Repetitive Advertising and the Consumer

Andrew S. C. Ehrenberg

INTRODUCTION

Advertising is in an odd position. Its extreme protagonists claim it has extraordinary powers and its severest critics believe them. Advertising is often effective. But it is not as powerful as is sometimes thought, nor is there any evidence that it actually works by any strong form of persuasion or manipulation.

Instead, the sequence, awareness/trial/reinforcement, seems to account for the known facts. Under this theory, consumers first gain awareness or interest in a product. Next, they may make a trial purchase. Finally, a repeat buying habit may be developed and reinforced if there is satisfaction after previous usage.

Advertising has a role to play in all three stages. But for frequently bought products, repeat buying is the main determinant of sales volume and here advertising must be reinforcing rather than persuasive.

These conclusions are based largely on studies of consumer behaviour and attitudinal response. They are important both to our understanding of advertising’s social role and to the execution and evaluation of advertising as a tool of marketing management.

In this paper I first examine advertising and the consumption of goods in general. I then discuss competition among brands and the factors affecting consumers’ brand choice, particularly for established brands of frequently bought goods.

The Demand for Goods

Advertising is widely credited with creating consumer demand. Sol Golden was quoted in 1972 in this Journal as saying:

'Advertising is the lynch-pin by which everything in the system hangs together - the consumer benefits, the economic growth, the corporate profits, the technological advancement.'

Some years earlier John T. Connor (1966), then Secretary of Commerce, said:

'Without advertising, we most certainly could not have had the unprecedented prosperity of the last 67 months, because advertising is an absolutely indispensable element in the economic mix of the free enterprise system that produced that prosperity.'
'We would not have had, without advertising, a drop in unemployment from over 7% to less than 4%.'

And we would not have had, without advertising, a rise in unemployment since. Many of advertising’s critics from Professor Galbraith downwards also believe it has such powers - to create demand, to manipulate the consumer, to build our acquisitive society. But let us look at these supposed powers

Product class advertising as a whole - 'Buy more cars, ''Drink more tea,' and so on - certainly cannot be held responsible for consumer demand. For one thing, there is relatively little of this form of advertising. For another it generally has only minor effects, increasing a market by a few percentage points or slightly slowing a rate of decline. These effects are worthwhile to the producer, but neither can be credited with creating demand or manipulating the consumer on any substantial scale.

The primary target of criticism is repetitive advertising for individual brands - 'Buy Fords,' 'Drink Lipton’s Tea,' and so on. This is where the bulk of mass advertising is concentrated. Such competitive advertising for different brands can lead to a higher level of consumption of the product class as a whole than would exist without it, but there is no evidence that such secondary or even unintended effects are either big or particularly common. There are not even any dramatic claims in the literature (if I have missed one, that is the exception). In many product classes with heavy competitive advertising, total consumption is rising little if at all; in some it is falling. On the other hand, there are many product classes with little if any mass media advertising - like sailboats or marijuana - where consumption is increasing quickly.

Advertising for new products cannot bear the blame for consumer demand either. Undoubtedly advertising can help to speed up the initial adoption of a new product by creating awareness and, indirectly, by gaining retail distribution and display. But advertising works as a lubricant in such cases - to ease and speed things - and not as the prime mover. Getting an initial purchase for a new product is not the point at issue in understanding society’s continuing demand for goods.

The key question is whether people continue to buy something after they or their friends and neighbours have used it. This applies equally to frequently bought goods like frozen foods and cigarettes and to once only or once in a while purchases like atomic power stations or lawn mowers, where the satisfied users’ influence makes itself felt through word of mouth recommendation over the garden fence (or the industrial equivalent), through retailer and press comments, and so on.

By and large, one cannot go on selling something which people do not like after they have had it. Sometimes people are sold a new kind of product, by advertising or other means, which they find afterwards they did not really want. Some initial sales volume may be created in this way, but generally that is all.

The usual reason why people buy things is that they want them. Anyone who has washed dishes knows that the demand for nonstick frying pans or dishwashers did not have to be created. Rather, suitable products had to be developed, and then advertising undoubtedly helped to speed their adoption.

There is no need to suppose that the role of advertising here is fundamental.  It is a peculiar form of snobbism to suppose that if other people want to smoke cigarettes, to smell nice, to have bathrooms, or to drive in motor cars, it is only because they have been manipulated by advertising. Sometimes this view can go as far as John Hobson’s statement at his Cantor Lectures (1964):

'Almost certainly the increase in motoring has been the result of competitive petrol [gasoline] advertising.'

The alternative is to suppose that people want to go from A to B, or like driving, or want to get away for weekends, and that rightly or wrongly they often find cars more convenient or pleasing than walking or other forms of transport.

An often-quoted example of the alleged effects of advertising in 'creating' demand is the growth of men's toiletries. But this has been part of the great non-advertised change in men's fashions: clothing, hair styles, and so on. Advertising by itself could not have created such a toiletries market 20 or 40 years ago. Instead of leading it, advertising generally follows fashion or product innovation. Anything else would be bad marketing - spending millions to convince people to buy something just because someone can produce it.

The effects of paid advertising or consumer demand must not be confused with the effects of the mass media as such or with people’s developing education and greater mobility. People increasingly see how other people live and this has led to vastly increased expectations.

People 'want' many things once they have become aware of their existence - food, warmth, good looks, money, power, to drive a car, to be a concert pianist, to avoid washing dishes, and so on. Some of these things are very difficult to achieve, others are easier. To acquire goods, one only needs some money, someone to produce them, and a precedent of other people owning them in order to overcome cultural habits or inhibitions.

People go on wanting things because they like them.  Increased if highly uneven affluence, increased availability of products, and vastly increased awareness through mass communication and education are three factors which account for the growing acquisitive nature of Western society. The glossy images of affluence shown in advertisements and in the media generally reflect a real demand. Eliminating advertising would not eliminate the demanding consumer.

The products he demands are mostly genuinely wanted or even needed by him. Manufacturers seldom create the needs, but they do attempt to fill them. As a result we have competition and competitive advertising among different brands or makes of the same product. This we now examine in more detail.

Competition and Persuasive Advertising

Most advertising aims to promote a particular brand or make of product in a competitive situation. Because it often takes an emotional instead of an informative tone, such advertising is generally thought to work by persuasion. A typical critic like Boulding (1955), as quoted for instance by Achenbaum (1972) in this Journal, wrote in his economics text:

'Most advertising, unfortunately, is devoted to an attempt to build up in the mind of the consumer irrational preferences for certain brands or goods. All the arts of psychology, particularly the art of association are used to persuade consumers that they should buy Bingo rather than Bango.'

It is generally recognized that advertising’s effects on sales are not necessarily immediate or direct. Instead, it is thought to work through people’s attitudes as an intermediary stage to changing their behavior.

Advertising therefore is often thought of as aiming to attach an image or some special consumer benefits to a brand, in an effort to distinguish it from its competitors in the mind of the consumer. This is attempted especially in situations where there are no physical or quality characteristics to differentiate it. Gasoline advertising that stresses 'extra mileage,' or 'smoothness,' or 'enjoyment,' or 'power,' is a case in point, and Rosser Reeves’ Unique Selling Proposition (USP) was an extreme version of the view that advertising can only work by offering buyers of Brand X something which no other similar brand has.

In the last 50 years, various theories have been put forward to try to explain how advertising works, taking attitudes into account (for example, Joyce, 1967). One simple version is the well-known AIDA model, which stands for the chain:

Awareness - Interest - Desire - Action.

This sequential pattern - or something like it put in different words - is treated as common sense: It only says that people need to be aware of a brand before they can be interested in it, and that they need to desire it before they can take action and buy it. This imputes two roles to advertising: (1) an informational role - making them aware of the product - and (2) a persuasive role - making people desire it before they have bought it.

In its informational role, it might seem that when there are no deeper benefits to guide a consumer’s brand choice, he will be influenced by the last advertisement seen or by the general weight of past advertising. This assumption has led to the use of awareness and recall measures in pretesting and monitoring advertisements. But there is little direct evidence that advertising for established brands works like this. The evidence that does exist is either negative (for example, Achenbaum, 1972) or at best shows effects which are not dramatically large and which still require confirmation (for example, McDonald, 1970; Barnes, 1971).

In its persuasive role, advertising is thought to create a desire or conviction to buy, or at least to 'add value to the brand as far as the consumer is concerned' (for example, Treasure, 1973). For this reason advertisements take on persuasive methods like creating a brand image, selling a USP, or informing consumers that they need a special product to meet a special need (for example, a special shampoo for oily hair). But again, there is no empirical evidence that advertising generally succeeds in this aim, when there are no real differences to sell.

In fact, these models of hierarchical or sequential effects have been generally criticized on the grounds of lack of evidence (for example, Palda, 1966). They also fail to explain many of the known facts.

For example, they do not explain stable markets where shares of advertising and shares of sales are roughly in line for each brand. The small and medium-sized brands survive year-in and year-out, even though their consumers are exposed to vast amounts of advertising for the brand leaders.

Nor do the models account for the situation where, following a drop in sales revenue, advertising expenditure is cut and yet no catastrophe results. If consumers must be continually persuaded to buy a brand, then surely a cut in advertising should turn a minor setback into a major disaster. But it generally is not so.

Again, the models fail to account for the fact that four out of five new brands fail. There is no suggestion that failure occurs less often for highly advertised new brands.

More generally, the models do not explain why advertising generally has only a marginal effect on total demand for a product group; nor why it is only rarely capable of shifting people’s attitudes and behavior on social issues like smoking, racial discrimination, voting, and so on.

It is not enough to claim that persuasive advertising depends on the quality of the campaign or that advertising in general is inefficient. What is needed is a new explanation of the ways in which advertising actually works.

In recent years a good deal of attention has been paid to alternative explanations of the advertising process, based on mechanisms like satisfaction after previous usage, reinforcement, reduction of dissonance and selective perception. The argument later in this paper is grounded on these processes. But the most direct advances have been in our understanding of consumers’ buying behavior and attitudinal responses in a competitive brand situation.

Buyer Behavior

Brand choice and repeat buying are regular and predictable aspects of buyer behavior.

The economic viability of any frequently bought product depends on repeat buying. It follows simple patterns. If 10% of consumers buy Brand X an average of 1.5 times each in a given time period, then in the next time period 45% of that group can be expected to buy the brand again on an average of 1.8 times each (as modeled for example by the 'NBD' theory [for example, Ehrenberg 1972, Table B4]). This is what is normally found under a wide range of conditions, both for food and nonfood products, in the U.S. and the U.K., for leading brands and smaller ones, and so on.

The 55% who do not buy the brand in the second period are however not lost for good. Instead, they are merely relatively infrequent buyers of the brand who buy it regularly but not often. No special efforts have therefore to be made either to bring them back or to replace them (the 'leaky bucket' theory). Few things about the consumer in competitive markets can be more important than knowing this, and a successful theory of repeat buying was needed to establish it.

The existence of regular and predictable patterns of repeat buying for a brand however does not mean that people mostly buy one brand only. Instead, the majority of buyers of a brand regularly purchase other brands as well. In general there are relatively few 100% loyal or sole buyers of a brand, especially over any extended period of time. A typical and predictable finding for frequently bought grocery products is that in a week, 80 or 90% of buyers of a brand buy only that brand, that in half a year the proportion is down to 30%, and that in a year, only 10% of buyers are 100% loyal (Ehrenberg, 1972). To expect any substantial group or segment of consumers to be uniquely attracted to one particular selling proposition or advertising platform would therefore generally seem entirely beside the point.

Although many consumers tend to buy more than one brand, this does not signify any dynamic brand switching. Instead, the evidence shows that individual people have a repertoire of brands, each of which they buy fair regularly. Consistent clustering or segmentation of the brands over the whole population is however relative rare. When it occurs, it is usually above normal tendency for buyers Brand A also to purchase Brand compared with the patterns for all the other brands, rather than any special tendency for buyers of one brand no to buy the other (for example, Collins, 1972). But consumers generally buy brands which are similar as if they were directly substitutable.

In general then, repeat buying an brand switching patterns do not vary materially from one brand or product to another. A particularly simple result is that in a relatively short time period the frequency with which consumers buy a brand varies only marginally within the same product group. The main difference between a leading an a small brand is that the leader has more buyers. With ready-to-eat breakfast cereals, for example, consume make on average three purchases of brand over a three-month period. This varies between only 2½ and 3½ for different brands (Charlton, et al, 1972 and this small variation is itself high predictable from buyer behavior theory, with the larger selling brands being generally bought slightly more frequently by their buyers.

This is what occurs in relative short time periods. In periods which are very long compared with the product’s average purchase cycle (for example, year or more), the opposite sort of feet appears to operate because mc consumers will have had some experience of most brands (even if only single purchase). This leads to the vie that a brand’s sales can only increase people buy it more often (for example, Iresure, 1973). But in a shorter period like three months for cereals, high sales show themselves in terms of having to have more people buy in that period.

These various results are no long isolated empirical regularities but are becoming increasingly well explained and integrated into coherent theory (for example, Ehrenberg, 1972; Goodhardt and Chatfield, 1973). The theory applies primarily when a brand’s sales are more or less steady. This holds true most of the time - it is a basic characteristic of the market structure of branded frequently bought products that sales levels are not in a constant state of flux.

Occasional trends and fluctuations caused by promotions, etc., may be important from a marketing management point of view, perhaps adding up to 5% more sales in a year or 20% more in a particular month. But they do not amount to big, dynamic changes in consumer behavior as such. The individual’s buying behavior remains broadly characterized as being steady and habitual rather than as dynamic and erratic.

Attitudes and Attitude Change

Since on the whole there are no large behavioral differences among brands except that more people buy one than another, there are not many things that need to be explained by differing motivations and attitudes. In fact, attitudinal responses to branded products tend to be fairly simple.

The evidence shows that most attitudinal variables are largely of an 'evaluative' kind, plus some highly specific 'descriptive' differences for certain brands (Bird, et al., 1969; 1970; Collins, 1973; Chakrapani and Ehrenberg, 1974).

An 'evaluative' response to a brand is equivalent to saying 'I like it' or perhaps even only 'I have heard of it.' Evaluative attitudes therefore differ between users and nonusers of a brand, but they do not differ between brands. For example, 67% of users of Brand A say it has the 'right taste' with only 6% of non-users of A saying so about it, and 69% of users of Brand B that B has the 'right taste' with only five% of nonusers of B saying so, and so on, as illustrated in Table 1. Brand A may therefore have more people in all saying it has the 'right taste' than Brand B, but only because more people use Brand A, not because its users look at it differently: To give an evaluative response about a brand largely depends on whether or not one is using it.

TABLE: 1 TYPICAL 'EVALUATIVE' AND 'DESCRIPTIVE' ATTITUDINAL RESPONSES TO DIFFERENT BRANDS

 

Evaluative

Descriptive (for Brand C):

 

eg. 'Right Taste'

eg. 'Convenient'

 

User of the stated brand

Nonusers of the brand

User of the stated brand

Nonusers of the brand

Brand A

%

%

%

%

Brand B

67

6

19

3

Brand C

69

5

17

2

Brand D

62

4

55

48

etc.

60

3

17

2

Certain large exceptions to this pattern occur. These usually reflect some physical 'descriptive' characteristics of one particular brand. For example, if a brand is fairly new, consumers tend to be aware of this and dub that brand exceptionally 'modern,' compared with older brands. If one brand of indigestion remedies can be taken without water and the others not, people notice this and far more regard it as 'convenient,' as is illustrated for Brand C in Table I. Promotional policies can also make a brand appear 'descriptively' different: A slim cigarette advertised in women’s magazines as being smoked by feminine women may be rated more 'female' than a standard full-flavored cigarette packaged predominantly in red, with advertisements placed in sporting magazines and featuring cowboys.

A 'descriptive' characteristic is usually perceived also by people who do not use the brand. A 'female' cigarette will be seen so by people who smoke it and by those who do not. Nonusers of an indigestion remedy which does not require water will also regard it as exceptionally 'convenient,' as for Brand C in Table 1, but they nonetheless do not use it. 'Descriptive' differences between one brand and another therefore seldom relate to whether anyone actually uses the brand. 'Evaluative' responses on the other hand, while distinguishing between users and nonusers, generally do not differentiate one brand from another. Such results are therefore simple but not very helpful in explaining brand choice.

Attitude change 

The conventional results of research into consumers’ attitudes show how they feel about products, but not how they change their feelings. Very little work has been reported about changes in attitude. What work there is is difficult to interpret (Fothergill, 1968).

It seems to be generally assumed that improving the attitudes of a nonuser towards a brand should make him use the brand, or at least become more predisposed to doing so. But this amounts to assuming that people’s attitudes or image of a brand can in fact be readily changed, and that such attitude changes must precede the desired change in behavior. There is little or no evidence to support these assumptions.

The example of a successful change in image that is commonly quoted is for Marlboro cigarettes —few people volunteer another. Marlboro as a brand dates back to the turn of the century. It was considered a 'ladies' brand, at one stage holding a major share of the 'older society women’s market.' But in the 1 950s, Phillip Morris, the maker, started advertising it very differently, in a male, outdoor manner - Marlboro Country, the Marlboro Man, and the famous tattoo. Sales rose dramatically and Marlboro became a market leader. There is little doubt that Marlboro’s advertising had much to do with its success. But there is no evidence that the advertising created a change in 'image' or that a change in consumers’ attitudes caused the vast increase in sales. The explanation is much simpler.

The change in Marlboro was a change in product. The new Marlboro of the 1 950s was a standard tipped cigarette, full-flavored, packed in the new flip-top package, with a strong design, and introduced at the start of the growth of the tipped market (the tipped sector of the U.S. market grew from one% in 1950 to more than 60% by the mid-Sixties). For the first half of the century, Marlboro had been expensive, high quality, and with a pink paper wrapper (so as not to show up lipstick). No wonder people thought of it as different.

Subsequent attitude surveys in fact showed that smokers thought of Marlboro not as a ladies cigarette but as male, outdoor, for young people, for people with average jobs, and so on. But it did not have a special image - it differed little in these respects from other brands of similar product formulation. It scored extra on points where its advertising was played back (male, outdoor), but these differences - some 11 or 12 percentage points in a recent survey - were 'not as great as might have been anticipated,' to quote Stephen Fountaine, Phillip Morris’ director of marketing research. The change in Marlboro was real - it became a standard tipped cigarette - and not one merely in the mind of the consumer.

Other factors 

Conventional thinking about how advertising works rests on the sequence,

Awareness - * Attitudes - Behavior Although this appears like common-sense, various studies in social psychology have cast doubt on it. There are well-established psychological mechanisms which can act in the opposite direction - with behavior actually affecting attitudes.

For instance, behavior (the act of buying or using a brand) can lead to greater awareness of information to which one is normally exposed (selective perception). Behavior can even lead to the deliberate seeking out of information, and to changes in attitude (notions of congruence and reduction of dissonance). The well known illustration is the study where buyers of Ford cars were found to look at Ford advertisements after their purchases. This is common.

Usually a consumer is not convinced that a brand he has not bought before has all the advantages over the alternatives. To reduce the 'dissonance' between what he has done and what he knows or feels, he changes his attitudes after the purchase to make his chosen brand appear adequate.  He needs to do this even more if the chosen brand in fact differs little from the others, because there is then tangible reason or 'reward' to justify his choiceb - for example, 'maybe it is not vet good but at least it cost less?’

These processes are consistent with the known facts of consumer attitudes, such as those illustrated in Table 1. We will now see how they also fit into the broader picture.

'The scope of advertising depends on the ignorance of the people to whom it is addressed. The more ignorant the buyer, the more he relies on advertising.' (Scitovsky, 1951)

Brand Choice and the Consumer

The consumer’s choice among different brands or products is widely thought of as irrational and based on ignorance. This is how advertising is supposed to get its effect:

'The scope of advertising depends on the ignorance of the people to whom it is addressed.  The more ignorant the buyer, the more he relies on advertising.' (Scitovsky, 1951)

No one doubts or criticizes advertising’s role when it is a question of supplying basic information or creating awareness—for example, a house for sale, a job vacancy, a play at the theatre, or even for a new consumer product. But where advertising is regarded as persuasive rather than informational, there is criticism because of the view that the ignorant consumer’s choice is influenced by the last advertisement seen or by the brand image he is being told to believe.

But this is all wrong. Buyers of frequently bought goods are not ignorant of them. They have extensive usage experience of the products - after all, they buy them frequently. As we have seen earlier, they usually have direct experience of more than one brand, plus indirect word of mouth knowledge of others. The average housewife is far more experienced in buying her normal products than the industrial purchaser buying an atomic power station. She is also far less likely to make a mistake.

In regarding the private consumer’s brand choice as irrational, the view seems to be that if there is little real difference among the brands, then it is not possible to choose rationally among them. This ignores the fact that the consumer knows there is little difference and that he wants to buy the product. In choosing between similar brands, it is equally rational to choose the same brand as last time, or to deliberately vary it, or even to toss a coin. Any brand would do because the differences do not matter.

Just because Brand X is advertised as having some specific 'consumer benefit,' it does not follow that anyone buying that brand must have believed or been influenced by that aspect of the advertising.

In practice, people seem to find it simplest to develop repeat buying habits covering a limited repertoire of brands. Our task is to discover and understand the consumer’s reasons for choosing brands, instead of imposing our own preconceptions of how he ought to think and behave and dubbing anything else as irrational. The questions are: How do these habits develop, and what is advertising’s role in this?

ATR :Awareness, Trial and Reinforcement

Three main steps can account for the known facts of brand choice behavior: (I) gaining awareness of a hand, (2) making a first or trial purchase, and (3) being reinforced into developing and keeping a repeat buying habit.

Some initial awareness of a brand usually has to come first, although occasionally one may find out a brand’s name only after buying it. Awareness operates at different levels of attention and interest and can be created in many different ways, of which advertising is clearly one. Awareness may build up into the idea of looking for more information about the brand, asking someone about it, and so on.

A trial purchase, if it comes, will be the next step. This does not require any major degree of conviction that the brand is particularly good or special. Buyers of Brand A do not usually feel very differently about A from how buyers of Brand B feel about B, as was illustrated in Table 1. If that is how one feels afterwards, there is therefore no reason why a consumer should feel strongly about a different brand before he has tried it. All that is needed is the idea that one might try it. A trial purchase can arise for a variety of reasons: a cut price offer, an out of stock situation of the usual brand, seeing an advertisement or display, boredom, and so on.

After trying a different brand, people usually return to their habitual brands as if nothing had happened. This is so even when new purchasers have been attracted on a large scale, with free samples or an attractive short term promotion (for example, Goodhardt and Ehrenberg, 1969; Ehrenberg, 1972).

But sometimes a repeat buying habit develops. This is the crucial determinant of long term sales. The way this habit develops for a particular brand is primarily a matter of reinforcement after use. Any feeling of satisfaction—that the brand is liked at least no less than the previously bought ones - has to be nurtured. Evaluative attitudes have to be brought into line with the product class norms. But no exceptional 'liking' need arise,

because similar brands are known to be similar and the consumer does not inherently care whether he buys Bingo or Bango (which only matters to the manufacturer).

According to this viewpoint, development of a repeat buying habit remains a fragile process, probably influenced by a variety of almost haphazard factors. The consumer knows there is little to choose between, but he must choose. The critical factor is experience of the brand and no other influences seem to be needed. Thus it has been found that something close to the normal repeat buying habits can develop without any explicit external stimuli such as product differentiation or advertising (Ehrenberg and Charlton, 1973), and preferences for particular price levels can also develop without any external support or manipulation, just by trial and the development of habits (McConnell, 1966; Charlton and Ehrenberg, 1973).

But this process does not in itself determine how many people become aware, make a trial purchase, and are reinforced into a repeat buying habit. This - and hence the sales level of a brand - can therefore be influenced by other marketing factors, including advertising.

The Place of Repetitive Advertising

Advertising can act in the various stages of the ATR process.

Firstly, it can create, reawaken, or strengthen awareness. Secondly, it is one of the factors which can facilitate a trial purchase. For an established brand, the consumer may already have been aware of it and even have tried it, but this would have been in the past. The problem is that now he is ignoring the brand and may even be imperceptive of the general run of its advertising. Typically, a special effort like a new product feature, a new package, a new price or special offer, or a new campaign - anything 'new' is needed to give the advertising an edge for this purpose and be noticed. Obtaining awareness and trial for a brand is nonetheless relatively easy.

The difficulty is at the third stage, of turning new tryers into satisfied and lasting customers. This generally has to be achieved in the context of consumers already having a repertoire of one or more other brands which they are buying more or less regularly.

What happens in detail is not yet known - do heavy buyers of X switch to being heavy buyers of Y, or is this a gradual process, or is it the light buyers who are most easily affected? What is it in fact that advertising has to try and support or accelerate? The knowledge of buyer behavior outlined earlier puts some constraints on the possibilities, but this is one of the purely descriptive features of consumer behavior which is not yet understood.

The process can, however, seldom amount to manipulating the consumer. Real conversion from virgin ignorance to full-blooded, long term commitment does not happen often. A substantial leap forward in sales occurs only once in a while and sales levels of most brands tend to be fairly steady. Trends and even short term fluctuations tend to be smaller and more exceptional than is often thought.

The role of repetitive advertising of well-established brands is therefore predominantly defensive —to reinforce already developed repeat buying habits. The consumer tends to perceive advertising for the brands he is already buying, and repetitive advertising enables the habit to continue to operate in the face of competition. The consumer does not have to be persuaded to think of his habitual brands as better than others, but has to be reinforced in thinking of them as at least no worse.

This view of repetitive advertising - mainly a defensive role of reinforcing existing customers and only occasionally helping to create new customers or extra sales seems in accord with many of the known facts. It deals also with some of economists’ fears about the social costs of advertising and its possibly oligopolistic tendencies (see Doyle, 1968, for a review).

It is consistent with the fact that advertising by itself generally is not very effective in creating sales or in changing attitudes. It also explains why most people feel they are not personally affected by advertising. They are right. Advertising for Brand X does not usually work by persuading people to rush out and buy it.

The primarily reinforcement function of repetitive advertising is in line with the fairly steady sales levels of most brands in most markets. Advertising is not produced by evil men trying to manipulate the consumer. (If it is, these men must be very ineffective.) No one is more eager to cut advertising expenditure than the advertiser himself, who actually has to pay for it. For an established brand he sees advertising mainly as a price that has to be paid for staying in business, but he dare not cut it, and he is right (unless all manufacturers act together for example, aided by government edict, as in the case of TV advertising for cigarettes). For the consumer, large fluctuations in a firm’s market share would also not be helpful, in terms of availability, quality control, or lower prices.

According to the ATR model, increasing the amount of advertising would not by itself have much effect on sales, but cutting it is likely to lose sales. This is because some reinforcing action would be withdrawn, allowing competitive brands to gain customers more easily. For an established brand the loss of sales would by definition be quite slow, and no special theory of lagged effects of advertising is needed. Furthermore, reducing an advertising budget after a drop in sales to bring the two in line would not necessarily lead to any further substantial drop in sales. The ATR model is consistent with a more or less constant advertising to sales ratio.

The model also explains the survival of a small brand with a small advertising budget. For its users, the large amount of advertising for a larger brand which they do not use performs no function and generally is not noticed. When a consumer buys two more brands, some more heavily advertised than the others, each brand’s advertising primarily reinforces that brand and the status quo can continue.

High levels of advertising mostly occur in product fields where consumer demand is strong and the product’ easy to supply (because of low capital costs, or excess capacity). This leads I active competition and hence the no to defend one’s share of the marks either by price cutting or by heavy advertising.

Economists are frequently concerned that high advertising levels act as a barrier to entry for new brands and hence deter competition. This is wrong on two accounts. Firstly, it it the high risk of failure with a nevi brand that acts as the barrier - 'four out of five new products fail.' The barrier is spending a million and probably having nothing to show for it. Secondly, heavily advertised product fields are in fact characterized by heavy competition and a high incidence of new brands - but generally launched by firms already in the market. Simply having a million to spend on advertising is not enough; general marketing skills and experience of the other factors in the marketing mix (for example, a suitable sales force) are also needed.

Remaining Problems

The ATR approach outlined here is no more than a broad verbal statement of how advertising works that seems consistent with the known facts. Detailed quantitative flesh needs to he put on the model, hut its differences with the theory of persuasive advertising already raise many questions - for example, about the content of advertising, about the setting of advertising appropriations and the evaluation of advertising, and about product policy.

As regards content for example, use of attitudinal research results to try to improve one’s image or to produce persuasive messages of how Brand X is 'best' seem mostly to mislead the advertiser and critic rather than the consumer. Advertising research has failed to show that individual consumers think their chosen brands are necessarily better than the brands chosen by other consumers. The consumer needs merely to be told that the brand has all the good properties he expects of the product, and there can be a renewed emphasis on creative advertising telling a good advertising story well.

More generally, since consumers rightly see competitive brands in most product fields as very similar, it seems unnecessary to strive compulsively to differentiate brands artificially from each other. The clutter of marginally different brands, types, and sizes and the corresponding costs of product development and distribution may be unnecessary. This is not a plea for uniformity but for real research into consumers’ attitudes and motives to gain a better understanding of their, rather than the advertiser’s, needs for product differentiation.

CONCLUSION

Most mass media advertising is for competitive brands. It is a defensive tool and a price the producer pays to stay in business.

Consumers’ attitudes to similar brands are very similar. Purchasers of frequently bought goods usually have experience of more than one brand and they mostly Ignore advertising for brands they are not already using.

It follows that there can he little scope for persuasive advertising. Instead, advertising’s main role is to reinforce feelings of satisfaction for brands already being used. At times it can also create new sales by reawakening consumers’ awareness and interest in another brand, stimulating them to a trial purchase and then sometimes, through subsequent reinforcement, helping to facilitate the development of a repeat buying habit, ibis is the main determinant of sales volume.

The Awareness-Trial-Reinforcement model of advertising seems to account for the known facts, but many quantitative details still need elucidation. Such developments could markedly influence the planning, execution, and evaluation of advertising.

With persuasive advertising, the task might be seen as persuading the pliable customer that Brand X is better than other brands. Under the ATR model, advertising’s task is to inform the rather e xperienced consumer that Brand X is as good as others. The language of the advertising copy might sometimes look similar (still 'better' or 'best'), but the advertiser’s aim and expectations would differ.

This paper is based on a report prepared for the J. Walter Thompson Company in New York.

REFERENCES

ACHENBAUM, A. A. Advertising Doesn’t Manipulate Consumers. Journal of Advertising Research, Vol. l2, No. 2, pp. 3-13.

BARNES, M. The Relationship Between Purchasing Patterns and Advertising Exposure. London: J. Walter Thompson Co., 1971.

BIRD, M., C. CHANNON, and A. S. C. EHRENBERG. Brand Image and Brand Usage. Journal of Marketing Research, Vol. 7, 1969, pp. 307-314.

BIRD, M. and A. S. C. EHRENBERG. Consumer Attitudes and Brand Usage. Journal of the Market Research Societe Vol.12 1970. pp. 233- 247; Vol.13, pp. 100-1, 242-3; Vol. 13, pp.57-8.

BOULDING, K. Economic Analysis. New York: Harper and Row, 1955.

CHAKRAPANI, T. K. and A. S. C. EHRENBERG. 3 he Pattern of Consumer Attitudes.’ AAPOR Conference, Lake George, May 974.

CHARLTON, P., A. S. C. EHRENBERG. and B.PYMONT. Buyer Behaviour Under MiniTest Conditions. Journal of the Market Research Societe, Vol. 14, 1972. pp.171-183.

CHARLTON, P., and A S. C. EHRENBERG. McConnell's Experimental Brand— Choice Data. Journal of Marketing Research, Vol. It, 1973. pp. 302-07.

COLLINS M A. Market Segmentation - The Realities of Buyer Behaviour. Journal of’ the Market Research Society, Vol. 13. l971, pp. 146-157.

COLLINS. M. A.. The Analysis and Interpretation of Attitude Data. Market Research Society, Course on Consumes Attitudes, Cambridge, March 1973.

CONNOR, J. T. 'Advertising: Absolutely Indispensable.' Address before the Cleveland Advertising Club, Cleveland, Ohio. New York: American Association of Advertising Agencies. 1966.

DOYLE, P. Economic Aspects of Advertising: A Survey. The Economic Journal, Vol. 78, 1966, pp. 570-602.

EHRENBERG, A. S. C. Repeat-Buying: Theory and Applications. Amsterdam: North Holland; New York: American Elsevier, 1972.

EHRENBERG, A. S. C. Data Reduction. London and New York: John Wiley, 1974.

EHRENBERG, A. S. C., and P. CHARLTON. The Analysis of Simulated Brand-Choice. Journal of Advertising Research, Vol. 13, No. 1, 1973, pp. 21-33.

EHRENBERG, A. S. C. and for example PYATT (Eds.). Consumer Behavior. London and Baltimore: Penguin Books, 1971.

FOTHERGILL, J. G. Do Attitudes Change Before Behaviour? Proceedings of ESOMAR Congress, Opitija. Amsterdam: ESOMAR, 1968.

GOODHARDT, G. J. and C. CHATFIELD. The Gamma-Distribution in Consumer Purchasing. Nature, Vol. 244, No. 5414, pp. 316.

GOODHARDT, G. J. and A. S. C. EHRENBERG. Evaluating a Consumer Deal. Admap, Vol. 5, 1969, pp. 388-93.

HOBSON, J. The Influence and Techniques of Modern Advertising. Journal of the Royal Society of Arts, Vol. 112, 1964, pp. 565-604.

JOYCE, T. What Do We Know About How Advertising Works? London: J. Walter Thompson Co., 1967.

McCONNELL, J. D. The Development of Brand Loyalty: An Experimental Study, and The Price-Quality Relationship in an Experimental Setting. Journal of Marketing Research, Vol. 5,1968, pp. 13-19 and pp. 300-03.

McDONALD, C. D. P. What Is the Si Term Effect of Advertising? Proceedings of the ESOMAR Congress, Barcelona. Amsterdam: ESOMAR, 1971.

PALDA, K. S. The Hypothesis of a Hierarchy of Effects: A Partial Evaluation. Journal of Marketing Research, Vol.3, 1966, pp. 13-24.

SCITOVSKY, T. Welfare. Chicago: Richard Irwin, 1951.

NOTES & EXHIBITS
ASC Ehrenberg

A.S.C. Ehrenberg