is i t mere chance to f i n d that during these two months, the dealer's advertising expenditures reached their highest peaks? ] Newspaper Expenditure . Radio Expenditure T e l e v i s i o n Expenditure WEEKS FIGURE 7 DEALER'S MEDIA MIX 70 Media a l l o c a t i o n i s an integral part of the advertising decision. The dealer's media mix is presented i n Figure 7. If his advertising strategy i s t y p i c a l of most automobile r e t a i l e r s , then the media mix i s a r e l a t i v e l y constant phenomenon. It has also changed l i t t l e over time. A 1961 study concluded that, in the opinion of automobile r e t a i l e r s , newspapers were the "most e f f e c t i v e medium". TABLE X AUTOMOBILE DEALER'S MEDIA CHOICE (SOURCE: ADAPTED FROM THE UNIVERSITY OF NEVADA, "THE EXTENT OF RETAIL ADVERTISING AS A MANAGEMENT TOOL-ITS SCOPE AND IMPORTANCE IN SMALL BUSINESS," WASHINGTON: SMALL BUSINESS ADMINISTRATION,1961, Pp.73-77. (MIMEOGRAPHED) MEDIUM FIRST SECOND THIRD Newspapers 29 9 7 Radio — 15 10 Tele v i s i o n 14 ,. 10 4 Directory 5 7 8 Is t h i s concentration on newspaper advertising j u s t i f i e d ? Multiple regression analysis should shed some l i g h t on which media aire best related to the sales c r i t e r i o n . In studying the sales effects of advertising, the analyst hopes to es t a b l i s h a causal r e l a t i o n s h i p . One might f i r s t examine the concomitant v a r i a t i o n hypothesizing t h a t a higher l e v e l of sales should be observed in the presence of adver-t i s i n g than in i t s absence. This i s a simple postulate to 71 prove, but interpretation i s hazardous. Figure 8 depicts the dealer's aggregated sales and advertising curves. Disregarding the many explanations for a moment, i t i s apparent that the amplitude of the advertising cycle mirrors that of the sales cycle i n eight of the twelve months. From May to September, the better months for automobile marketing, the sales and advertising curves appear correlated. That promotional expend-itures tend to r e f l e c t sales rates p a r t i a l l y explains this f i n d i n g . J F M A M J J A S O N D MONTH FIGURE 8 DEALER'S MONTHLY SALES AND ADVERTISING CYCLE When the advertising cycle i s lagged one month, the two curves appear to be fluctuating i n an t i t h e s i s . Diagnosis i s d i f f i c u l t . Perhaps sales are completely i n e l a s t i c to 72 advertising pressure—however, i n t u i t i v e judgment would suggest a more erudite explanation. I t seems plausible that the cumulative effects of dealership advertising decay very rapidly and a f t e r one month approach zero. This hypothesis accords with theory: promotion conveying "sale" and "action" i s known to decay quickly. I f this i s true, continuous advertising i s necessary. I t i s equally possible that a month is too lengthy a period in which to analyze the sales effects of dealership advertising. Are weekly data more sensitive? In Figure 9, the r e l a t i v e change in weekly sales i s plotted on the v e r t i c a l scale The ordinate represents the percentage change in advertising expenditures during the same period. Because the points are spread out i n Panel A, the sales-advertising relationship i s not strong. As Panel B demonstrates, lagging promotional expenditures one week reduces the scatter. The apparent d i s t r i b ution of points from lower l e f t to upper right evidences the existence of some positive c o r r e l a t i o n . Since the* clustering indicates no curvature, a linear r e l a t i o n s h i p between sales and advertising i s assumed. Unfortunately a simple co r r e l a t i o n c o e f f i c i e n t r e f l e c t i n g the association between advertising and sales does not d i s t i n g -uish cause and e f f e c t . If advertising i s to stimulate sales i t must c l e a r l y precede the actual sale. In Figure 10 there i s some evidence to suggest that increased advertising during week t resulted i n heightened sales the following week. However, c H A N G E I N V E E K L Y s A 1 . 0 0 L- -E . 7 5 S .. A • 50 s . A . 2 5 0 0 F - . 2 5 A N N - . 5 0 U - . 7 5 A L - K O P j 73 # PANEL A: S a l e s $ and A d v e r t i s i n g -1.50 -1.00 -.50 .50 1 . 00 1.50 CHANGE I N WEEKLY ADVERTISING AS PERCENTAGE OF ANNUAL 1,. rlQ -1-00° --r-o S 1 . 0 0 c A H A L E • 75 N " S G A • 50 E A s " " ' 2 , I N • 25 ¥ 0 0 E • r _ E A - . 2 5 K N L N - . 5 0 Y U .-A L - . 7 5 - i - onl ' # PANEL B: • « S a l es $ etrici Advert i s ing" *$ FIGURE 9 REL A T I V E CHANGE I N . SALES VERSUS RELATIVE CHANGE I N ADVERTISING S a l e s A d v e r t i s i n g ^ Change S a l e s i n Week t / T o t a . l Y e a r l y S a l e s A d v e r t i s i n g i n Week t / T o t a l Y e a r l y A d v e r t i s i n g S a l e s $ t - S a l e s $ f c_ 1 ; A d y t t - A d v t t _ ] 74 g r a p h i c r e s u l t s are not easy to i n t e r p r e t , t h e r e f o r e m u l t i p l e r e g r e s s i o n was used f o r f u r t h e r a n a l y s i s . TIME FIGURE 10 DEALER'S WEEKLY SALES AND ADVERTISING CYCLE S a l e s and T o t a l A d v e r t i s i n g Ten percent of u n i t s a l e s v a r i a t i o n was e x p l a i n e d by the r e t a i l e r ' s t o t a l a d v e r t i s i n g expenditure, as shown i n Table X I . Is a b e t t e r f i t to the data given when t o t a l adver-t i s i n g i s lagged one week? Regression r e s u l t s are i n c o n c l u s i v e — PREDICTOR TABLE XI (RESULTS : UNITS SOLD CORRELATED WITH-CONSTANT REGRESSION COEFFICIENT t-VALUE R" F-RATIC (1) Total Advertising 31.28 .006 2.43 # .10 5.92 * (2) Total Advertising t-1 (3) Dealer Newspaper Lineage 30.93 29.22 (4) Dealer Newspaper Lineage t-1 28.22 006 oo: .003 2.81 9 4.02 3 5.06 9 13 24 7.94 # 16.23 C 25.68 $ # s i g n i f i c a n t at <: s i g n i f i c a n t at * s i g n i f i c a n t at @ s i g n i f i c a n t at .01 l e v e l 001 l e v e l .05 l e v e l 005 l e v e l 1 TABLE XII PREDICTOR RESULTS : SALES DOLLARS CORRELATED WITH -CONSTANT REGRESSION COEFFICIENT t-VALUE F-RATIO (1) Total Advertising (2) Total Advertising t-1 (3) Dealer Newspaper Lineage (4) Dealer Newspaper Lineage t-1 (5) Average Price 87 .86 90.47 89.57 84.07 2.67 .025 .024 .009 .011 .002 3.14 {§> 3.15 @ 3.31 @ 4.57 @ 2.53 # .16 .16 .17 .29 .11 9.87 # 9.93 # 10.96 # 20.97 $ 6.42 * # s i g n i f i c a n t at C s i g n i f i c a n t at * s i g n i f i c a n t at (a> s i g n i f i c a n t at .01 l e v e l 001 l e v e l .05 l e v e l ,005 l e v e l 7? the value of R moves from .10 to .13, hardly a s i g n i f i c a n t increase. The low c o e f f i c i e n t of determination i s probably caused by c o l l i n e a r i t y between the components of the t o t a l advertising appropriation. I f the dealer allocates more do l l a r s to radio advertising, he usually spends less on news-papers. The co r r e l a t i o n c o e f f i c i e n t s support t h i s conclusion. Sales and Dealer Newspaper Advertising The existence of cumulative advertising effects i s i l l u s t r a t e d by the dealer's newspaper lineage variable (Table XI, equations three and four). While 24 percent of the unit sales v a r i a t i o n i s explained by the coincidental newspaper advertise-ments , 34 percent i s explained when the variable i s lagged one period. Both regression equations are s i g n i f i c a n t at the .0001 l e v e l . The dealer newspaper expenditure variable, lagged one week, was not as good a predictor of sales as dealer newspaper lineage lagged one period (R —.14 versus .29). The higher p. correlations for lineage probably r e f l e c t the influence of p r i c e — t h e expenditure variable i s a conglomeration of various rates per agate l i n e . The lineage factor includes t h i s price d i f f e r e n t i a l while the expenditure variable does not. SALES AND DISTANCE Sales are strongly oriented to the region immediately surrounding the dealership. However, a s i g n i f i c a n t amount of patronage was received from transients, that i s , customers who 78 purchased automobiles beyond t h e i r normal trading areas. The magnitude of this transient trade confirms that Vancouver i s the automobile r e t a i l i n g centre of B r i t i s h Columbia. TIME FIGURE 11 ADVERTISING EXPENDITURE AND OUT-OF-TOWN SALES Although the transient sales curve roughly mirrors the advertising curve i n Figure 11, there are important deviations. The most notable exception occurs during the l a t t e r weeks of August—whereas advertising remained almost constant, out-of-town sales rose dramatically. This phenomenon r e f l e c t s an annual recreational event near the dealership, which doubtless draws thousands of people past the showroom. There i s a positi v e c o r r e l a t i o n between distance and s a l e s — t h e average distance t r a v e l l e d by consumers t y p i c a l l y increases as advertising expenditures r i s e . This i s the 7 9 implication drawn from Figure 12. Median distance, being less affected by extremeties, remains s u r p r i s i n g l y constant through-out the year. S p e c i f i c s t a t i s t i c s appear in Tables XIII and XIV. Simple co r r e l a t i o n c o e f f i c i e n t s between the distance and sales variables were t y p i c a l l y about .25. Although correlations between advertising and distance data are also very low, i t i s impossible to conclude that promotion does not " p u l l " consumers from a wide geographical area. Units/ Miles TIME FIGURE 12 UNIT SALES, MEDIAN AND AVERAGE DISTANCE ) RESULTS TABLE XIII DISTANCE AND ADVERTISING A. CRITERION - AVERAGE DISTANCE PREDICTOR CONSTANT REGRESSION COEFFICIENT t-VALUE R F-RATIO (1) Total Advertising 16.26 .0027 1.52 04 2.32 (2) Dealers Newspaper Lineage 16.87 ,0009 1.58 .04 2.50 (3) Dealers Newspaper Lineage t-1 17 .98 0007 1.46 02 1.31 B. CRITERION - MEDIAN DISTANCE PREDICTOR CONSTANT REGRESSION COEFFICIENT t-VALUE R F-RATIO (1) Total Advertising (2) Dealers Newspaper Lineage 9.85 9,78 0003 .0001 .86 1.20 01 ,02 .75 1,45 TABLE XIV RESULTS : SALES AND DISTANCE A. CRITERION - UNITS SOLD PREDICTOR CONSTANT REGRESSION COEFFICIENT t-VALUE R F-RATIO Average Distance 32 .52 369 1.85 * 06 3.44 --Median Distance 19.89 1.932 1.13 09 5.53 * B. CRITERION - SALES DOLLARS PREDICTOR CONSTANT REGRESSION COEFFICIENT t-VALUE R F-RATIO Average Distance Median Distance 57 .88 2 .67 5.893 .002 2.22 * 2.53 # 09 11 4.94 * 6.42 * # s i g n i f i c a n t at C s i g n i f i c a n t at * s i g n i f i c a n t at @ s i g n i f i c a n t at .01 l e v e l .001 l e v e l .05 le v e l .005 l e v e l 82 SALES AND PRICE ' L i s t price i s an unsatisfactory unit i n which to measure the dealer's weekly sales volume. A given automobile's r e t a i l price varies as a function of: i) the customer's willingness and a b i l i t y to pay, i i ) the accessories he chooses and i i i ) the commissioned salesman's a b i l i t y to exact a higher p r i c e . The many models offered by the dealer assures a continuum of prices. Weekly sales data should generally display an inverse relationship between average price and units sold: i f average price declines, unit sales should increase. In the regression analysis, p o s i t i v e c o e f f i c i e n t s for average price contradict economic theory. However, a number of factors may be operating i n the market to explain this s i t u a t i o n . For instance, during periods of heightened marketing a c t i v i t y , intense "trading-up" by consumers w i l l cause a d i r e c t relationship between average price and units sold. When new models are introduced, prices are r e l a t i v e l y high, but so are sales. I f a p a r t i c u l a r model i s i n heavy demand, there w i l l be a d i r e c t c o r r e l a t i o n between price and sales. As a further example, price may be lowered to stimulate s a l e s — i f consumers do not react quickly, or are ' influenced by other patronage motives, unit sales w i l l not increase substantially to assure the inverse relationship. Price i s known to be an omnipotent factor in 1 automobile buying behavior. Approximately eleven percent of the weekly sales volume can be explained by differences in the average / ' 1 . ' 1 i • 83 price of the units sold. In Table XII, equation f i v e , average price i s a very s i g n i f i c a n t predictor of sales (t 2.53). Although, correlations between average price and sold units are low, t h i s probably r e f l e c t s the averaging process. OTHER PREDICTIVE EQUATIONS In the multiple regression equations, low t-values^ahd negative c o e f f i c i e n t s were consistently observed for the dealer' broadcast expenditure variables. I f radio and t e l e v i s i o n variables are added to the equations of Table XV, the percentage of explained variance does not improve. Unfortunately, zero advertising expenditures on broadcast media during many weeks makes present evaluation of radio and t e l e v i s i o n impossible. A low l e v e l of prediction was also associated with the manufacturer's newspaper advertising variable. Appendix A i l l u s t r a t e s the poor performance of equations containing this factor. Removing the e f f e c t of the dealer's newspaper adver-t i s i n g from each regression in Table XVI, the insi-gnificant p a r t i a l c o r r e l a t i o n c o e f f i c i e n t s show that almost none of the dealer sales variance 'can be attributed to the variance i n the manufacturer's l o c a l advertising expenditure. In fact, Table XVTII suggests there was l i t t l e c o r r e l a t i o n between the manufacturer's sales and his l o c a l newspaper advertising p o l i c y . L i t t l e coordination was also observed between the advertising appropriations of franchisor and franchisee. However, th i s analysis i s incomplete as the manufacturer's national advertising figures could not be secured. TABLE XV RESULTS : SALES VOLUME REGRESSED ON -PREDICTOR CONSTANT REGRESSION COEFFICIENT t-VALUE PARTIAL CORRELATION COEFFICIENT 9 R~ F-RATIO (1) Dealers News. Lineage t - l . 010 4.69 @ .564 Median Distance -48.91 1 .440 .58 .085 .40 10.51 C Average Price 39.653 2.77 @ .375 (2) Dealers Lineage t - l .010 4.7 3 ^ .568 Average Distance -42.78 .476 .92 . 133 .40 10.97 C Average Price 39.478 2.78 @ . 375 (3) Dealers Average Lineage t - l Price -34.31 .011 39.510 4.95 @ 2.78 @ .581 . 373 .39 15.81 $ # s i g n i f i c a n t at C s i g n i f i c a n t at * s i g n i f i c a n t at (3) s i g n i f i c a n t at .01 l e v e l .001 le v e l .05 l e v e l .005 le v e l CO TABLE XVI RESULTS : UNITS SOLD REGRESSED ON -PREDICTOR CONSTANT REGRESSION COEFFICIENT t-VALUE PARTIAL CORRELATION COEFFICIENT F-RATIO (1) Dealers News. Lineage t-1 .003 4.74 .569 Median Distance 21.94 ( .498 .62 .091 . 34 8 .41 # Average Price .462 .00 .014 (2) Dealers News. Lineage t-1 V .003 4.7 9 .573 Average Distance 24.44 .143 .84 .122 .35 8 .50 # Average Price i .403 .08 .012 (3) Dealers News. Lineage t-1 .003 5.01 @ .586 Average Price 26.99 .412 .08 .012 . 34 12 .58 # s i g n i f i c a n t at C s i g n i f i c a n t at * s i g n i f i c a n t at !3> s i g n i f i c a n t at .01 l e v e l .001 l e v e l .05 l e v e l .005 l e v e l CD 86 SEVERAL VARIABLES AS PREDICTORS OF SALES VOLUME Successive elimination of uninteresting results reveals that the more s i g n i f i c a n t multiple regression equations contain the p r i c e , distance and lagged dealer newspaper lineage var-i a b l e s . The best predicting equation i s that represented by regression three, Table XV. Thirty-three percent of the variance i s explained by the dealer's lagged newspaper lineage, variable and almost fourteen percent by the average p r i c e . The pr o b a b i l i t y of obtaining an F - r a t i o as great as 15.81 i f the two parameters were random occurrences i s less than .001. Consequently, we must conclude that dealer newspaper lineage and price have an e f f e c t on sales volume. Of course, t h i s leaves 60 percent of the variance to be explained by a l l other factors. I f an additional factor, namely average distance i s added to equation three (Table XV), i t may help account for part of the variation' previously unexplained. As equation two, Table XV shows, the addition of the distance variable has not given a much better f t t . S i m i l a r l y , no other factor examined in the regression analysis improved the s t a t i s t i c s of equation three. ' SUMMARY AND CONCLUSIONS Remembering that conclusions based on regression analysis are tentative, here i s a summary of the major findings of th i s exploratory study. 87 Sales and Dealer Newspaper Lineage The automobile dealer appropriated most advertising dollars to newspapers. Regression estimates indicate that sales volume i s s i g n i f i c a n t l y related to the newspaper lineage variable. The behavior of the lineage variable also confirms the existence of cumulative advertising e f f e c t s — c o n s i s t e n t improvements in s t a t i s t i c a l measures res u l t when this variable i s lagged one week. Greater lags do not improve predictive a b i l i t y . I f cum-ul a t i v e advertising effects are s l i g h t , a p o l i c y of continuous advertising seems appropriate. Distance Two axioms of marketing geography are substantiated in the study: i) other things equal, people prefer to purchase automobiles near th e i r residences ,i_f they are able to do so? i i ) as the s p a t i a l d i s t r i b u t i o n of customers increases, the greater must be the a t t r a c t i o n to p u l l transient trade. At higher levels of advertising pulsation, the average distance t r a v e l l e d by automobile purchasers t y p i c a l l y increases. That median distance remains surp r i s i n g l y constant suggests existence of a time-distance l i m i t , beyond which consumers are unwilling to t r a v e l . Weekly o s c i l l a t i o n s in out-of-town sales also suggest: that such patronage i s related more to holidays and other events than i t i s to advertising. In t h i s instance, r e t a i l location i s an important determinant of sales. 88 Price Throughout the year, the average price of an automobile approximated $3000. This figure does fluctuate from week to week, crudely r e f l e c t i n g the actual purchase price of each model. Yet the addition of average price to the multiple re-gression equations reduces the unexplained variance s i g n i f -i c a n t l y . Positive c o e f f i c i e n t s for average price are disturbing, but many extraneous influences are probably operating. Other Variables Inadequate data gathering and measuring procedures probably reduced the usefulness of other variables subjected to regression analysis. I n t u i t i v e reasoning would imply that weather i s related to automobile sales—aggregating d a i l y r a i n f a l l data no doubt mars this influence. Because radio and t e l e v i s i o n were used infrequently, this paper i s unable to ascertain whether or not the dealer's broadcasting variables are c r u c i a l determinants of sales. F i n a l l y , the manufacturer's "r. 9 l o c a l newspaper advertising apparently bears l i t t l e relationship to his sales or those of the dealer. However, this finding is speculative in that i t neglects the contribution of the franchisor's national advertising campaign. Further Research The findings of this study are interesting as a confirm-ation of a p r i o r i reasoning. Models of the advertising process must consider many more variables i f the sales effects of 89 advertising are to be measured. Chapter IV outlined some of these factors which may apply to the automobile r e t a i l i n g s i t u a t i o n . Regression estimates on one year's data are hardly c o n c l u s i v e — t h e sales response to advertising i s complex and dynamic. More research on the unique data of many firms i s needed df advertising axioms are to be supported. Only then w i l l advertising theory grow from i t s present f e t a l stage to maturity. Only then w i l l the findings of advertising research have general a p p l i c a b i l i t y . 90 BIBLIOGRAPHY I i BIBLIOGRAPHY A. BOOKS Alderson Associates, Inc. A Basic Study of Automobile R e t a i l i n g . Dearbron, Michigan: The Ford Motor Company, 1958. Bogart, Leo. Strategy i n Advertising. New York: Harcourt, Brace and World, Inc., 1967. Borden, N e i l H. The Economic Ef f e c t s of Advertising. Homewood, I l l i n o i s ! Richard D. IrwTn, Inc., 1952. Colley, Russel H. (ed.). Defining Advertising Goals for Measured Advertising Results. New York: Association of National Advertisers, Inc., 1961. Cri s p , Richard D. How To Increase Advertising Effectiveness. Consultant Reports on Current Business Problems. New York: McGraw-Hill Book Company, 1959. Davies, Owen L. (ed.). S t a t i s t i c a l Methods i n Research and  Production. London: Oliver and Boyd, Limited, 1947. Ez e k i e l , Mordecai and Karl A. Fox. Methods of Correlation and Regression Analysis. New York: John Wiley and Sons, Inc., 1959. Ferber, Robert and P. J . Verdoon. Research Methods i n Economics and Business. New York: The MacMillan Company, 1962. Firestone, 0. J . The Economic Implications of Advertising. Toronto: Methuen Publications, 1967. Frey, Albert W. How Many Dollars For Advertising. New York: Ronald Press Company, 1955. Gallop, George. "Measuring Advertising's Sales Effectiveness: The Problem and Prognosis," i n Evelyn Konrod and Rod Erickson (eds.). Marketing Research—A Management Overview. New York: American Management Association, 1966. Pp. 145-151. Howard, John A. Marketing Management—Analysis and Planning. Homewood, I l l i n o i s : Richard D. Irwin, Inc., 1963. King, William R. Quantitative Analysis For Marketing Management. New York: McGraw-Hill Book Company, 1967. 92 McCarthy, E. Jerome. Basic Marketing—A Managerial Approach. Honewood, I l l i n o i s ! Richard D. IrwTn, Inc., 1964. Nelson, Richard L. The Selection of R e t a i l Locations. New York: F. W. Dodge Corp., 1958. Palda, K r i s t i a n S. The Measurement of Cumulative Advertising  E f f e c t s . Englewood C l i f f s , New Jersey: Prentice-Hall, Inc., 1964. Robinson, Patrick J . and David J . Luck. Promotional Decision  Making—Practice and Theory. New York: McGraw-Hill Book Company, 1964. Starch, Daniel. Measuring Advertising Readership and Results. New York: McGraw-Hill Book Company, 1966. Smith, G a i l . "How GM Measures Ad Effectiveness," i n Keith K. Cox (ed.). Readings In Market Research. New York: Appleton-Century Cr o f t s , 1967. Pp. 170-180. Tinter, Gerhard. Econometrics. New York: John Wiley and Sons, Inc., 1952. T u l l , Donald S. "A Re-examination of the Causes i n the Decline i n Sales of Sapolio," in H. G. Barksdale (ed.). Marketing i n Progress—Patterns and Potentials. New York: Holt, Rhinehart and Winston, Inc., 1964. Pp. 240-252. Weinberg, Robert S. An A n a l y t i c a l Approach to Advertising Expenditure Strategy. New York: Association of National Advertisers, 1960. Wolfe, Harry D., James K. Brown and G. Clarke Thompson, Measuring Advertising Results. Studies i n Business Policy No. 102. New York: National I n d u s t r i a l Conference Board, 1962. B. PERIODICALS Alderson, Wroe. "Measuring the Sales Effectiveness of A d v e r t i s i n g — A Progress Report," Proceedings: Fourth  Annual Conference of the Advertising Research Foundation. New York: Advertising Research Foundation, Inc., 1958. Pp. 93-99. Applebaura, William. "Methods for Determining Store Trade Areas, Market Penetration and Potential Sales," Journal of  Marketing Research. 3:127-141, May, 1966. 93 Bachelder, J . "How Many Dollars For Advertising," Dun's  Review and Modern Industry, 72:54-5, October, 1958. Banks, Seymour. "Some Correlates of Coffee and Cleanser Brand Shares," Journal of Advertising Research, 1:1-10, March, 1961. Becknell, James C. and Robert W. Mclsaac. "Test Marketing Cookware Coated with Teflon," Journal of Advertising  Research, 3:2-8, September, 1963. Benesch, Charles. "Sales Tests As A Measure Of Advertising Effectiveness," Journal of Marketing, 17:178-81, October, 1952. Berreman, Joel V. "Advertising and the Sale of Novels," Journal of Marketing, 7:234-41, January, 1943. Brown, George H. and Frank A. Mancina. "A Note on the Relationship Between Sales and Advertising of Department Stores," Journal of Business, 13:1-16, January, 1940. Buzzell, Robert D. "Operation Research for Measurement of Advertising Effectiveness—The F i r s t Five Years," Proceedings Ninth Annual Conference of the Advertising  Research Foundation. New York: Advertising Research Foundation, Inc., 1963. Pp. 78-83. Carrick, Paul M. "Why Continued Advertising Is Necessary," Journal of Marketing, 23:386-98, Ap r i l , 1959. Coffin, Thomas C. "A Pioneering Experiment in Assessing Advertising Effectiveness," Journal of Marketing, 27:1-10, July, 1963. Cover, John H., et a l . "Department Stores' Sales and Advertising," Journal of Business, 4:227-44, July, 1931. Cowan, Donald R. G . "Sales Analysis," Journal of Business, 6:15-35, January, 1933. Cox, Donald F. "Clues for Advertising Strategists II," Harvard Business Review, 39:160-82, November-December, 1961. Dean, Joel. "Does Advertising Belong in the Capital Budget," Journal of Marketing, 30:15-21, October, 1966. . "How Much to Spend for Advertising," Harvard Business Review, 32:120-30, January-February, 1954. 94 Dickens, Dorothy. "Advertising Dairy Products i n Rural Grocery Stores," Journal of Marketing, 19:268-70, January, 1955. Dorfman, Robert and Peter 0. Steiner. "Optimal Advertising and Optimal Quality," American Economic Review, 44:826-36, December, 1954. Eldridge, Clarence E. "Advertising Effectiveness—How Can I t Be Measured," Journal of Marketing, 22:241-51, January, 1958. Feder, Richard A. "How to Measure Marketing Performance," Harvard Business Review, 43:132-42, May-June, 1965. Forrester, J . W. "Advertising: A Problem i n I n d u s t r i a l Dynamics," Harvard Business Review, 37:100-10, March-A p r i l , 1959. Green, Paul E. "Bayesian Decision Theory i n Advertising," Journal of Advertising Research, 5:15-20, June, 1965. Ealbert, Michael H. "A P r a c t i c a l and Proven Measure of Advertising Effectiveness," Proceedings:Sixth Annual  Conference of the Advertising Research Foundation. New York: Advertising Research Foundation, Inc., 1960. Pp. 77-82. Henderson, Peter L. , e_t a l . "Sales Effects of Two Campaign Themes," Journal of Advertising Research, 1:2-11, December, 1961. Hollander, Sydney A. "A Rationale for Advertising Expenditures," Harvard Business Review, 27: 7 9-87, January, 1949. Hoofnagle, William S. "The Effectiveness of Advertising for Farm Products," Journal of Advertising Research, 3:2-10, December, 1963. Jastram, Roy W. "Advertising Ratios Planned by Large-Scale Advertisers," Journal of Marketing, 14:13-21, July, 1949. Jenssen, Ward J . "Sales E f f e c t s of TV, Radio and P r i n t Advertising," Journal of Advertising Research, 6: 2-5, June, 1966. Jessen, R. J . "A Switch-Over Experimental Design to Measure Advertising E f f e c t , " Journal of Advertising Research , 1:15-22, March, 1961. K o l l i n e r , Sim A. "New Evidence of Ad Values," I n d u s t r i a l  Marketing, 48:81-4, August, 1963. 95 Kuehn, A l f r e d A. "How Advertising Performance Depends on Other Factors," Journal of Advertising Research, 2:2-10, March, 1962. , and Albert C. Rohloff. " F i t t i n g Models to Aggregate Data," Journal of Advertising Research, 7:42-7, March, 1967. , Timothy W. McGuine and Doyle L. Weiss. "Measuring the Effectiveness of Advertising," i n Raymond M. Haas, (ed.), Science Technology and Marketing. F a l l Conference Proceedings of the American Marketing Association. Bloomington, Indiana: American Marketing Association, 1966, Pp. 185-194. Lavidge, Robert and Gary Steiner. "A Model for Predictive Measurements of Advertising Effectiveness," Journal of  Marketing, 25:59-62, October, 1961. L i t t l e , John D. C. and Russel Ackoff, "How Techniques of Mathematical Analysis Have Been Used to Determine Advertising Budgets and Strategy," Proceedings: Fourth Annual  Conference of the Advertising Research Foundation. New York: Advertising Research Foundation, Inc., 1958, Pp. 19-23. Marantz, Marcel. "Evaluating Department Store Advertising," Journal of Advertising Research. 7:16-21, March, 1967. Marchner, Donald C. "Theory Versus Practice i n A l l o c a t i n g Advertising Money," Journal of Business, 40:286-302, Ju l y , 1967. Martineau, P i e r r e . "Automobiles: What They Mean to Americans," Michigan Business Papers No. 30, Ann Arbor, Michigan: University of Michigan, 1954, Pp. 24-36. May, Frederick E. "Buying Behavior: Some Research Findings," Journal of Business, 38:379-95, June, 1965. . "Adaptive Behavior i n Automobile Brand Choices," Journal of Marketing Research, 6:62-5, February, 1969. Meissner, Frank. "Sales and Advertising of Lettuce," Journal of Advertising Research, 1:1-10, March, 1961. Merriman, Jones F. "Evaluating Advertising Appeals Through Sales Results," Journal of Marketing, 23:164-7, October, 1958. Mueller, H. P. "Should Advertising Budgets Lead or Follow Sales," I n d u s t r i a l Marketing, 44:115-6, September, 1959. Murray, J . Alex. "Canadian Consumer Expectational Data: An Evaluation," Journal of Marketing Research, 5:54-61, February, 1969. Nerlove, Marc and Kenneth J . Arrow. "Optimal Advertising P o l i c y Under Dynamic Conditions," Economica, 39:129-42, May, 1962. , and Frederick V. Waugh. "Advertising Without Supply Control: Some Implications of a Study of the Advertising of Oranges," Journal of Farm Economics, 43:813-37, October, 1961. Palda, K r i s t i a n S. "The Hypothesis of a Hierarchy of E f f e c t s : A P a r t i a l Evaluation," Journal of Marketing Research, 3:13-24, February, 1966. : . "Sales E f f e c t s of Advertising," Journal of Advertising Research, 4:12-16, September, 1964. . "The Evaluation of Regression Results," i n Stephen A Greyser (ed.). Toward S c i e n t i f i c Marketing: Proceedings  of the Winter Conference of the American Marketing  Association. Boston; American Marketing Association, 1963, Pp. 279-290. Purbis, L. E. "Activation Research—A New Development i n the Analysis of Influences on Purchase," i n Robert Hancock (ed.). Proceedings of the 43rd. National Conference of the  American Marketing Association. Minneapolis, Minnesota: American Marketing Association, 1960, Pp. 349-370. Rasmussen, Arne. "The Determination of Advertising Expenditures," Journal of Marketing, 16:439-46, A p r i l , 1952. Raymond, Charles K. "Are We Downhearted," Journal of  Advertising Research, 6:64, September, 1966. Roberts, Harry V. "The Management of Advertising Results," Journal of Business, 20:131-45, July, 1947. Salveson, Melvin E. "Management's C r i t e r i a for Advertising Effectiveness," Proceedings: F i f t h Annual Conference of the Advertising Research Foundation. New York: Advertising Research Foundation, Inc., 1959, Pp. 23-28. Schoenberg, E. H. "The Demand Curve for Cigarettes," Journal of Business, 6:15-35, January, 1933. Semon, Thomas T. "Assumptions i n Measuring Advertising Effectiveness," Journal of Marketing, 28:43-4, July,1964. Simon, J u l i a n L. "Are There Economies of Scale i n Advertising?" Journal of Advertising Research, 5:15-20, June, 1965. , and George H. Crain. "The Advertising Ratio and Economies of Scale," Journal of Advertising Research, 6:37-43, September, 1966. Telser, Lester G. "Advertising and Cigarettes," Journal of  P o l i t i c a l Economy, 70:471-99, October, 1962. Tousley, Raymond. "Advertising Fresh F r u i t s and Vegetables I I , " Harvard Business Review, 23:79-94, Autumn, 1944. V a i l e , Roland S. "Use of Advertising During Depressions," Harvard Business Review, 5:323-30, A p r i l , 1927. Vidale, M. L. and H. B. Wolfe. "An Operations-Research Study of the Sales Response to Advertising," Operations Research, 5:370-81, June, 1957. Wagner, Louis C. "Advertising and the Business Cycle," Journal of Marketing, 6:124-35, October, 1941. Wallace, James M. "A Perfect Measure of Advertising's Contribution to Marketing," Journal of Marketing, 30:57, J u l y , 1966. Weiss, Doyle L. "Determinants of Market Share," Journal of  Marketing Research, 5:290-5, August, 1968. Wirth, Edward D. "Developing Marketing Strategy Through Multip l e Regression," Journal of Marketing Research, 4:318-9, August, 1967. C. UNPUBLISHED MATERIALS Advertising Research Foundation, Inc. Minutes of the second meeting of the Operations Research Discussion Group, February 16, 1960, at Advertising Research Foundation, Inc., Headquarters, New York. (Mimeographed). . Minutes of the t h i r d meeting of the Operations Research Discussion Group, A p r i l 22, 1960, at Advertising Research Foundation, Inc., Headquarters, New York. (Mimeographed). Clement, Wendell E. "An Analysis of the Advertising Process and i t s Influence on Consumer Behavior," Paper presented at the 1968 F a l l Conference Proceedings of the American Marketing Association, Denver, Colorado. (Mimeographed). 98 Loekhart, David C. "An Examination of the Franchise System of D i s t r i b u t i o n , " Unpublished graduating essay, The University of B r i t i s h Columbia, Vancouver, 1968. Miracle, Fordon E . "Measuring the Productivity of Advertising," Unpublished d i s s e r t a t i o n , The University of Wisconsin, 1963. Puffer, Frank. "Multiple Correlation and Regression Program," The University of C a l i f o r n i a at Los Angeles, October, 1964. Revised for zero data by J . D. Forbes, Univeristy of B r i t i s h Columbia, Vancouver, February, 1967. University of Nevada. "The Extent of R e t a i l Advertising As A Management T o o l — I t s Scope and Importance i n Small Business, Washington: Small Business Administration, 1961. (Mimeographed). 99 APPENDIX A TABLE XVII RESULTS : REGRESSION OF UNITS SOLD ON DEALER AND MANUFACTURER ADVERTISING VARIABLES PREDICTOR CONSTANT REGRESSION COEFFICIENT t-VALUE PARTIAL CORRELATION COEFFICIENT R 2 F (1) Dealers T o t a l Advertising .006 2.27 * .309 31.56 .106 2.90 * Manu. News. Expenditure -.000 -.08 -.011 (2) Dealers News. Expenditure .009 3.93 ® .489 27 .92 . 248 8.10 # Manu. News. Expenditure .000 .03 .000 (3) Dealers News. Lineage .003 3.89 @ .486 29.87 .246 8.00 # Manu. News. Lineage -.000 -.27 -.039 (4) Dealers News. Lineage t-1 .003 4.92 @ .579 28.35 .343 12.58 # Manu. News. Lineage t-1 -.000 - .058 -.008 # s i g n i f i c a n t at .01 l e v e l C s i g n i f i c a n t at .001 l e v e l * s i g n i f i c a n t at .05 l e v e l (§> s i g n i f i c a n t at .005 l e v e l 101 TABLE XVIII COMPARISON OF MANUFACTURER'S MONTHLY SALES AND ADVERTISING CYCLES MONTHS % OF MANUFACTURER'S % OF ANNUAL ADVERTISING CARS SOLD BUDGET January 8.52 1.38 February 9.32 2.23 March 7 .52 6.57 A p r i l 7 .65 13.31 May 10.44 12.71 June 8.79 16.74 July 7 .58 13.98 August 7.69 2.21 September 6.44 13.02 October 8.30 4.67 November 9.03 9.89 December 8.72 2 .99 1 0 2 APPENDIX B [ 0 3 PROGRAM USED TO COMPUTE MEDIAN AND AVERAGE DISTANCES DIMENSI^NX(70) ,Y(70) ,XY(70) ,XBAR(70) C X=XC00RD,Y=YC#0RD,SY=C0MPUTED DISTANCE,SBAR=AVERAGE DISTANCE Ka-18 1 P0RMAT(I2,2X,19(2F2.O) ) XD»14 YD= 9 10 CONTINUE RsK+19 L*K\rt8 READ(5,1) IWEK, (X(I) ,Y(I) ,I=K,L) IF (Y(L)-O.) 11,11,10 11 CONTINUE K=-18 C TEST FOR N0. 0F AUT0S WITHIN METR0 AREA L=l 12 CONTINUE IF<X(L)-0.)13,13,14 13 XY (L)= (( (X (L) -XD) **2)+ (Y (L>) -YD) **2) ) **0.5 L=Lt 1 Qff> T0 12 14 CONTINUE C CHECK F0R 0UT 0F T0WN SALES LCAL=L-1 15 CONTINUE IF(Y(L)-0.)17,17,16 16 XY(L)=Y(L) L=L+1 G0 T0 15 16 LT0T-L-1 C HAVE C0MPUTED ALL DISTANCES C N0W RANK 0RDER DISTANCES CALL LS0RT(X,LT0T,) C COMPUTE MEDIAN FT0T=LT0T FTAT« F T 0 T /2 . F T I T - L T 0 T / 2 . NT0T=FTTT IF (FTAT-FTTT) 17 ,17 ,18 17 NMED=NT0T FMED=(XY (NT0T) +• XY (NT0T l ) ) / 2 . G0 TQ) 19 18 FMED=XY (NTOT+1) 19 C0NTINUE FSUM=0. D 0 2 O I=1,LT#T 20 FSUM=FSUM+XY(I) FDIS»FSUM /FT0T C WRITE 0UTPUT 99 F0RMAT(14H00WEEK NUMBER©,12) 100 P0RMAT(15HO©TOTAL SALESO0,F10.5) 101 F0RMAT(19HOOAVERAGE DISTANCE©,F10.5) 102 FORMAT(18H0GMEDIAN DISTANCE©,F10.5) WRITE(6,99) IWEK WRITE (6,100) FT0T WRITE(6,101) FDIS WRITE(6,102) FMED 103 F0RMAT(1H©,19F4.1) WRITE(6,103) (XY(L) ,L l.LTOT G0 Tp 10 ST0P END 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0102341/manifest

Comment

Related Items